Why I'm here, and what I'll be talking about

4am. Sleep-deprived. Pouring rain. Desperately tired. And yet, I still can’t shake the feeling that I just have to get this out. So without further ado, let’s get into it.

What is ‘this’? That’s an valid question which I still haven’t figured out the answer to yet. I feel like by the time I’ve filled out this blank page I’ll have got there, so bear with me. If you do make it to the end, and you’re a human with a job in or around tech - and getting that tech into the hands of other people - then there might be something here you’ll be able to take away too.

To say the working world has changed over the past year would be an understatement. I haven’t come here to hype or scaremonger - I’d much rather stick to the practical and pragmatic - but it’s not an exaggeration to say that as I write, the ground is quite literally moving beneath our feet. In the 15 years or so I’ve been at it, I’ve never seen or experienced anything quite like it. The LinkedIn Lunatics will you the ‘playbook has changed’, with a few short. snappy. sentences (that were definitely not written by a bot) about how to keep up and what you can do about it. Perhaps that’s where I should be too, but I don’t really fancy subverting myself to the whims of the LinkedIn algo, and we all know it feels a bit disingenuous and self-indulgent over there anyway. So the intention is that by being here, in my own space, you’ll be able to digest these thoughts as honestly as the intention behind sharing them: without alternate intention or agenda, and only as a means to writing down what I think.

Why does it matter what I think? That’s another valid question. The short answer is: it doesn’t. It matters what you think. But I have been in the game for a while, and have been fortunate enough to have had somewhat of a front seat for a lot of that. That probably gives me some kind of unique insight that might make it an interesting tale to tell.

I suppose for one, there aren’t many other founders of Too Good To Go, and there can’t be many growth stories like the early-stage Too Good To Go growth story. I built that from the very beginning and have quite the many learnings to share from along the way - from getting started with absolutely nothing, to scaling up to teams of hundreds.

Then it was Futureplay - a completely different journey in a completely different country with a completely different culture. I was the first marketing hire there, and, while the transition was tough (it takes a lot to step away from your own startup - again, the parts that LinkedIn doesn’t tell you about which I’ll perhaps get into later) I like to think I had a small part to play to helping the company to a successful acquisition by Plarium.

After four years, that was the right moment for me to join Metaplay, another startup of just ten people at the time building backend tech for the games industry. And it’s at Metaplay I’ve felt most at-home since the Too Good To Go days - being able to draw upon all of my learnings from growing companies and games in both B2B and B2C. And amidst all of this I suppose that it goes without saying that along the way I’ve built teams, hired people, fired people, managed up, managed down, closed deals, lost deals, launched products, ran campaigns - and, well, just about anything that anyone working in or leading a tech company without any experience of hands-on engineering and product development could expect to do.

That was until this year at least. Up to now, for literally all of my career so far, people like me have always relied on developers to get whatever we needed to do done. A case in point: we spent three years running our website at Metaplay with an web development agency, and while the agency was great and I’ll always have time for the people there, I look back at that window and wonder how the hell we got by working that way for so long. Design briefs. Development briefs. Lead times. Development times. And at the end of it all, a nice invoice from the agency to boot. This is where I get back to the ground shifting: where a non-technical person used to have to rely on technical people for implementation and execution, you can now do those things yourself. And having spent our entire careers living so close to the customer, we’re actually in a unique position to be able to harness the power of the opportunity in front of us in meaningful ways that can have the kind of impact on your work and your business few others will be able to help - if you can manipulate the tools right.

