I’m sure everyone has worked with a person who (somehow) is able to achieve more in 1 hour than the average person achieves all day and, conversely, people who seem to spend an entire week on something most people can get done before 10am on Monday.

All that being said, most people would agree that being present is at least a pre-requisite for being productive, i.e., nothing gets done if you’re not at work.

And so, typically, people spend their time trying to optimise their working day: streamlining their workflows, shuffling tasks into a more logical order, carving out blocks of “focus time”, etc., etc. That’s all sensible. Don’t stop.

But, as AI “agents” and tools like Claude Co-work become capable of running autonomously for multiple hours at a time, is this really the best way to optimise productivity?

By now, you can probably see where I’m heading with this.

Since everyone else has taken all the cool names, I’m going to default to my standard strategy (second-tier wordplay) and call this “Sleepworking”.

Quick Arithmetic

Some quick numbers to set the scene here. These are all approximations; I know everyone will have a slightly different work-life balance, but I believe they represent the majority of full-time workers reasonably well.

1. A human works ~40 hours per week (outside France).

2. An earth-week contains 168 hours.

3. An earth-week contains 128 non-working hours.

Just to emphasise how big a deal this could be, I’ve shown how big of a boost you get if you manage to successfully convert different proportions of your non-working hours into productive hours.

By Boost, I just mean how many unproductive hours you’ve managed to convert (second column) and how big this is relative to the average forty-hour week (third column).

Note: this is a MASSIVE simplification. It doesn’t factor in the productivity gains that AI gives you during the working day OR the time required to plan/review. It’s still directionally correct IMO.

And, theoretically, since AI agents can work in parallel (something you can’t easily do as an individual), the conversion rates could go beyond 100%. We’ll get to that later, but for now, I think we’re better of targeting a more conservative conversion ratio. Baby steps.

Ok, so how do you actually do this?

As Balaji so eloquently put it here, AI works middle-to-middle, not end-to-end. In other words, the work that AI probably can’t (or probably shouldn’t) do is usually at the start and end of a process — specifying and reviewing the final outcomes.

Note: I think everyone can (and should) figure this out as quickly as possible; I believe it will feel more natural to you if you’re from a management background and have experience delegating tasks, especially in technical domains.

This is a simple way to think about any significant piece of work…

In later articles, I’ll dive into each of these steps in a bit more detail and create some reusable instructions/prompts to help. For now, you should combine the thoughts below, along with AI and some conversations with your team, to figure this out.

Specification

You describe the required outputs in rigorous detail — the better job you do of this, the better result you’ll get: the content, the structure, the layout, any required conventions, in-house standards or tests, file formats, tone-of-voice, where to get data from, where to put data, colour palette, logic, rules, etc., etc., etc.

Of course, you might use Claude or ChatGPT to help you refine and ideate on this, but ultimately, the human is the one who needs to specify the goals and outputs required.

Plan

Once you know what you’re working on, you need to decide on the how. AI should be involved here, but you still need to review the details before letting it loose. Work with AI to help break down the work into smaller concrete sub-tasks; think about what sources of information or tools will be needed.

Think about any milestones or “checkpoints” where AI should save its current progress. Think about how AI finalises each task and marks progress against the bigger goal.

Remember, the idea here is that you won’t be present, so you also need to provide guidance on what the AI should do if it can’t achieve one of the sub-tasks. Think about how it should log its progress and decisions once it begins implementing the plan.

Most decent agentic tools have a Plan mode now — learn how that works and start there. In my experience, it won’t enable multi-hour runs, but it will get you off the starting blocks.

Implement

With a detailed plan in place and solid guardrails (see next section), you pull the ripcord, stand back and let the AI run. If your plan is well documented, the AI has access to the data, tools and connections it needs, and the guard rails have been well designed, it will should work autonomously to accomplish the goals you’ve laid out.

Validation

This is a BIG one. I think the two biggest reasons people fail to get good results when asking AI to run for long periods of time, in an unsupervised fashion, are:

• Incomplete specification

• Missing or vague validation processes.

You must provide a concrete set of instructions, rules, tests and standards in a way that AI can apply reliably. Validation should be done regularly at checkpoints along the way — not just once at the end. This keeps the AI on track.

• In software engineering, this might be a suite of tests.

• In copywriting, it might be in-house style guide, automated grammar and punctuation checking.

• For design, it might be analysing image composition or verifying colour choices and font-faces, etc.

I’m not going to list everything here; I can’t. I do plan to create some resources to help people self-serve on this, but for now, just think about what you would do yourself if you had delegated this sub-task to another person and wanted to verify what they’d done before they spent too long going down the wrong track.

Review

You must do this yourself. Of course, you may use AI to assist, but fundamentally, if you care about the final output and want it to be done to a high standard, you are responsible for reviewing it yourself.

Use AI to help you approach the review phase in a sensible, logical order (e.g., you might start at a high-level and work into the lower details). Ask AI to point out important decisions that were made and any trade-offs involved.

You also, obviously, need a way to incorporate the review feedback — this will depend on the feedback itself:

• If it’s small, you might edit the work yourself, or ask AI to fix up a few things.

• If it’s bigger, you might specify and plan another cycle of work (this time with enhanced guard rails or new requirements based on the review outcomes).

Timeline

This exact approach was inspired by the Elves framework by John Ennis at Aigora. That Elves framework is targeted at software engineering work, but it’s generic enough that it can be reworked for many long-running, autonomous workloads.

What’s Next

I’m going to stop here as there’s a lot to think about here already.

If you’re still using AI in a transactional/conversational style, you must take the time to explore these more autonomous, agentic tools (Claude Co-work, Claude Code, Microsoft Co-work, Manus, Codex etc.).

My goal here is to adapt what I’m seeing in the engineering space and bring it to a wider audience. If you’re a manager or someone with lots of experience delegating, this will be an easier ride, but it will still be a lot to take in.

If you are from a technical background yourself, I’d encourage you to take a look at the following resources as they serve as the basis for this series.

• Human Layer

• Factory AI

• Elves Framework

Until next time.

If a picture speaks a thousand words, then a custom tool, built in real-time might speak at least ten times that!

Before you carry on, if you prefer to watch instead of read, I’m experimenting with video. Many are mash-ups of videos I send clients, others are new.

Whilst the tech community accelerates into the brave new world of agents and AI driven software-engineering, I’ve found there to be a bit of a content gap in how other smart people (outside this bubble) get started.

Perhaps the simplest way to put this is: it’s now possible to build useful software tools without writing a single line of code. If you can articulate something in natural language, you can now build a tool for yourself or your business.

And, though I’d push back on the narrative that you should go out and replace your entire software stack with a vibe-coded hobby project, there is immense value in learning how to solve firm-specific problems, or finally building out those 'nice-to-have' tools that never made the budget cut.

Whilst many engineers are cynical of this new trend, it doesn’t change the fact that, for the first time ever, people without any background in computing whatsoever can create software.

I do believe it’s easier for those with an engineering background to build applications with AI but it’s certainly possible for everyone to build something at this point.

So, what should you be building?

Well, I don’t think there is any one-size-fits-all approach here, but there are two (repeatable) patterns that seem to have worked really well for the SMB leaders I’ve been working with. Neither require a lot of time, money or resources to get started and, in most cases, so long as you can articulate your problem clearly, you’ll likely come out the other side with something pretty impressive.

Pattern 1 - Unlocking Underutilised IP

David is the MD of a fantastic US-based consultancy that helps other consultancies grow, win more clients and become more profitable. He has authored several books, regularly appears on podcasts and has an extremely large corpus of high-quality IP related to business development and scaling consultancy firms.

I’m working with David’s firm to look at both his internal operations and how we can use the firm’s existing IP to create even more value for his clients.

A few days ago, David asked how feasible it would be to create a tool he’s previously tried to build (several times) without a huge investment in time. The tool - “Vizidiz” - is a visual decision making tool that helps simplify complex decisions and rank candidates along multiple dimensions.

• You’re assigning employee annual bonuses for dozens of staff and want a consistent approach to evaluating each person.

• You have 10 vendors bidding for your new contract and want to make sure you pick the vendor that is, on balance, the best fit.

• You have 4-5 technical approaches for a given engineering problem and want to rank-order them.

• Or purely personal choices, like rank-ordering houses you want to buy based on safety, green space, and local amenities.

Because he’d never been able to produce a tool he was 100% happy with, this IP was just sort of sitting there. Such a shame.

David was able to provide a recording of an old presentation he’d put together on the topic, and sent me a couple of clarifying emails on the finer points.

Only 15 minutes later, we were using the first version of the tool! From here we can incorporate it into the firms internal operations, product offerings and other learning materials.

I’ve left the original version we created here (warts and all). Try it out!

The point here is not the tool itself, it’s that the cost of running this experiment was so small that it completely changed the risk-reward ratio of this entire initiative.

Two years ago, this sort of experiment would have taken too much time away from other initiatives that it simply wouldn’t have been done. The value would have never been unlocked and a good tool would have continued to exist in theoretical-form only.

Where could you do this in your own business or team? What’s stopping you?

Pattern 2 - Back of the Envelope Analysis

Whilst David works predominantly in the knowledge-realm, Mark works mainly in the physical realm of nuts and bolts. Mark has a history of investing in alternative and distressed asset-classes and is constantly looking for ways to extract value from otherwise wasted resources.

I think it’s fair to say that Mark is still very early in his AI adoption journey - he knows it can be extremely powerful but is still figuring out how to extract that power in order to improve his business outcomes.

Naturally, in the physical realm, a failed project or experiment can be costly - in some cases, enough to put someone out of business. And, whilst we like to think, there is a lot of complex modelling behind every decision - the reality is that there just isn’t the time!

In this particular case, the question was whether there was any merit in the idea that large electrical batteries could be used to smooth out retail electricity prices (e.g. charging the batteries at night and consuming/exporting during the day).

The team needed a quick “back of the envelope” answer to determine whether it made sense to explore this investment at all, or whether it was simply a fool’s errand.

