10 mistakes I made as a seed stage CEO

A list of mistakes I made at seed stage that lengthened our time to Product-Market Fit.

Sorry I’m sharing this month’s newsletter a little later than usual. I wrote a long, light-hearted piece on Daniel Ek, King Gustavus Adolphus, and Spotify’s strategy of manoeuvre warfare at the end of last month, which you can find on my website, so I de-prioritised this update.

At Lingumi we spent >2 years trying to get to product-market fit (henceforth ‘PMF’), and survived a near-death experience during that period. For reasons with which I’m not familiar, 18 months has become the established ‘ideal time to PMF’ with funding cycles and runway targets built around that. We did it in longer, and I’ve seen it done much faster, too.

In this post I'm going to run you through a list of mistakes I made at seed stage that lengthened that time to PMF. The good news is we made it through, but we could have done it quicker and better. By learning these lessons over time, I think I have become a better CEO. Many of my readers are early-stage CEOs, so I thought this might be interesting for you. For those of you who are later-stage, do send me comments on mistakes you’ve made later on - I’d like to learn what lies ahead. And I hope you find reading this quite therapeutic…you might have made some of these too!

To summarise the mistakes I’ll write about below, they were:

  1. liability to easily-avoidable cognitive biases

  2. not setting a non-fluffy strategy earlier

  3. firing too many cannonballs, and not enough bullets

  4. not hiring a product manager earlier….

  5. …and not hiring more experienced talent earlier…

  6. …and not offering competitive compensation earlier

  7. thinking that 'long runway' is a destination, not an optional route

  8. being in the trenches too much, and not in the map room enough

  9. thinking that “growth” is the goal too early

  10. not relaxing, ever

Let’s dive in.

1. Liability to easily-avoidable cognitive biases

I think I heard tens of talks, lectures and YouTube videos about how founders should focus on a single metric, take time to hire the right people, iterate fast, avoid cognitive biases, etc. They’re all in these great YC Startup School videos which I now often re-watch.

I’d consider myself a reasonably attentive person. But I didn’t follow the advice. This seems to be a common trend: many first time founders have watched those talks, yet ploughed on the same way, and not heeded the advice about those “startup basics”.

Why? I have re-written this next sentence a few times trying to answer that question.

I was too afraid to pick everything we’d done up, bin it, and start again from a different angle. I guess I was very liable to sunk cost fallacy. I have reacted sharply against this tendency - maybe over-corrected (?!) - with my current approach to strategy. Let’s see what lessons I learn about that in the coming years.

I promised not to offer much advice in this newsletter. But here’s one bit. If you are an early-stage founder and are looking at the metrics and the runway, and thinking “uh oh”, pick it up, bin a lot of what you know and have built, and fix the problem, or build your vision of the world, from a different direction. It’s not giving up. It’s the opposite.

2. Not setting a non-fluffy strategy earlier

Even more of a blocker to our ability to iterate towards PMF, and the subject of an earlier piece in this newsletter, was an absence of a non-fluffy strategy.

I have been a semi-competent generalist, 'product builder' type, leading the team primarily through my actions, and my ability to work deep in the details, across everything (including writing some of the early software, thankfully later deleted...). Part of the workload was sharing the unofficial role of Product Manager. I don't think I made a good product manager back then. For one, my liability to cognitive biases (see above for the first!), particularly confirmation bias and "I've got a brilliant new idea" bias (which needs a snapper title), overwhelmed my ability to really deeply listen to customers, and slowed Lingumi down in the early years.

What’s more, because I was so heavily involved in the details and focused so much on not hiring anyone and reducing burn (see below!), I didn’t have the headspace or the time to think - or learn anything - about setting a strategy.

I've stepped away from working on >95% of the details now, and now, getting that strategy right is my job. Realising this error, and focusing on strategy, helped us turn the ship around as we worked towards product-market fit, a better way, in more recent years.

3. Firing too many cannonballs, and not enough bullets

A great strategy is, in part, a guide on what not to do. We didn’t have that, and did everything that popped into our heads. We’d try loads of ideas.

This was exacerbated by raising our seed round. Suddenly having a hiring budget for the first time in our life opened an opportunity: we could hire people to make more ideas happen! The early Lingumi team were (and are) brilliant, passionate designers, developers, and builders, but between us we had very limited experience. We didn't know what we were doing, which is both good (we had what Paul Graham calls 'schlep blindness'), and a bad thing, because we ran in too many directions, lacked clear tactics and processes for product building and user research, and fired cannonballs when we should have been firing bullets.

“But iteration is the key to PMF, right?”. Yes, it is, but focused iteration, using really lean, hacky test-it-in-a-day prototypes and deleting what was invalidated before. Our Head of Product, Rich, is a master of this practice of quick, scrappy testing now, and spends a lot of his time coaching it. In the pre-Rich years, we did the opposite, piling cannonball-sized new ideas on top of old cannonballs into a messy product experience that took us further away from product-market fit, not closer.

4. not hiring a product manager earlier…

5. …and not hiring more experienced talent…

6. …and not offering competitive compensation earlier

What a trio of mistakes! My mindset of keeping the business incredibly lean and eaking out runway had turned me into a very frugal capital allocator. I imagine post-first-funding-round founders split two ways on this, because I know others who bought beer taps and travelled business class.

Though the spirit of frugality remains in the company, and is a great cultural asset in many companies (“Jeff Bezos’s desk is made from a door”, etc), it meant we initially moved slowly on the opportunity to hire great talent.

Lesson learned? In the words of Lingumi’s Head of Strategy, Mads, “Allocate more capital to strategic costs, and less to non-strategic costs.” For more wisdom from Mads, particularly on budgeting, head to our Inside Lingumi’ blog.