So that brings me back to what’s changing now. And I’ve just realised this is the first time I’ve written these two letters in this whole piece: AI. (Was it the elephant in the room?). The world is awash with it, and if you don’t know where to start, it probably feels impossible to catch up. Heck, even the people working in AI feel like they’re falling behind. But there’s literally so many gains to be made. As a long-time Wordpress tinkerer I threw myself into Lovable when it first launched in November 2024, got helplessly addicted, built an endless amount of new projects, and, slowly but surely in the time from them to now, have since graduated from Lovable through to Cursor and now, Claude Code. This hardcore vibe-coding ramp has meant that while not being a coder, I’ve been able to pick up a lot of the thinking and mental models required for building the systems that allows marketing and commercial to be self-sufficient, without having to rely on engineers to do simple development tasks when they’d much rather be working on bigger and more meaningful problems.

Concretely, what does that look like in practice? Well, since the start of 2026 alone, our tiny commercial team of two has been able to do many things. It feels a bit like we’re cheating, and that’s why I wanted to get some of these learnings out in the open - I know leaders and operators will be facing similar challenges to the ones we’ve been overcoming, and I want to use the journey we’re going on to help others. Here’s what we’ve been up to and what, over the coming weeks and months, I’ll dive into more practically, so if you’re just getting started with AI then you’ll have a useful reference to follow from:

• We moved to an AI-systems first way of operating, leaving people in the loop only for things like judgement, decision-making and customer-facing content edits and sign-off. Everything else - all the boring, manual, tedious background work - has been or is being automated. This means we don’t have to worry about things like updating CRMs, and, in the long-run, keeps a small commercial team focused on the most impactful things likely to move the needle
• To do this, we’re operating now entirely through Claude Code, Github, and Linear. It’s like work, but on acid. There have been plenty of gotchas and things to be mindful of when setting this up (one wrong decision at the start can cost weeks of fixing later on), so it only feels right to share those sooner rather than later to avoid others falling into the same trap. (If, like me, Github is just a word you’re familiar with but have no idea what it means or what it does then fear not - I was in the same position 18 months ago.)
• We moved our website from Hubspot to a completely self-hosted custom Next.js project. We used to pay thousands and wait weeks to change some modules on our homepage, or design a new landing page. Now we can ship them in a minute.
• We migrated our CRM from Hubspot to Twenty, a new open-source CRM. We used to pay thousands for something that we never even enjoyed using - and much of the time we were with Hubspot was spent trying to optimise our contact database so as not to exceed our marketing contacts tier (IYKYK). That sucked, and I’m so happy we’re not there anymore. Twenty costs like 10 bucks per person per month, so is not only an order of magnitude cheaper, but also an order of magnitude better. We’ll get into the whys of that later on, and how you can take the steps to do so yourself.
• With the hygiene, upkeep and data bricks in place, we then built a team of agents to sit on top of them and do the proactive work that would help us scale. In our line of work - getting Metaplay into more game studios - this means finding the right prospects, tailoring the right messaging, and building the right outreach so that a human can send it at the right time. There’s a million-and-one ways to do this - most of them live in some off-the-shelf tool that never quite does what you want it to in the way that you want it to (there’s no one-size-fits-all for a B2B go-to-market and sales motion, especially when you’re selling complex tech). If you build it yourself, you’ll have to learn the hard way about what works and what doesn’t, how to make your agentic team a help rather than a hindrance - and you can only even start thinking about doing that once you’re familiar with building apps, integrations, CI/CD pipelines, runners and serverless actions, triggers, and a whole load of other technical systems. That’s the journey I’ve already been on, so I think I can give you a bit of a headstart.

Of course, as well as all of this AI and agentic stuff, we’re a people-first B2B sales and marketing org. We can’t build a business without selling to people, and selling to people in the games industry is really quite difficult. Over the years we’ve learned that offline channels are the main levers - events, conferences, mixers, meetings - and everything we do around it should be designed to support that, and be as frictionless as possible. That’s what the above is built in aid of - ultimately, we don’t want to have to think about anything other than actually doing sales and marketing. If you’re in a similar boat, what’s to come will be useful for you.

Anyway, that’s enough for this very long-winded intro post. If you’re still here, hopefully you have an idea of the practical things you’ll be taking away with you over the coming months, as well as the softer side of things too.