Mark spent some time trying to figure this out himself with a bit of help from Google and a little bit of AI support but he wasn’t getting anywhere fast. The issue in this case was that Mark was trying to use his AI assistant as a calculator and numerical estimator. If you have used AI for any length of time, you know that LLMs are terrible at arithmetic! They are language engines, not calculators.

I suggested that, instead of asking the AI for the answers conversationally, Mark should ask the AI to write the code (as an interactive app) that would allow Mark to explore the problem in more detail. The AI writes the logic, but the code does the math. Essentially, we introduce an additional step so that:

Fifteen minutes later (and with a few more pointers), Mark and his team had built their first (simple) interactive tool. After a few more iterations, they ultimately decided this particular idea didn’t have any legs but, perhaps more importantly, the team had added a powerful new approach to their tool-belt and, I’m happy to report, they’ve built more since!

Warnings

If it’s not already obvious to the reader by this point, there are of course some risks to this approach. The main one being that there is every possibility that AI is translating your words and requirements into the wrong code and thereby giving you a false sense of security/accuracy.

This is a hard problem to solve completely and there’s a hell of a lot to say on this topic. That being said, in some ways this approach can actually be more transparent than using AI in a conversational style; the internals of the tool are available for inspection and, even if you can’t write computer code, it’s often possible to sanity check the key calculations involved (or even ask the AI to elaborate).

For now, just know that these tools are not always perfect and that caution must be exercised when using any AI output as part of a critical workflow or decision.

But please don’t let that put you off having a go! Almost every firm I work with has tricky estimation problems or fantastic IP to unlock. What they don’t always have is the time to do this stuff. Now they do.

P.S. I’m experimenting with video posts right now. I do these a lot for clients and figured there’s no reason not to have a go here. It goes into a bit more detail and includes a realtime example of tool-creation.

As someone working in the space, the outlandish claims that land (on an almost hourly basis) often do more harm than good. Of course, it’s good to be positive but the level of exuberance and click-baiting has killed the signal-to-noise ratio and made it very difficult for clients like mine to discern what is (and is not) viable with current technology. I love AI, and I’ll be writing a lot more about it … but we should try to stay grounded in reality.

All that being said, we are clearly in the midst of an absolutely enormous social, technical and economic rewiring. There’s a hell of a lot to say on the topic but for now, I want to throw out a few takeaways based on how this rewiring is unfolding across the small businesses I’m working with.

Erm. So… Are we cooked?

Alongside the optimistic buzz that comes along with every new engagement, there is also (normally) a moment where we step across the bright line from hope, into worry.

Naturally, no two clients are the same but you could probably guess some of the big ones…

• Security and Data Protection

• Process Disruption (the destructive kind)

• Internal Resistance to AI

• Mistakes in Critical Outputs/Artefacts

However, if you talk to the business owners and key decision makers, this pales into insignificance behind a bigger worry…

• Yeh. So…anyway…is my business about to become obsolete?

This is especially true in the professional services space that I tend to work in. There’s a genuine (and very rational) concern that customers will choose to use ChatGPT, Copilots and Agents in place of the services, skills, and experiences that the business has spent decades developing.

Small businesses worry that their edge will be commoditised, that the “big players” will just throw money at the problem until they win, and that they’ll be left picking up the crumbs.

So, time to pack-up shop and retrain as a plumber? No. Not yet at least.

Small but mighty

The small businesses I work with are often better positioned than they realise. Their edge lies mainly in the things AI can’t easily replicate or train on - the trust they’ve built, the nuanced social dynamics, the relationships they’ve nurtured, and the unique knowledge acquired from diving deep into a particular niche.

Not to mention that small businesses are nimble in a way that larger organisations envy - they can adapt faster, experiment sooner, and thread A-grade tools into their workflows without a 400-year roll-out plan (if you’ve ever been committed to using sub-par tools by the board of a large organisation - you know this pain). Decision makers in small businesses typically feel the consequences of their decisions more directly - it’s why I like working with them.

Anyway, I know you’ve probably heard something like this before and I’m not intending to regurgitate stale AI tropes but I really do believe these are the super powers of small business. Especially relative to the faceless Gigacorps.

With the above in hand, add in some thoughtful automation and AI tooling to get acceleration, dexterity and leverage - not obsolescence. Firms that lean into that combination usually end up doing more of what they’re great at, not less. Naturally, this will likely change the shape of the business as certain departments shrink and others grow … but it will not eliminate it.

Like most things, this is an opinion based on the information available right now. Of course, some revolutionary new innovation could (and probably will) change the game (again) but, even then, it takes a long time for this sort of technology to diffuse through the economy (I’m still waiting for the self-checkouts to reliably weigh my loose vegetables).

We have to face the problems in front of us today and, with the current rate of change, anyone telling you they know how things will look in 5 years is dishonest or deluded.

Personal Reflections

Recently, I’ve been fortunate enough to start working with an internationally renowned advisory firm (I would put them squarely in the “elite” tier of small businesses). They know their stuff; they regularly appear in podcasts and newspapers, they’ve published best-selling books, and some of the most elite businesses in the world have been paying for their advice for decades.

By doing all the above. Consistently. For decades. They have established a huge network of high-value clients that really trust them. With that solid foundation to build on, I’m now working with them to introduce AI and Automation, and I have extremely high hopes for where this is going to take us.

From my own side, I’ve learnt a lot from working with this business. For whatever reason, I’ve never prioritised networking, public writing or any other form of “authority building” - part of that is my character, part of it is circumstance. I’m trying to address that now as I think it’s necessary for the success of my own firm. We live and learn.

If you’re a small business owner and you already have all the foundational stuff in the bag … I think you’re really, really well placed to be a beneficiary of AI rather than a casualty.

Be courageous! Go for it!

On April 28, 2025, the lights went out across Spain and Portugal. Not a flicker, but a near-total collapse, plunging me and ~55 million of my Iberian friends into darkness and chaos. Trains stopped, traffic lights died, hospitals scrambled onto backup power, mobile networks went silent, and the internet became unavailable. It was one of the most severe blackouts in recent European history.1

Aside from a small number of (very brief) power-cuts in the UK, this was a totally new experience for me and a fantastic reminder of how reliant we have become on electrical and communication networks in general.

The electrical grid is one of the most complex structures built by mankind; the more you learn about grid engineering, the more miraculous it becomes. The fact that when you flip a switch, you can expect, with high confidence, that you’ll receive magic electrical power has amazed me ever since I learnt about it as part of my Engineering degree.

So, what better time than a multi-national scale outage to dust off a few mental cobwebs and work out what the hell just happened?

Although no one has all the answers right now, what we do know reveals a story not just about a single failure, but about the inherent fragility of our modern power grids as we race towards a renewable future without properly addressing the engineering realities in front of us.

The Collapse: A Five-Second Catastrophe

Just after midday, Spain’s grid operator, Red Eléctrica (REE), detected an event (we’ll call it N-1) - likely a loss of generation. The system momentarily stabilised. Then, 1.5 seconds later, a second event (N-2) hit. The instability escalated. Within another 3.5 seconds - a mere five seconds from the initial trigger - the crucial power line connecting Spain to France automatically tripped, isolating the Iberian Peninsula to protect the rest of Europe.

Cut off and destabilised, the Iberian grid haemorrhaged power. In those same five seconds, 15 gigawatts of generation tripped in Spain (60% of its demand at that instant), along with 5 gigawatts in Portugal. The grid frequency, the vital heartbeat of the system, plummeted. Protective systems across the peninsula fired, generators tripped offline, and the entire network cascaded down into collapse.

What’s interesting here is the occurrence of two generation loss events (N-1 and N-2) at the same time; generator losses are widely considered to be uncorrelated, independent events, and to see two should be extremely rare! In fact, situations like this are so rare that grid operators generally only consider the possibility of a single generator event when designing contingency plans.

Details on the specifics of these events have, disappointingly, not been provided yet.

Grid 101: The Never-Ending Balancing Act

To understand what happened, a quick Grid 101. As I said earlier, power grids are extremely complex machines demanding constant, perfect balance. Generation must instantaneously match consumption (demand). Too much generation, and the grid’s frequency (normally a rock-steady 50 Hertz in Europe ) rises. Too little, and it falls. Voltage, the electrical "pressure," must also stay within tight limits. Deviations risk damaging everything from your fridge and home appliances to the grid infrastructure itself.

Inertia: The Grid’s Unsung Hero

But there’s another crucial factor: inertia. Think of it as the grid’s physical shock absorber. Traditionally, this inertia comes from the massive, spinning turbines and generators in conventional power plants (coal, gas, nuclear, hydro). These heavy rotating masses store kinetic energy. When a disturbance hits and frequency starts to change, this stored energy is automatically released (if frequency drops) or absorbed (if frequency rises), resisting the change.

Crucially, inertia slows down how fast the frequency changes (the Rate of Change of Frequency, or RoCoF). This buys precious seconds for slower control systems to react and stabilise the grid. High inertia means stability; low inertia means fragility.

The Renewable Conundrum & The Trigger Debate

Here’s where it gets complicated (and a bit political). Solar panels and most modern wind turbines connect via inverters. They don’t have those large, spinning physical masses. So, as we replace conventional plants with these inverter-based resources (IBRs), the grid’s natural inertia decreases.

On April 28th, renewables hit nearly 80% of Iberian generation, with solar alone near 60%. This meant the system was operating in a critically low inertia state.

This specific point has been seized by the anti-renewables contingent as a means of dismissing solar and wind as a problematic, impractical source of energy, with the implication being that we should use common sense and go back to good ol’ synchronous generation (Gas, Coal, Nuclear), etc. And, though I think the case for Wind power is dubious (at best) - this largely feels like unhelpful point scoring right now.

On the other hand, the more ideological supporters of renewable energy have been quick to dismiss these claims as biased junk science. It isn’t.