After our seed round, we should have hired a product manager, we should have offered better compensation to hire strong talent (as we now do), and we should have reduced runway to accelerate towards PMF. Those would have been strategic costs and decisions.

Looking back on this, I feel I have learned to allocate capital better over the years. And over time, seed rounds have grown. We raised £1.7m over a pre-seed and a seed round. On £1.7m, your hiring budget is not going to be very big. The average seed round has substantially grown since then. But if you'd given me £3m in 2016, I don't think I'd have known how to use that capital better. We might have paid the same people (and ourselves) more, or hired somewhere in the middle, but I think the only way to really learn these skills is to make the mistakes.

7. Thinking that 'long runway' is a destination, not an optional route

Did someone say runway? That reminds me of another mistake.

Hindsight is a wonderful thing, but I hope by writing this, someone one day learns the thing I should have learned: the goal at seed is NOT a long runway. The goal is product-market fit. It sounds obvious, but 'long runway, slowly iterate' is just one combination of the levers we could have pulled. And I'd argue, on reflection, that it's a route that slows things down; the urgency of the three months before we ran (nearly) out of cash were more productive than the preceeding twelve. Other options would have been 'medium runway, shorter-term PMF targets' or 'short runway, hire the best in the world, go for broke'. When the market is big and your problem well-defined, the main thing holding a company back from finding product-market fit is speed to iterate. We now iterate really fast on new ideas and experiments (see this example), but we’ve hired the talent now (see mistakes 4-6) and I got out of their way.

8. Being in the trenches too much, and not in the map room enough

“Good-morning, good-morning!” the General said
When we met him last week on our way to the line.
Now the soldiers he smiled at are most of 'em dead,
And we're cursing his staff for incompetent swine.
“He's a cheery old card,” grunted Harry to Jack
As they slogged up to Arras with rifle and pack.

But he did for them both by his plan of attack.

- The General, Siegfried Sassoon

This mistake is a re-articulation of what I touched on in (2), but I like the metaphor. I grew up in the British schooling system, where the metaphor ‘lions led by donkeys’ is often quoted in history books to describe useless WWI generals. The Sassoon poem above is a wonderful and chilling example.

I didn’t want to be a donkey, so I jumped in with the lions.

My coach, Hanna, who is wonderful (and is reading this - but you really are), identified early on that one of my weaknesses is wanting to be recognised as an “in the trenches, get sh*t done” kind of leader, which makes me naturally weak at delegating, and taking time out (for holidays, management work, strategy thinking, having a life).

But the point is not that generals are, by virtue of their role, donkeys. The point is those generals were donkeys. I wish I had spent more time learning to become a strong general, and plotting a path of attack in the map room, not just diving into the trenches to show I could fight like the rest of them.

We’ve put a LOT of work into this together, and I hope my team (some of whom read this) have seen the change even during this year.

9. Thinking that “growth” is the goal too early

I’m saving the best until (almost) last! When Lingumi was going through the Entrepreneur First 5 cohort in London, we managed to scrape through the first investment committee by getting traction in the form of sales growth. We sold a whopping £64 of product in the week leading up to the IC, making us the highest-grossing EF5 startup (at that particular moment, at least).

From that moment forth, we tried to keep growing by selling. Increasing sales, albeit without a specific set of revenue goals or associated usage key results, became our obsession. But selling a product that doesn’t have PMF is really difficult, and a focus on doing that, rather than learning from lessons 1, 2, 3, and 5, meant we didn’t admit that we needed to iterate, or focus on a different north-star metric, or really define what PMF would look like for us.

We barely grew at all, banged our heads hard against a wall, and eventually shut down that product and built towards our vision in an entirely different way. The mistake? Focusing on sales, and thinking we had to “get to Series A” or “get to £1mn ARR” or whatever else we had told ourselves was the next milestone.

What should we have done pre-PMF? We should have picked a metric that meaningfully indicated product-market fit, and done anything required (see lesson 1) to grow that. Pushing sales as our metric back then was like trying to hammer in a screw. It might work if you hit hard enough (eg burning cash on ads), but if we’d have looked around a bit, we’d have found a much better tool, and made our lives a hell of a lot easier.

I don’t think seed stage investors necessarily hammer (or screw?) this point home enough to first-time founders. If they do, I missed the memo. The goal is not “growth” in the way that post-PMF startups define growth. The goal is PMF.

10. Not relaxing

Boring but important one to close. I didn’t take an actual holiday (where I switched off) until nearly three years into starting Lingumi, and when I came back from that holiday, with a clear mind, we adopted a new strategy, focused the team on it, totally re-built our product around it, and began gaining momentum.

I think it’s feasible to work quite long hours, and still unwind and take proper holidays. Endurance athletes train really hard, but they make the recovery count.

I’ve heard it a million times, but I’ll write it here again because I’m still not good at learning from lesson 10. The brain is a muscle, it grows stronger when it is given time to relax after a tough workout. I’m still working on this one.

— End note —

One of the objectives of this newsletter is to continually learn. I'm conscious that now I'm into a second cycle: learning to become a great growth-stage CEO. My readership includes a large number of early-stage CEOs, but does also include some later-stage CEOs, and some VCs and angel investors. If you are connected to a later-stage CEO who has learned the lessons of that second stage in a startup's lifecycle and is interested in comparing notes, please do hit reply. I'd love to know who they are.

Thanks for the many wonderful replies to last month’s newsletter about our virtual offsite - it seemed to strike a chord, particularly the “I like cheese” game…

Please do share this if you enjoyed it, or hit reply with any thoughts!