So, was low inertia the cause? Well, right now it’s unclear. Experts are divided. Many point to the initial N-2 contingency - two generators tripping almost simultaneously within 1.5 seconds - as the real, highly unusual trigger. No system, they contend, is designed to withstand such an extreme event, especially when isolated from European support.

However, it’s delusional to downplay the possibility that low inertia played a critical role in the consequences. Once the N-2 event happened and Iberia was islanded, the frequency likely dropped so fast (high RoCoF) in the low-inertia environment that protective relays on a massive 15 GW of other generation tripped almost instantly. These relays are designed to protect equipment from damagingly fast frequency changes.2

So, while low inertia might not have caused the initial N-2 trigger, it almost certainly made the system vulnerable to a rapid, total collapse following that trigger. The grid lacked the resilience to cushion the blow, and highlighting this doesn’t make someone “anti-renewable.”3

Ignoring the Alarm Bells

This fragility wasn’t a secret. The warnings were clear and numerous:

• ENTSO-E (Europe’s TSO Network) repeatedly warned that declining inertia increased the risk of system splits and blackouts.4

• REE/Redeia (Spain’s Operator/Parent Co.) flagged the need for more storage and interconnections back in 2020. As recently as February 2025, Redeia warned investors of "generation disconnections due to the high penetration of renewables" and reduced inertia.5

• CNMC (Spain’s Regulator) reported in Jan 2025 that grid voltage levels were already hitting dangerous thresholds.6

• Experts warned about phasing out inertia-providing plants too quickly and the "unplanned and haphazard integration" of renewables without adequate grid upgrades.

The risks were documented. The question is why they weren’t adequately addressed before the system reached breaking point.

Politics, Spin, and the Price of Stability

In the blackout’s aftermath, the predictable blame game erupted. Government officials and REE’s leadership publicly denied that renewables were the issue, sometimes contradicting earlier internal warnings. Opposition parties accused the government of prioritizing green policies over grid security.

Was it a cover-up? Maybe not active concealment, but the inconsistencies suggest significant “narrative management”. Acknowledging that the pace and manner of the energy transition, without sufficient parallel investment in grid stability, created the conditions for collapse is politically difficult. It forces uncomfortable questions about who pays for grid upgrades and stability services, costs often postponed or passed to consumers.

It's Not Hopeless: Modern Solutions Exist

The inertia problem doesn’t mean we ditch renewables. It means we need to engineer solutions alongside them. Fortunately, the toolkit exists:

• Synchronous Condensers (SCs): These are essentially large rotating machines run without fuel, specifically designed to add physical inertia and voltage stability back into the grid. Think of them as spinning stabilisers. Deployment is already happening, including by REE on Spanish islands.

• Fast Frequency Response (FFR) & Synthetic Inertia: Using smart power electronics, IBRs (like solar, wind, and batteries) can be programmed to rapidly inject or absorb power in response to frequency deviations, mimicking inertia’s effect.

• Grid-Forming Inverters (GFIs): The next generation of inverters. Instead of just following the grid’s lead, they can actively form it, establishing their own stable voltage and frequency, much like traditional generators. Crucial for very high renewable grids, but large-scale deployment is still emerging and untested.

• Energy Storage (Batteries & Pumped Hydro): Essential partners. Batteries provide the ultra-fast response needed for FFR and synthetic inertia. Pumped hydro offers large-scale storage and traditional inertia. Often scoffed at by renewables sceptics as impractical, decades of work on battery tech has made this more feasible.

The point is: we can build stable, high-renewable grids. But it requires deliberate investment in these solutions, not just assuming the grid will cope. It also requires we move past some of the unhelpful ideologies and childish hysteria surrounding topics like Climate Change.

As much as a successful renewable energy future is something everyone should be able to get behind, fossil fuels and nuclear are not the devil. Of course, carbon emissions and pollution are a problem that we should continue to address, but let’s not forget that those very same molecules are responsible for lifting more people out of poverty than virtually anything else in history.

The ridiculous idea that energy use is a bad thing has warped so many minds that we’ve strayed into some very strange places psychologically (remember those articles telling us that the best thing to do to reduce carbon emissions was to avoid having kids … I mean, seriously, do people fall for this stuff?).7

It’s fine to toss around a few dumb ideas in WhatsApp, but when it comes to designing the power grid, let’s try to be a bit less silly, please.

Bright Side

To round out this fairly concerning piece, I wanted to add something a bit more upbeat.

After a few hours of power loss and especially once the internet was lost, I noticed a real change in the behaviour or people out on the streets - without access to a super-stimulatory device in their pocket, people were chatting (to strangers) more and engaging in more “wholesome” activities - groups in the park, reading books on the balcony etc.

Drivers were also being quite considerate of pedestrians, and there was a general feeling of community and people uniting around a shared problem.

It was a nice (though brief) reminder that switching off is good for us sometimes (and that I need to make sure I always have access to an FM Radio, powerpacks, and firelighters!)

Summary

The Iberian blackout is a powerful reminder (pun intended). Renewables are fantastic, but they aren’t plug-and-play replacements for the inherent stability that conventional plants have provided for decades. The transition demands massive, proactive investment in grid modernization, storage, and specific stability technologies.

Ignoring the physics, downplaying risks, and underinvesting in resilience is, as one analyst warned, like “playing Russian roulette with the grid”. On April 28th, the chamber wasn’t empty. We need to ensure it doesn’t happen again.

Automation Consultancy ⚙️🚀

Codefield helps businesses build and implement automated solutions for an extremely wide range of challenges.

Regardless of sector or size, we find that there are many, many places where leveraging Automation and AI can help businesses reduce (or even eliminate) the time spent on repetitive and/or tedious tasks, allowing them to focus on the more valuable strategic work that takes the business forward.

If you think there are areas of your business that could be automated or streamlined, then feel free to reach out for a friendly (and in no way pushy) chat at automation@codefield.io

Week One

Remembering that I included April 1st in last month’s numbers, the first week started on April 2nd and, though I had started pretty strong, it fizzled out quickly. Honestly, I’d put this down to prioritisation and lack of effort on Twitter.

This first week was dominated by April 2nd (riding the wave from March), which kept the follower numbers elevated despite little engagement for the other six days. After Week 1, I was at about 31% of the monthly goal. Not bad.

Week Two

This week was quite disrupted, and the results reflected that. Between some other responsibilities, a new 3D graphics project, and some unforeseen family events, I never really gained any momentum with my posts and replies.

As you can see, it was slightly better than the last half of week one, but because there were no strong days to offset the mediocrity, it was a bad week. I netted 19 new followers and crawled along to reach ~45% of my monthly goal.

Week Three

This is where the wheels really came off. I deprioritised the social media activity to work on things I thought were probably higher ROI (marketing, 3D graphics projects, family, Portuguese). I don’t regret it, but it does reiterate the fact that consistency is a huge factor in building an audience. You continuously lose a small number of followers, so if you don’t gain more than that, you’ll be net-negative.

In truth, this week barely changed the overall numbers; there was no chance of hitting the monthly target by the end of week three. Ah well. Can’t have it all.

Week Four ⚡️🪫

This week continued much like week three: minimal input from me, and a loss of inspiration for what to post and engage with. Results were quite dire until Monday, when, along with ~55 million of my Portuguese, French, and Spanish pals, I received a last-minute invitation to take part in the second-largest power cut in European history!

Despite losing access to electricity, mobile coverage, and (perhaps most horrifyingly) internet access, for the brief period I was able to stay online, I had a great time!

For around thirty minutes, there was very little mainstream media coverage of what had happened, and so, with Twitter being perhaps the most up-to-date platform for news, people were jumping online and working together to find out what was happening.

The situation was very rapid fire - lots of diagnosing issues and trying to triangulate who did (and did not) have power. Reminded me of work in some old jobs in some ways.

I’m going to do a proper write up of this shortly (I’m a bit of a power-grid geek) but, despite only having access to the internet for about one hour, I was able to consolidate information quicker than many others, broadcast it out quite quickly and pick up a decent number of follows from people who were not getting answers from elsewhere.

Though I do post about energy and infrastructure, I suspect many of these new “emergency” followers may be short-lived. Let’s see.

So in total, after that crazy episode, I had scooped up an additional 23 permanent followers. Bringing me to exactly 66% of my target (40% total growth)

Summary

Honestly, I think I could have hit the 60% growth this month if I’d been more dedicated - but obviously … I wasn’t 😆. Next month I have a lot of travel, and I think it’s unrealistic to expect better than this month.

At the risk of permanently lowering my standards, I am going to aim for ~30% growth in May - I need a win, and there are just a lot of other things I need to prioritise.

The main goals for this month are:

1. Finish the MVP of my 3D Graphics Project - this is an idea I’m hoping will grow my audience significantly, and even if it doesn’t, it’s super interesting for me.

2. Dust off my electrical-engineering books and do a write-up on the European power cut. There’s a lot to learn here, and it’s been a while since the second-year module on Electrical Power and Engineering.

3. Continue my automation-services journey. Ideally, I’d like to acquire a real customer/client for this side of my business in the next 2-3 months. It will be hard going, but it’s a fascinating field.

Okay, that’s it.

Stay safe (and keep those battery packs charged)

Automation Consultancy ⚙️🚀

Codefield helps businesses build and implement automated solutions for an extremely wide range of challenges.

Regardless of sector or size, we find that there are many, many places where leveraging Automation and AI can help businesses reduce (or even eliminate) the time spent on repetitive and/or tedious tasks, allowing them to focus on the more valuable strategic work that takes the business forward.

If you think there are areas of your business that could be automated or streamlined, then feel free to reach out for a friendly (and in no way pushy) chat at automation@codefield.io

Oh! I had a wisdom tooth removed too so I guess I lost a few IQ points.

The hurrier I go, the behinder I get.

As I mentioned last time, March was always going to be a tricky month; I had some personal travel to the UK, a new product to build, and a significant number of chores to get through!

After my (pleasantly) unexpected 186% follower growth in February, I was confident my approach was solid and didn’t need any major tweaks. The initial plan was basically the same:

1. Engage with a mixture of small, medium, and large accounts.

2. Stay away from rage-baity topics and mud-slinging.

3. Generate a couple of pieces of interesting and/or valuable content.

For the most part, I think elements (1) and (2) were achieved, and (3) was not. Without making too many excuses, I think this deviation from the plan was due to an extended period of travel to the UK and some last-minute dental surgery that required a bit of logistical juggling.

Let’s have a look at what has happened since last time.

Same old, Same old

At the start of March, I carried on as planned - engaging with a mix of small, medium, and large accounts across my interests. This provided steady growth and I was still benefitting from the tail-end of the widely-shared Stripe post.

After one week, I had gained 22 followers and lost 5 - so +17 followers.

Given my 60% growth target, I was 22% of the way through March and about 17% to my target. This wasn’t too alarming, especially since things seemed to be picking up towards the tail end of the week.

Distracted by Building

In the second week, I began to start building the MVP for another MicroSaaS product idea I’d had at the end of Feb. A tool for improving the Stripe Invoicing download workflow…

Stripe only provides users with the ability to download Invoice PDFs 1-by-1 and this gets very tedious in large numbers. I’d previously written a script for this and it felt like the sort of tiny tool I could (a) build quickly and (b) sell to a frustrated business for a few quid a month.

By adopting the latest AI coding-tools, leveraging some code I’d used for earlier projects, and adopting a less perfectionist attitude, I was able to build and launch the MVP in under 10 days - blazingly fast tbh.

P.S. For those interested, this MVP is now live » Invoice Powerup 🚀

Unfortunately, one unintended consequence of building this product so quickly, was that I didn’t maintain my efforts on Twitter and engaged less and less as the week went on.

Unsurprisingly, this lack of engagement led to pretty poor growth.

After adding 17 users in the first week, I’d only added 6 users in the second week and was now waaay behind the target (I should have been at about 45% by this point)

Jetset Lifestyle (on a Shoestring)

Normally, falling this far behind the target would have been a reason to re-evaluate and perhaps throw more time and energy into the mission. This wasn’t really possible as I had already arranged a 7-day tour of the UK to visit friends and family.

For all the right reasons (family and friends are actually important y’know!) - Week 3 was even worse than Week 2. I barely posted or replied to anything and this was reflected in my account metrics.

Honestly, at this point, I think I knew the gig was up and that March was destined to be an underperforming month in terms of account growth. I was now 67% of the way through March but only 27% of the way to my target.

Thank God for Anesthetic

As part of my UK roadshow, I also dropped into my old (and more trusted) dentist - there are a lot of good things about living in Portugal but having to navigate dental surgery in a second language is not one of my favourites!

Bad News: The dentist confirmed that I was overdue for a wisdom tooth extraction. Good News: The dentist said he could expedite my extraction if I could extend my visit for a few days.

I was keen to get the extraction done with the dentist I knew well BUT this meant I would now be in the UK for another week and, unless something drastic changed, I was probably not going to reach even 33% of my target (20% total growth)!

Despite the gloomy outlook, the ad-hoc nature of the trip extension did allow for more time to work on my own stuff (I didn’t have many planned visits in the calendar and everything was a bit more informal).

Alongside putting some finishing touches on the MVP, I tried to get back on track with my Twitter engagement and the results were actually … very good. It’s hard to know exactly why this week was successful but I think it was led by a single “funny” reply that started to receive thousands of likes, drove users to my profile and, ultimately, to follow me.

All that being said, there was no way a ‘good’ week was going to offset the terrible performance in the previous two. Nevertheless, I’m playing the long game and followers are followers. It’s all good.

I picked up 27 followers and lost 6 for a total gain of +21. A good week but, sadly, with the month virtually over, I hadn’t even hit half of my target 😔

Content is King

Without time to create any novel content, big threads, articles, or videos, I was prepared to walk away from this month at 48% of my target. You can’t do everything and life without friends, family (and painless teeth) is just not worth it.

But then, just as I’d started to pencil in a time to write this progress-report, a massive account posted about something I’d previously created content for…

The AI company Anthropic (creator of Claude) released a video a few months ago about how to write “stellar” prompts for LLMs; for whatever reason, the big account was sharing this video now (months later) and it was being well-received by his massive (200k+) follower-base.

Fortunately, it just so happened that many months ago, I’d been working through Anthropic’s Prompt Engineer Guide (if you’re techy, check the Jupyter Notebook out here) and had created a stylised illustration and summary of the guide.

The first time I posted this, it basically went nowhere and I’d written it off as a bad experiment but, since it was so directly applicable to the content posted by the big user - I thought I may as well add it into the replies.

BOOM! 💥

Completely unexpectedly, this reply turned into my most successful reply ever (in terms of followers). Hundreds of high-quality accounts started to like, bookmark, and repost the material and the followers started to flow in.

Disclaimer: Technically some of this happened on April 1st but I’m going to include it as (a) it was POSTED in March (b) the mostly-American users liking it were still IN March and (c) it’s MY blog and I don’t care what YOU think (I do really, please don’t unsubscribe!)

It still wasn’t enough to hit my monthly goals but it was a massively positive finish to an otherwise disappointing month 💪

You can see the massive jump on Apr 1 as my content was picked up on the US side and followers started to drop in.

In just 4 days, I picked up 28 followers and lost 2 for a gain of +26. As well as being a much-needed rescue package to bring the March numbers up - the final day was my best day ever in terms of new followers.

This last day brought in almost double the previous record set back in January!

And, although I didn’t quite hit my target of 60% growth I still grew by 44% (about 73% of my monthly target).

More than just the numbers, I’m now starting to see some return on investment when repurposing existing content - this is a clue that some of my content is actually quite valuable to people but that they just aren’t seeing it via my personal posts yet.

So, really, really happy with that last-minute twist 🥳

Conclusion

As I said last time, my account is still minuscule and there is a long way to go. Consistency is key and, although real life must come first, it pays to stick to a schedule and intensity where you can.

The fact that a piece of content I created months ago (with little fanfare) also turned out to be the most successful single piece of content ever, shows that it’s not the quality of my content that’s the issue but the reach and distribution - this isn’t a shock but it’s good to have things confirmed.

April is going to be a busy month!

• I’m expecting a lot of financial turmoil so I will need to spend more time than usual managing and hedging some of my financial investments.

• I want to get the Invoice Powerup MVP off the ground with some marketing, promotional activity and SEO work (all of which I’m a novice at)

• I’ve just built a new HIIT and weight-training plan that will require a bit more personal time. Hopefully, the weather plays ball.

• I’m looking into Newsletter-based businesses on Beehiiv - it won’t be for this newsletter as this is largely a personal account - it’s just an idea.

• I have a neat idea for a (hopefully viral) dev tool/visualisation that I might sneak in if there’s time.

The Fractional-CTO work has been delayed for a few weeks as Loopli is making some priority changes - no bad thing as I was worried about the distraction this month anyway.

In terms of audience building, I’m going to aim for 60% again in April. It seems like a sensible number and I don’t want to give up on it just yet. I’ll also need to ensure I create fresh content this week (ideally my viral idea above will serve that purpose) as relying on replies alone doesn’t seem to be enough now.

Onwards and upwards!

Automation Consultancy ⚙️🚀

Codefield helps businesses build and implement automated solutions for an extremely wide range of challenges.

Regardless of sector or size, we find that there are many, many places where leveraging Automation and AI can help businesses reduce (or even eliminate) the time spent on repetitive and/or tedious tasks, allowing them to focus on the more valuable strategic work that takes the business forward.

If you think there are areas of your business that could be automated or streamlined, then feel free to reach out for a friendly (and in no way pushy) chat at automation@codefield.io

Pivoting from the Scheduled Posting Approach

As I mentioned last time, starting from a tiny follower-base in month one, I managed to grow the number of authentic followers by 60% - a number I was both happy with and thought I may be able to repeat in month two.

Yes, 60% was ambitious but it felt possible … at least at the time.

Week One Woes

At the start of February, I carried on as normal - churning out content, posting daily, and, basically hoping for the best. But the engagement? Small.

After one week, I had gained 7 followers and lost 5 - so +2 followers.

Given my 60% growth target, I was 25% of the way through February but only 3% to my target! Not good, I needed to try something else.

If At First You Don’t Succeed … Try Throwing Money At It

After having written a (semi) popular thread about GitHub contributions in Europe vs. the US and UK (data-driven, visually snappy, topical), I felt like I had a solid candidate to try ‘boosting’.

‘Boosting’ is essentially X’s paid-advertising model where you can pay to have your content displayed to more users in exchange for cold-hard-cash. I know at some point I may want to use paid advertising to sell things, and so I was curious to try this out in anger. I’m not keen on throwing cash at this long-term, but I figured a modest $50-$100 could bootstrap my audience and give me a crash course in online advertising along the way.

The first campaign (Feb 7th) was a bit of a trainwreck; the X interface is pretty unclear (at least to a newbie) and I struggled to target the content to the relevant audience. After this failure, I grokked a few marketing resources and reconfigured my ad - the second campaign (Feb 10th) was definitely more successful but was still an extremely inefficient use of capital in terms of number of valuable accounts reached.

It’s now Feb 11th and I’ve acquired 11 followers - this is about 27% of my way to the goal. Not horrific but not on target for success.

You just need to do it more

Mid-month, I stumbled across a comment from another startup/tech account describing how he’d finally gotten some uptake: be more active. To be honest, it sounded almost insultingly obvious - reply more, engage more, just do more. But after reading through the open-sourced ranking algorithm, it did make some sense: I’d been dutifully posting every day but I wasn’t networking. X is a social network, after all, and the algorithm loves a chatterbox.

So, I ditched the old approach and committed to making 25-30 replies a day to relevant accounts in my niches—Tech, AI, Innovation, Europe, etc. To my surprise … it worked!

Within hours of starting, I’d been quoted by @levelsio (a big-deal account with a massive following). That single reply netted me more followers than my promoted post did in two days. Maybe I was on to something?

The results were steady, consistent growth, real engagement, and best of all, actual conversations with smart, funny, influential people. It’s not an experience I’ve had anywhere else online tbh.

I kept it up for a week, turning it into a game: each day, I’d aim for five more “reps” (replies) than the day before. Between Feb 11th and Feb 18th, I had acquired an additional 35 followers!

Not only was the 60% target looking achievable, I was likely to overshoot! 🚀💪

The Engineer Strikes Back

Scrolling aimlessly through my feed to find reply-worthy posts was getting inefficient. By the time I chimed in, conversations were often stale, and I needed a system if this was to work for more than a few weeks. Here’s roughly what I came up with:

Reply-o-Tron 3000™

• Identify role models: I picked accounts I admire and want to emulate.

• Dig into their networks: I scoured their followers and following lists for gems.

• Build three lists: I sorted 75 accounts into Low (1-10k followers), Med (10-50k), and High (50k+). That’s the sweet spot I can handle while keeping my replies sharp and authentic.

• Leverage X features: I used the Lists tool to organize them and the X Pro dashboard to monitor all three at once—like multi-tabling in my old online poker days.

• Timebox it: I set aside two hours daily (one at EU midday, one at US midday) to dive in, reply, like, repost, and follow.

This kicked things up a notch. Regular engagement started paying off, and - surprise - I was now noticing some of my standalone posts (not replies) were gaining a small amount of traction. For the first time, they weren’t just ghosting into the void.

I needed to capitalise on this if possible.

The Stripe Thread: Timing Is Everything

I decided to attempt one more high-effort thread. I wasn’t sure exactly what the topic would be, but then, as if by magic, Stripe (a multi-billion-dollar fintech company) dropped their Annual Letter!

It’s a fantastic letter, and for anyone starting an online business, I’d highly recommend staying on top of anything Stripe publishes - their size and scale means they have economic insights that no-one else can match.

What was most interesting was that, despite being US-centric, around 30% of their letter was dedicated to the problem of EU growth, pinpointing opportunities and bottlenecks in the region’s tech and finance ecosystem. This was extremely relevant to me: Tech, Finance, Europe, Innovation—all the things I’ve been talking about.

So, what to do? Speed-read it, distill it into a punchy, digestible thread with a few punchy visuals, and tweet it out before the masses jumped on the bandwagon.

I pulled it off. Within minutes, Patrick Collison (Stripe’s co-founder with 500k+ followers) had liked it. That like seemed to be a bat-signal to the algorithm - “This is good stuff” - and the thread started popping up in high-value feeds.

The thread caught the eye of an Irish News company, who reposted it on their X account as part of a segment on Ireland’s role in Europe’s fintech boom. A European think tank focused on digital policy also shared it, framing it as a “must read”.

Then Andreas Klinger, a linchpin in EU tech innovation with 60k+ followers, retweeted it to his network of policymakers and VCs. He even slid into my DMs with some kind, practical tips on sharpening up my content for the EU growth crowd - absolute gold.

Suddenly, my little thread was bouncing around newsrooms and policy circles. The result? A (small) flurry of new followers and - better yet - a budding network of high-quality contacts tied to the “grow Europe” mission.

Even including four days away from X for personal reasons, the final week-and-a-half was enough to push me comfortably across the finish line with +42 followers and smashing my original target - I hit 186%!

Conclusion

My account is still minuscule and there is a long way to go. The process of building an online audience is more art than science, but it’s becoming clear to me now that certain aspects of it can be wrapped in processes that make things more reproducible.

I’m going to aim for 60% again in March. This may seem cowardly given this month’s results but, now that the baseline is getting bigger, I may struggle to keep hitting 100% gains. On the flip side, a bigger audience and network effect may actually accelerate things … I guess we’ll find out in about four weeks.

I’ve also landed a Fractional-CTO gig with a small health-tech startup - one day a week, nothing crazy, but it’ll nibble at my time. Balancing that with X and other things should keep me on my toes.

For now, I’m going to take this as a small victory whilst remembering to keep in mind that I’m still a pico-account and have a lot to learn.

I’m pleasantly surprised that it’s been possible to apply aspects of my research-driven, engineering mindset to generate solid results in a domain I’d previously thought was completely driven by emotion (or at least impossibly complicated algorithms that would be impossible to crack).

Thanks for reading.

Automation Consultancy ⚙️🚀

Codefield helps businesses build and implement automated solutions for an extremely wide range of challenges.

Regardless of sector or size, we find that there are many, many places where leveraging Automation and AI can help businesses reduce (or even eliminate) the time spent on repetitive and/or tedious tasks, allowing them to focus on the more valuable strategic work that takes the business forward.

If you think there are areas of your business that could be automated or streamlined, then feel free to reach out for a friendly (and in no way pushy) chat at automation@codefield.io

Why now? Well, after several conversations with peers, I’ve concluded that now is a good time to transition this newsletter away from being (exclusively) a personal account of what I’m working on and toward writing about the topics that really fascinate me. I find that writing about things helps to clarify my own (often wooly) understanding of them too!

I’ll continue to write about my personal progress so that folks who only care about that can simply skip this series.

Europoor

Sometime around the start of 2024, I became aware of the term “Europoor” - it’s a derogatory term used by some Americans and (bizarrely) Brits to convey the vast gap in wealth, innovation and living standards that has opened up between Europe and the US over the last few decades (particularly post GFC).

I wasn’t really bothered about the online name calling - that’s a phenomenon as old as the internet itself and, Lord knows, everyone gives as good as they get. What bothers me is that, putting aside the playground insults, when it comes to innovation, this is a totally valid claim. Europe and the UK are falling behind and, if anything, the problem is getting worse not better.

This isn’t a commentary on whether the US is overall a better place to live and work - that’s a complex topic and there are many, many aspects of life in which Europe/UK rank more highly (I’d highly recommend this piece if you want a balanced view on that) but on the topic of innovation, the US is pulling ahead.1

• Since the 2008–09 crisis, US labor productivity has surged by 30% - more than three times the pace recorded in the Eurozone and the UK.

• American companies, buoyed by an aggressive culture of risk-taking and heavy R&D spending, have driven GDP expansion by 11.4% since 2019

• The US repeatedly produces trillion‐dollar tech giants, while Europe hasn’t birthed a single new firm valued above €100 billion in the last half-century. 2

• US investors capture 83% of venture capital in G7 economies.

• To bridge this innovation and productivity chasm, the EU is estimated to need an extra €800 billion a year—about 4.7% of its GDP—just to start catching up, with the UK facing similar uphill battles after decades of underinvestment.

And this is to say absolutely nothing of other international competitors like China and India (but let’s save that story for another day).

Yikes! So, is there any, um, good news?

Well, honestly, I love Europe and the UK and I don’t think we should just be giving up and accepting a managed decline of the economy.

There are (despite all the doomers) reasons to be optimistic. It will require major changes in policy, tonnes of new infrastructure, piles of cash and rather a lot of time BUT lots of people are already beginning to push for a more innovative and ambitious Europe! 🔥

Edit: France’s Mistral released this just a few days before I sat down to write this.

With all that said, the rest of this series will be filled out with observations, ideas, solutions and problems as I bump into them.

I’ll try to keep things balanced - some pieces will challenge the Europoor narrative, some might support it and others will be more ambiguous. It won’t always be big-hitting stuff either - some pieces (like this one) will be small gems that aim to make you stop and re-check your assumptions.

So… let’s git to it! (sorry, that was the last Git pun, probably)

GitHub Innovation Dataset

As many of you may know, each year, GitHub releases a report that highlights statistics and trends derived from all the public activity happening on GitHub. The data is explained on the GitHub Innovations site and it’s a really interesting read - it slices the data in lots of ways and gives you a feel for what people are working on, what languages are being used and so on.

After reading through the data, it occurred to me that much of it is presented at the aggregate level (i.e. you can see activity at the country or economy level) and then separately they provide data on the number of users/developers etc.

I downloaded the data, did some of my own processing, charting and found some really interesting patterns. I posted these on Twitter and it was well received by the technical community. I’m going to present the findings here without too much in way of “explanation” as I’d like people to draw their own conclusions.

Like any metric, it’s important to understand the limitations - this data includes public activity only and does not account for:

• Private activity/contributions (only ~7% of GitHub repos are public3)

• Quality of contribution

• Incorrectly labelled location and user data (e.g. users forgetting to update their location when they migrate)

All that is to say that you should be cautious about extrapolating these results to make claims about private activity (which is the majority). Anyway, let’s plow on…

1. Total Git Pushes for US, UK and EU

Typically, activity metrics are published at a per-economy level with the bigger blocs leading in terms of total public activity.

And, that’s exactly what we see when we chart it: 🇪🇺 > 🇺🇸 > 🇬🇧

But, whilst interesting, this doesn’t tell us the whole story…

2. Git Pushes Per Capita for US, UK and EU

I gathered the historical population data (via Eurostat, US Census) for the last four years and joined it to the GitHub dataset so that I could see what things look like at a per-capita or per-person level. Note that this is total population and not working age population.

Now you see a different story. The order completely flips so that the Brits are now leading the charge and Europeans rank last: 🇬🇧 > 🇺🇸 > 🇪🇺

So, should we all start drinking English Tea and warm beers? Maybe. Although, again, this isn’t telling us the whole story…

3. Git Devs Per Capita for US, UK and EU

Naturally, the more digital and service-led economies will generally have a larger number of developers per capita. i.e. more tech workers and fewer manual workers. This would give them a positive bias on the previous chart.

We can check this assumption by charting the number of GitHub users (per 1000 people in the population) for each of these economies.

It’s exactly as you’d expect; the most digital economies are on top: 🇺🇸 > 🇬🇧 > 🇪🇺

4. Git Pushes Per Developer for US, UK and EU

So, why don’t we look at the activity per GitHub developer? Well, when I joined those two datasets, the ordering switches again and we see that the Europeans are actually contributing more than their US and UK counterparts: 🇪🇺 > 🇬🇧 > 🇺🇸

Holy Crêpes Batman! Despite the stereotype of the mega-industrious US engineering teams, when it comes to public activity, it appears that the Europeans (and to a lesser degree the Brits) are contributing more per engineer than the Americans.

On top of that, regardless of which economy you’re looking at, all three of these economies are pushing less per developer than ever before - that’s definitely interesting. Perhaps people are using multiple accounts or more non-developers are using GitHub for other reasons etc.

Anyway, at the risk of becoming repetitive, I want to say again that there are huge limitations and assumptions baked into this data - I’ve deliberately resisted adding too much opinion to this. Nonetheless, it’s at least worth being aware of this result when reading future reports on this type of activity.

5. Europeans take holiday seriously

Lastly, and just for fun, I couldn’t help but notice the absolutely massive drop in activity that occurs every Q3 in Europe (Jul, Aug, Sep)

In fairness, who wants to write code when it’s 40°C and you need to get to the beach before all the deck-chairs are taken!

P.S. I’ve been asked by a few people how I generated these graphs with the hand-drawn look and feel (à la Excalidraw). It’s actually a custom tool I’m working on which allows D3.js data visualisations to be given the rough/hand-drawn look - when it’s finished I might make the tool available to others but, right now, it’s not user friendly!

Until next time.

Automation Consultancy ⚙️🚀

Codefield helps businesses build and implement automated solutions for an extremely wide range of challenges.

Regardless of sector or size, we find that there are many, many places where leveraging Automation and AI can help businesses reduce (or even eliminate) the time spent on repetitive and/or tedious tasks, allowing them to focus on the more valuable strategic work that takes the business forward.

If you think there are areas of your business that could be automated or streamlined, then feel free to reach out for a friendly (and in no way pushy) chat at automation@codefield.io

A Month of Trials, Tribulations, and Tweeting

As you might remember from last time, I’ve recently committed to dive head first into the choppy swamps waters of Twitter. I expected to hate it but, in truth, it has actually been quite a good laugh.

I was (and still am) very critical of many aspects of social media and I’m very unsure about whether history will look back on this “innovation” as a net-positive or not. All that being said, I don’t think social media is going away and so, in my view, it’s something we should spend time trying to improve, contain and iterate on.

Zooming in on Twitter specifically, I think that despite its myriad flaws, it seems to be the best way of ingesting realtime information on almost every topic; there’s a good chance that the majority of experts and public figures in your field of interest are on Twitter.

Obviously, there are HUGE issues with misinformation, echo-chambering and the whole “freedom of speech vs freedom of reach” dynamic but, for me, I believe the quality of information and discussion is still the best that is currently available.

Previously, my only real foray into social media was Facebook - this is almost the exact opposite experience to that of Twitter - it’s largely an alternative way to communicate (less effectively) with people that you already know. I very rarely learnt anything more useful than that my second-cousin now lives in Twickenham or that I’ve forgotten my auntie’s birthday (again). Hmmm, I guess it’s also good for sharing pictures.

Subscribe now

So, in summary, although it’s a very flawed place, I’ve enjoyed the first month of this Twitter/X experiment and will continue to engage with the community there.

Progress Review

First off, I think having the metrics-based approach to finding out what works and what doesn’t has been really interesting and, without that, I’d be quite disillusioned already.

Whilst the numbers are useful, it’s obviously early days and drawing too many conclusions at this point is unwise. Let’s take a look though …

That don’t impress me much (oh, oh, ooh)

I possibly have the brains but I don’t yet have the touch - here is a snapshot from the first month (January).

I don’t want this piece to be a forensic investigation into every datapoint but this graph tells some believable stories:

• Video content is watched, liked and shared a LOT

  • The peaks around Jan 5th and Jan 16th were both following the release of a small tech-focused animation (one about programming, one about AI)

  • X has claimed they want to become a video-first platform and I believe they’re encouraging this by boosting the reach of video content. This is good for small accounts who don’t have many followers (like me).

• Until you have a large following, text-posting (not replying) is barely worth it

  • The impressions I get from posting under my own account are tiny and have barely received any engagement at all. Basically, until you become identified as a high quality voice - you just get drowned out by everyone else.

  • The plateaus around Jan 7-13 and Jan 22-28 were periods where I wasn’t able to create visual content and just made sure I stuck to my daily schedule by putting something out in text.

  • Posting images is slightly better but still seems to garner little attention for small accounts like mine. Memes generate poor results too.

• Making relevant/funny/intelligent replies to large accounts can bring massive results. Almost every other peak on that graph is due to a large-account reply.

  • If you can reply quickly, reply quickly - I now set up notifications for large accounts that are relevant to my audience.

  • If there is something major happening in your space (e.g. DeepSeek v3 fallout) then consider preparing some words/ideas in advance that may be relevant. You can then roll these out if/when the right person also posts on the issue.

  • If you can be witty, be witty.

  • Unless you want to be hounded by bots, trolls and scammers - stay clear of extremely controversial topics or mega accounts (Musk, AOC etc)

Everything is relative

Whilst my account is still absolutely minuscule, I think I’m getting somewhere - I am starting to work out what works, what doesn’t and what else I should be trying. Here are some month-over-month metrics:

Now obviously, most of these 1000%+ gains are base-effects explained by the fact that I’m starting from such a low level. Regardless, I’m taking it as a positive for now, and over more months, I should be able to see a more informative pattern emerge.

You’ll notice the engagement rate (engagements divided by impressions) has actually dropped. Simply speaking, this means people are liking less of what they see than before (not good) but, after looking into this more, I believe it is due to two things:

1. Base Effects - I only made a handful of posts in December but received a few likes on them. I don’t think more than 10% of those engagements were real (see 2)

2. Bots - I had a massive percentage of bot-followers in December. They add ‘likes’ to everything and pollute the signal. I don’t care about vanity metrics or likes from bots. That’s not why I’m here.

Ah, yes! Bots! Let’s talk about them again

As I mentioned last time, at the start of this quest, I spent a while blocking any follower that I expected to be a bot - sadly, it turned out to be the majority of my following!

Although I wasn’t foolish enough to think that would be the end of it, I wasn’t sure how quickly these bots would replenish themselves - turns out, pretty damn fast.

• I acquire new bot followers every single day.

• I receive likes from bots every single day.

• I occasionally receive full-blown (ai generated) replies from bots.

Right now, I’m able to stay on top of this - I block and report any account that I’m confident is a bot. It’s tedious but manageable and there are some tools that make it simpler. Hopefully this will become a higher priority for the X engineering team at some point - until then, it’s just one of those things.

Experimentation

One thing I am having some fun with is the experimentation - what works, what doesn’t etc. For example, I’m trying videos with background music and videos without music, posts with different capitalisations, posts with line-breaks, posts with emojis, posts at different times of day etc … I don’t have enough data to draw conclusions yet but I think that will come.

There’s a limit to how far I’m willing to go with this; I don’t want to completely sacrifice the content at the expense of engagement - that’s a road to nowhere and, as someone who detests clickbait, I really don’t want to become part of the problem.

There’s a lot of research on diminishing attention spans and why it seems to be a growing problem - much of that is attributed to the sheer volume of information that people are exposed to nowadays BUT much of it is also attributed to the format and delivery of that content too. It’s hard to say whether this is inherently bad or not - perhaps it’s a requirement of our inevitable transition to Homo Technologicus - but I want to exercise some level of responsibility and caution on this point.

I remember, many years ago, I started digging around on the topic of A/B testing for some e-commerce related stuff; I stumbled across this comment on a podcast that, until now, hadn’t quite landed properly:

What we know is that if you A/B test a website enough times, it will turn into a porn site. Because when you A/B test, you will end up with clickbait. You will end up with prurient images that people sort of click on in the short run because… ‘ehh.’ That is no way to do the work that you care about - Seth Godin

After seeing the other side of these metrics, I can see how it could become very easy to optimise for engagement, degrade yourself, your content and your online-persona in the name of clicks; I want to get that on the record here - almost like a reminder to myself to be alert to the risk and to hold myself a little more accountable. If this sounds like virtue-signalling, then sorry, I guess it kinda is.

Next month

I have genuinely enjoyed putting together the animations and want to double down on a sort of animated-infographic approach. Finding the data, charting the data, and animating the charts has been really fun and has given me an excuse to play around with some technologies I haven’t done much with before (manim and D3). Having content based on data also serves as a jumping off point for my own commentary and interpretation too - eventually this could lead to some interesting discussions with others.

I’ll continue to engage with large relevant accounts where it makes sense and, although it seems largely a waste of time, I’ll continue to regularly post on my own account for now.

This month I increased my genuine user following by 60% and (with a fair-wind) I think I can repeat this growth in February too.

Thanks for reading.

Automation Consultancy ⚙️🚀

Codefield helps businesses build and implement automated solutions for an extremely wide range of challenges.

Regardless of sector or size, we find that there are many, many places where leveraging Automation and AI can help businesses reduce (or even eliminate) the time spent on repetitive and/or tedious tasks, allowing them to focus on the more valuable strategic work that takes the business forward.

If you think there are areas of your business that could be automated or streamlined, then feel free to reach out for a friendly (and in no way pushy) chat at automation@codefield.io

Honestly, as someone who is extremely cynical of online phenomena such as “Influencers” and “Professional Youtubers”, the prospect of engaging more with this universe is not massively attractive.

Despite my reservations, I do believe that the Creator Economy is a very legitimate concept and that, as AI begins to bathe us in synthetic content, there will be significant opportunities for people who are able to establish some sort of online social capital and build an audience that, not only trusts them, but assigns a sort of human-premium to their more authentic content.

I may not love the idea, but, if this is what it takes, I’m always happy to give something a crack.1 I absolutely draw the line at the Youtube Face though. Grim.

The Plan: Systematic Approach to the Chaos

After a rather sensible New Year’s Eve, I awoke on January 1st with a clear head and an unallocated bank holiday. So, although I don’t normally defer life changing behavioural tweaks to the start of the new year, in this case, I actually indulged a little. I bought myself a nice notebook and laid out some “Creator Economy Consistency” habits/goals for 2025.

1. Post to Twitter/X (three times per day)

2. Engage with relevant communities (three times per day)

3. Engage with influential accounts (three times per day)

4. Reply to followers or community members (daily)

I actually have a more detailed plan here but it’s overkill for this article. You’ll notice I haven’t set any targets for things like “Total Followers” and so on — there’s no good reason for that other than that any numbers I pick at this point would be complete guesswork; after a month or so of growth, I’ll re-assess and try to pin down a number that is ambitious but achievable.

Scheduling and Efficiency

I really don’t want to have to log in to Twitter three times per day, generate some content and then post it out. Organising my life around 3 strict slots would be extremely inconvenient and, if anything, I’m actually trying to have phone-free periods in each day (that’s another tale).

Fortunately, this is an easily solvable problem; just use a scheduling tool such as Typefully and consolidate all the ideation, creation and posting effort into a single block at the start of the day. I genuinely believe moving to this model will make a huge difference to my consistency and ability to hit the NY goals (above) - in fact, I think it already has.

Tracking, measuring and experimenting

A secondary benefit of using one of these scheduling tools is that they often come with a reasonably high-res view of your account engagement metrics (e.g. Follower Growth, Likes, Replies, Bookmarks etc).

Though I think it’s totally possible to ignore all the metrics and just use a more brute-force approach, it seems silly to throw away a potential goldmine of realtime data - especially if you have a tendency to be analytical by nature anyway.

For anyone who was forced to went through a basic business or economics module at university, you probably already know that this is actually a famous misquoting of - “It is wrong to suppose that if you can’t measure it, you can’t manage it – a costly myth.” - but, putting that pedantic side-note to one side, it’s still very much the case that people and organisations tend to do better when they adapt and refine their strategies based on objective measurements.

With that in mind, I’ll be doing regular (probably weekly or fortnightly) reviews of the data to see what seems to work better - for example:

• Is a reply to a big account better than tagging the account in my own tweets?

• What time of day does my target audience tend to engage the most?

• Does a short, punchy tweet work better than a diligent, detailed thread?

• Are my second-rate memes really as funny as I think they are?

• Do I need to post more video content?

Obviously, there are lots of ways to slice the data and some times I’ll have to conduct a more formal A/B test type process to really know for sure - but you get the point, it’s an iterative thing that I’ll be using to direct my overall strategy.

When Robots Attack

I couldn’t really finish this piece without commenting on one of the more frustrating aspects of X/Twitter - the ridiculous volume of bot/fake traffic! I’ll probably write something about this specifically in due course but it really is out of control … I won’t go so far as to say it’s killing the platform, but it just makes things a bit icky.

For context, I have had my Twitter account for 7 years and it has only really become regularly active in the last few days; only two days into this active-period, my account had already amassed approximately 160 followers - that seemed very suspicious to me and so I dug a little deeper.

After screening my follower list, I removed more than 50% for being either bots, impersonators, or inactive accounts! I have to say, though it was satisfying to kill off the bots, it was also kinda depressing; I had been hoping only a small percentage of my followers weren’t real. Alas, I now have a clean slate and I’ll prune as I go.

You might ask, why I bothered to remove them in the first place? Perhaps, having a bot follower is better than not having a follower at all? Right? Well, no.

As it happens, many metrics use the Follower Count as the denominator and so, if you increase this number, but don’t increase engagement on your content (bot’s generally don’t interact much) then the algorithms will assume your content is poor-quality or uninteresting to people and bury it with all the other cruft. Perhaps counter-intuitively, it’s better to have a smaller follower base with smaller bot percentage than a higher follower base with a higher bot percentage.

Fortunately, these bots are extremely easy to detect - I’ll write about this later in more detail but basically… if an account has 2 or more of the following, it’s very likely to be a bot:

• Generic (but attractive) female Profile Picture

• Massive Following count but small Follower count

• Zero/minimal activity on their Feed

• Large number of numeric (or seemingly random) characters in their Username

Honestly, I don’t understand why the engineers at X don’t implement a simple Bot Score or similar. It seems like a few simple heuristics and a cheap image classification algorithm could eradicate the majority of simple bots - maybe there is some commercial advantage to having higher volumes of activity on the platform? Or maybe they want to push more people towards the paid/verified service? I don’t really know - but it is very annoying right now.

Wrap up

Anyway, that’s all for now. I’m cautiously optimistic that this strategy will pay off - if you’re in a similar boat, have a look at the detailed plan I linked earlier - it’s based on material collected from other (actually successful) accounts and tweaked for a Tech-based audience.

Happy New Year.

Automation Consultancy ⚙️🚀

Codefield helps businesses build and implement automated solutions for an extremely wide range of challenges.

Regardless of sector or size, we find that there are many, many places where leveraging Automation and AI can help businesses reduce (or even eliminate) the time spent on repetitive and/or tedious tasks, allowing them to focus on the more valuable strategic work that takes the business forward.

If you think there are areas of your business that could be automated or streamlined, then feel free to reach out for a friendly (and in no way pushy) chat at automation@codefield.io

Looking back, I think I’ve been having fun with this sort of thing since I was young enough to believe that making money was simply a matter of having a clever idea. I won’t bore you with the whole list but, just to give you a glimpse …

The Early Years: Santa Merch and Dodgy Downloads

1. Selling Santa’s golden buttons on Ebay (along with the half-eaten mince pies, my parents had scattered some of Santa’s buttons around the fireplace - I thought they would be worth something!). I never listed them.

2. Saving up for a CD-Writer (probably museum exhibits at this point) and “burning” free versions of CAD software for my fellow GCSE Technology classmates. My price? A mere £5. Two whole weeks of pocket money!

The Pre-Unicorn Portfolio: Almost Inventing Things That Mattered

Joking aside, there were some ideas that, over time, have proved to be quite prescient…

There was my proto-Uber moment. My best friend (working as a taxi-office call-operator) and I spent one evening detailing how brilliant it would be if people could just press a button on their phones to summon a cab and then the cab drivers would just select the jobs themselves. “Genius!” we agreed, before promptly doing absolutely nothing about it.

Similarly, while working my first “proper” job, I became the unofficial sandwich coordinator for our floor, manually collecting orders and funds before making the pilgrimage to the local sandwich shop. “Wouldn't it be brilliant,” I thought, “if we could just do all this online?” JustEat (not me) would later prove that yes, indeed it would be.

The pattern here is probably obvious: having ideas, as it turns out, isn’t the problem. The problem is what comes next – or rather, what usually doesn’t come next. Each idea met the same fate: a brief moment in the spotlight of my enthusiasm, followed by a hasty retreat into the shadows of “yes, but” thinking. “Yes, but someone else must have thought of this” - “Yes, but the technical challenges are enormous” - “Yes, but my coursework is overdue and I don’t even know what drop-shipping is anyway!”

The Procrastinator Strikes Back

After years of perfecting the art of having good ideas and doing precisely nothing with them, I finally decided it was time for action. I assembled my tools with the kind of meticulous attention to detail that only a caffeinated engineer knows how. Keyboard hotkeys optimised? Check. Notion workspace configured to the level that would make a library scientist weep? Absolutely. Three different work-tracking methodologies researched and compared? But of course.

What followed was a masterclass in what I'd describe as “technical procrastination” – a peculiar state where one is simultaneously doing everything and nothing. As an engineer, I’d discovered that building anything was a splendid way to avoid doing the right thing. Need to work on customer acquisition? Perfect time to refactor that authentication system. Marketing plan looking a bit thin? Clearly, this calls for a new GraphQL API.

This implementation trap is particularly devious because it feels productive. You’re writing code, solving problems, implementing features. You really are busy. But busy doing what, exactly? In my case, I was often busy avoiding the harder, scarier parts of actually launching a business. The parts that couldn’t be solved with a clever bit of middleware or an elegant database schema.

It’s basically like re-organising your spice rack while the kitchen’s on fire. Yes, the cinnamon is now adjacent to the cumin, but perhaps the flames blazing out of the air-fryer should be addressed too?

The Triangle of Tech Founder Suffering

If we distill the myriad components of a successful business down to its essential elements – ideation, building, and selling (including marketing and distribution) – a rather horrid triangle emerges. For myself, and I suspect many others, generating ideas proves to be the most comfortable challenge, followed by the actual building of the thing. Then, trailing a distance best measured in light-years, comes the task of selling, marketing, and getting the bloody thing into customers’ hands.

Unfortunately, what is most comfortable is not what is most necessary. Not at all. The elements grow progressively more challenging as you descend through the Triangle of Tech Founder Suffering.

1. Ideation - This sits at the top of the Triangle of Tech Founder Suffering – the delightful process of thinking up clever solutions to problems that may or may not exist.

2. Building - Moving down – this is our comfort zone as technical founders. Yes, it’s challenging, but it’s a known quantity. We can measure it, test it, and most importantly, hide behind it when the scarier parts of building a business come knocking.

3. Selling - And then... oh dear. The bottom layer. That vast expanse of activities that can't be solved with a clever algorithm or a particularly elegant use of dependency injection. Marketing. Sales. Distribution. The parts that require actual human interaction and, heaven forbid, the possibility of rejection.

Most technical founders (myself included) tend to invert this triangle in terms of time spent. We'll happily spend months perfecting our pixel-perfect responsive design whilst our marketing strategy consists of a vague notion to “post it on Reddit when it’s ready.”

Idea Selection

Despite establishing that generating ideas is possibly the least challenging part of the entrepreneurial equation, what is still very important is to be able to validate these ideas before you sink weeks, months or even years into them.

There will be plenty of time to write about engineering and sales over the course of this newsletter but, to synchronise you readers with where I’m personally at in my journey, it makes sense to talk about the “Idea Validation” process I’m working through right now.

The Idea Repository: Bringing Order to the Chaos

To organise and capitalise on the ideas whizzing around my mind and chrome tabs, I maintain two distinct repositories of ideas. The first is what I’ll call the “Lightbulb List” – capturing everything from shower thoughts to pub revelations that occur almost unprompted throughout the day. The second is more deliberate: a researched compilation derived from market analysis and studying successful founders - I’m just gonna call this the “Research List” for now.

The key difference? One list is essentially the result of mixing all of my own thoughts about life, business and technology together and generating ideas that are likely to be unique or novel (even if not very good). The other, more observational Research List keeps me grounded in market reality and mitigates all the personal bias contained in my Lightbulb List. Occasionally, when I cross-reference these lists, a genuinely strong “Candidate” emerges that warrants further investigation and validation.

This systematic approach might seem over-the-top but there’s method to the madness: by bringing structure to idea selection, I’m reducing the likelihood that I fall into another “Yes, but” cycle and, ultimately, do nothing with any of them.

The Matrix: A First Attempt at Quantification

My first serious attempt at idea evaluation involved creating an unnecessarily complex scoring system in Notion. The criteria, in all their questionable glory, were as follows:

1. Skill Alignment - Can I actually build this, or am I flattering myself?

2. Passion Alignment - Am I (and will I remain) interested in this?

3. Customer Acquisition - How difficult will acquiring users be?

4. Revenue Potential - Are we talking beer money or real money?

5. Trend Alignment - Am I fashionably late (or even too early) to this market?

6. Development Simplicity - Is this feasible for a one-person tech team?

7. Supportability - How much ongoing support will this require if it does work?

8. Scalability - Can the user base scale with effort and/or time?

9. Market Competition - Is this space already overcrowded?

Now, this is by no means perfect, but it did allow me to stack-rank my ideas and keep some momentum. If I went back in time, would I come to the same conclusions? Absolutely not. But this framework wasn’t about finding the perfect idea, it was about breaking free from the paralysis of endless possibilities.

The real value wasn’t in the scoring system itself, but in forcing me to confront the uncomfortable questions about each idea. Questions like “Do I actually understand who would pay for this?” and “Am I prepared to spend my weekends explaining to users why I can’t send an email to their WhatsApp?”

The Refined Search: A More Systematic Approach

Now that my first venture has been released, I’m experimenting and refining a more structured approach to idea validation for the next one - it’s split into two distinct phases:

Phase 1: Idea Candidate Selection

The process begins with cross-referencing my Lightbulb List against emerging patterns noted down in my Research List. This phase often generates a shortlist of possibilities - either through direct overlap or through ideas that simply cannot be ignored from either source because they’re so strong.

Phase 2: Candidate Validation

Each Candidate then undergoes a more rigorous examination. Right now I’m using two tools for this (though I’m sure I’ll introduce others as time goes on)

1. I use DataforSEO to understand search engine usage for terms related to the candidate. I look at how those keywords and terms are trending over time and who else is occupying that market space.

2. I use GummySearch to surface actual user discussions and pain points from related Reddit communities. This tool is surprisingly insightful and removes a lot of the grunt work normally required to find market niches and unfulfilled user needs.

Spoiler: In future issues, I’ll go into more into detail on these (and other) tools.

Process Outline

The process is deliberately short as I prefer fast feedback loops for this kind of thing. Honestly, it’s still more art than science at the moment - but it does the trick.

1. Cross-reference Lightbulb List with Research List

2. Create initial shortlist of Candidates.

3. Validate Candidates using toolchain (currently DataforSEO and GummySearch)

4. If nothing compelling emerges:

   • Explore topics that emerged during (3) and add to Lightbulb/Research List

   • Adjust filtering criteria (essentially lower the bar for Phase 1)

   • Return to Step 1

Since I’m basically writing this in realtime, the above is the process I’m currently engaging in to find and validate my next idea (after releasing the MVP of Handle Hawk) earlier this year. I’m trying to time-box this stage to a few weeks so, with a fair wind, I should have selected and validated my next candidate soon.

What else helps?

Like it or not, even if you have a great product and strong market fit, it can still be very difficult, expensive and/or time consuming to find and attract users. An obvious way of addressing that is to just, generally, have a much bigger online presence such that you can direct people to your products and services directly.

And so, I’m throwing caution to the wind, shredding my lurker-skin and turning up my social media game to 11*. This means that, in parallel to all the work above, I have a long-term commitment to building an audience here, on X and (more recently) on Bluesky. It’s a very experimental process at this stage but the medium term goal is just to remain consistent for the next six months and post something (somewhere) quite regularly. The reality is that unless I strike viral gold, it’s going to be a marathon rather and not a sprint. I’m OK with that - endurance is my thing.

*Well, realistically 6 or 7 - I’m still a Software Engineer after all!

That’s all folks…

In the coming weeks, I’d like to streamline the process and tool-usage so that it’s quicker and more repeatable. I’ll playback how I’m using each tool to find market opportunities and eavesdrop on future customers - with some luck, this will turn into a neat mini-framework that can help people cruise past the “Yes, but” stage.

As per the first issue, don’t read this because you think I’ve discovered some secret formula for success - I really haven’t (yet) - read it because, by understanding my own thought processes, successes and failures, you’ll be able develop an approach that works for you.

Until next time, keep building – but perhaps spend slightly less time optimising your development environment and slightly more time talking to actual humans. I know, I know... but they’re not as scary as they seem.

Automation Consultancy ⚙️🚀

Codefield helps businesses build and implement automated solutions for an extremely wide range of challenges.

Regardless of sector or size, we find that there are many, many places where leveraging Automation and AI can help businesses reduce (or even eliminate) the time spent on repetitive and/or tedious tasks, allowing them to focus on the more valuable strategic work that takes the business forward.

If you think there are areas of your business that could be automated or streamlined, then feel free to reach out for a friendly (and in no way pushy) chat at automation@codefield.io

If you’re looking for another “How I quit my job and made millions with this simple SaaS” story, you can stop reading now. This isn't that. This is a real-time account of what actually happens when you trade the structured world of corporate engineering for the much, much (much) less structured world of solo founding.

No growth hacks, no success-theatre, and definitely no courses teaching you how to sell courses about selling courses. My goal here is to work out a (repeatable) process to:

• Identify problems worth solving.

• Build the solutions.

• Get the solutions into the hands of users.

• Generate cashflow.

It’s going to be hard, but I think it’s also going to be pretty interesting. If all else fails, at least I tried and learnt a lot in the process.

Subscribe now

Given this is the first edition, it’s probably worth providing some context about the current state of affairs and my observations so far …

The Accidental Escape

The plan was reasonably simple: Take a sabbatical. Learn Portuguese. Maybe build some side projects. Take some time to travel around with my new found free time. The reality? Within three months of leaving, I’d somehow acquired a CTO role I wasn’t even looking for, started a company in Portugal (an adventure in bureaucracy that deserves its own newsletter), and discovered that having unlimited freedom can be both energising and paralysing in equal measure.

Confessions of a Reformed Problem-Solver

The thing they don’t tell you about leaving a senior engineering role: Being good at solving problems doesn't help much when nobody’s giving you the actual problems to solve!

In corporate life, I was one of the "firefighter" types - the person you call when everything is on fire. Production outage? I’m going to be useful. Technical decision paralysis? I can help settle the discussion. War between management and engineering breaking out? I’ll dust off my diplomat’s cap. But now? I’m staring at a blank canvas wondering what the hell to paint.

Unsurprisingly, it turns out that finding the right problem to solve requires a very different skill set to solving the problem itself. And, whilst I love the technical activity of taking something from empty-whiteboard to stable-product, figuring out what product people actually want? That’s a whole new game.

The Content Creator Industrial Complex

Speaking of games, let’s talk about the elephant in the room: the endless parade of content creators selling the dream of digital entrepreneurship. You know the type - “Just create valuable content!” they say, from their perfectly lit home office (complete with a bookshelf creaking under the weight of its own pretentiousness)

The reality? Finding genuine market opportunities is hard. Marketing is harder. And don't get me started on distribution. It’s not that I think these skills are impossible to learn - they’re just not as straightforward as deploying a new micro-service - at least not to me.

What I'm Actually Doing About It

So far, I've:

1. Set up a proper company (Portuguese bureaucracy: 1, My Patience: 0).

2. Landed a part-time CTO role that keeps the lights on.

3. Built a small social media tool as a first dip-of-the-toe.

4. Created this newsletter to document the journey.

Going forward, I’m going to develop and (crucially) test frameworks for:

• Finding problems worth solving (beyond the obvious “X but for Y” ideas).

• Validating markets before writing all the code.

• Building distribution from a modest budget and tiny social media presence.

• Generating significant cashflow without going max-long Dogecoin.

What's Next

This newsletter will follow my attempts to crack this puzzle systematically. No success theatre, no pretence of expertise - just honest reflections and some frameworks, tools and processes for figuring it out.

Future dispatches will dive into:

• A retrospective on my first product - what has gone well vs what hasn’t.

• Systematic approaches and tools for finding market gaps.

• Technical decisions that actually matter.

• Marketing for people who'd rather be doing virtually anything else.

Most importantly, I’ll be designing, refining and testing everything I discuss in real-time, for myself. I have genuine “skin in the game” and if what’s written here doesn’t work, I’ll be the first to feel it.

So, if you’re a technical person trying to build something real, or think you’d derive value watching someone work through this systematically, you’re in the right place.

P.S. If this resonated, hit subscribe. No pressure - we both know there are already too many newsletters in your inbox. That being said, I think this one could turn out to be worth it.

Automation Consultancy ⚙️🚀

Codefield helps businesses build and implement automated solutions for an extremely wide range of challenges.

Regardless of sector or size, we find that there are many, many places where leveraging Automation and AI can help businesses reduce (or even eliminate) the time spent on repetitive and/or tedious tasks, allowing them to focus on the more valuable strategic work that takes the business forward.

If you think there are areas of your business that could be automated or streamlined, then feel free to reach out for a friendly (and in no way pushy) chat at automation@codefield.io