🏑 Home

Startup Life Lessons

Four years of running a tech startup have taught me a few things. Startup life throws a lot of punches and there's nothing like learning first-hand. Here's what I've learnt over the past four years.

πŸ“– You can't judge a book by its cover

I learnt that a good CV is no guarantee of a good hiring choice. But I also learnt that a bad CV is no guarantee of a poor choice.

I employed two recruits at Humanise.AI who had probably the worst CVs I've ever seen. The college drop-out that tried not once, but twice, to start a uni course but gave up before the first term had finished and was then bumming around doing nothing. Or the (very) amateur programmer who turned out to have a hidden artistic skill they'd never mentioned at interview. All the interviewing rules would have had us reject them both.

But, we gave them a chance. We could see a spark of something. We weren't sure what, but there was something there. "Come and work for us as an intern, show us what you can do", we said. And so they did. To cut a long story short, in both cases they were given a chance when nobody else was willing to take a risk. We took that risk and and it paid off β€” they both worked hard and showed us the spark was more than a spark. We gave them full-time jobs.

Hiring processes often focus on reducing risk. Eliminate anyone without the good grades, eliminate anyone with a gap in their employment history, eliminate anyone who doesn't fit the conventional career path. In short, the rules are designed to find excuses to reject people.

The trouble with those rules is that they don't actually work.

The risk the rules don't insulate you from is that you end up employing someone with an exemplary history, but no spark. No "I want to show you what I can do". No "you took a risk, let me demonstrate why that risk was the right thing to do". No emotional commitment to making your business thrive. The A1 choices might be smart and supposedly risk-free, but they also demand a high salary and come with an inbuilt "if this doesn't work out, I'll jump somewhere else" attitude β€”Β an attitude that results in decidedly "b-team players" who are coasting along.

So I've learnt to favour looking for that special something, to take a risk, to give someone a chance to shine. These are the better choices. They aren't risk-free choices, but nothing ever is. I learnt to take on risk because it's far more rewarding.

🀩 Recruit for attitude, not specific knowledge

When we interview engineers, we learnt to look for attitude and approach to problem solving, not perfect recall of obscure javascript syntax.

Technical skills can always be learnt. If the person you've employed has the right attitude, they will learn almost anything really quickly.

But attitude very rarely changes. Someone with the wrong attitude can be a constant thorn in the side. I've never found a way to re-wire someone's approach to life. That stuff you're stuck with, so I learnt that it's really important to understand what makes people tick, how they communicate, what their attitude to life is. If that's all good, they can learn that javascript syntax if they need to.

However, someone with perfect recall can be a nightmare. Always prioritise attitude over particular skills when recruiting. Because skills can be learnt and attitudes rarely change.

⏱ It's important to make accurate decisions, but sometimes it's more important to make fast decisions

My gut feel on someone or something has proven to nearly always be right. Sometimes I've not taken decisions based on that gut feeling, instead waiting for evidence to back it up. However, every time I've done that, I've regretted it. Poor choices need to be lanced quickly. If they aren't, they fester. They suck energy and enthusiasm. They cause endless debates and disputes.

I learnt very clearly that sometimes we need to make tough choices, based mainly on gut feel. We can't always wait for irrefutable evidence.

It's sometimes difficult to rationalise our gut feelings, but that doesn't mean they aren't right. It's worth getting in tune with those feelings and listening to them. Our brains are strange things and work in mysterious ways. If we only rely on rational analysis, we're locking ourselves out of half of our brain's function and will always take decisions later than we perhaps need to. Sometimes it's better to take a decision early, even if it might be the wrong one. Certainty of direction is sometimes the most important thing.

I've had to make tough decisions at times when running a business. I've never regretted those decisions once. I have at times, however, regretted not taking those decisions earlier.

🧠 People need to feel responsible

I once rejected a set of objectives given to me by my boss. I explained to him that what he'd given me weren't objectives, but a set of instructions on how to achieve a higher objective he hadn't written down. At the time I was insistent that it was his job to tell me what I needed to achieve and my job to work out how to do it.

Why is this important? It's important because I need to own my decisions. If I've worked out how to do something, I'm much more committed to making it work than if I'm told to follow a set of instructions imposed on me.

People need to own things. I've watched when an engineer has made a set of bad decisions and boxed themselves into a corner on a Friday afternoon. By the Monday morning they've come back with a resolution to the problem they'd created. They owned the problem and they owned finding a solution. If I'd told them how to do everything, there's no way they would have spent the weekend working out how to fix it. Ownership drives behaviour.

Too many organisations are like my old boss β€” attempting to design solutions at the top, rather than empowering those at the bottom. Once empowered, people will tap into their inner reserves and creativity. When micro-managed, they will wait for the next instruction.

πŸ‘¨β€πŸ‘¨β€πŸ‘§β€πŸ‘§ The primary job of a leader is to build the team

I've learnt that the overriding job of a leader is to build a team and culture that allows that team to perform.

Sometimes roles and responsibilities aren't clear, meaning team members tread on each other's toes.

Sometimes working practices and processes or culture need change because the team is inefficient.

Somtimes a team has the wrong mixture of skills - too much skill in one area, not enough in another.

Sometimes individuals might be lacking in motivation or underperforming for personal reasons β€” a leader needs to be tuned into personal dynamics, spot when someone's not happy and act as councilor to better understand how that person can be given support.

Sometimes an individual might be a "squeeky wheel", undermining others and creating dissent that impacts the productivity of others.

Sometimes a team's strategic objectives are unclear, meaning they're left floundering for direction.

All of these things are the responsibility of the team's leader. A leader needs to be alert to team dynamics, identify when something is wrong and be willing to take decisive action. Decisive action won't always be popular with everyone, but corrective actions are always best taken quickly and with conviction. Issues left festering only get worse, almost never better β€” it's always best to nip things in the bud early.

πŸ—£ Talk early and often

At Humanise.AI we were early to the remote working revolution, starting out 100% remote back in 2017. After four years of this, I've learnt one overriding lesson: talk.

When we're remote, people sometimes hide behind a messaging app like Slack. But Slack is no substitute for talking. Messaging apps can turn us into keyboard warriors, making us say things and taking positions we would never take face-to-face.

Slack can act as a kind of insulating layer, behind which it's possible to create our own alternate reality. In this alternate reality we can over-interpret the significance of what others say, over-react, dig our heels in and entrench our position. Over time that alternate reality gets more vivid in our minds until it explodes on contact with reality.

But if we'd just pick up the phone, we'd realise we'd got it all wrong β€” our alternate reality is not reality. The words of others didn't have the significance we thought they did. When we hear the voices of others, we realise their position is not so dissimilar to our own. And we give a little ourselves.

If we do use an email or messaging app to pursue a disagreement, only one thing happens β€” at each turn of the conversation things are amped up another step. Debates become more heated. Disagreements become stronger. Positions become more entrenched. Alternate realities become more vivid.

I've learnt to always pick up the phone. I've learnt to never use written words in anger or frustration and to never let others do the same. Online disagreements need to be diffused quickly by being taken offline. My emphasis has become to break the chain of disagreement and to never allow either myself or others the luxury of creating an alternate reality. The phone, not messaging or email apps, is our friend in building bridges.

😫 Opinion without experience is just guesswork

Sometimes people have opinions on things they have no expertise in. Some people have too many opinions. It's said that we were born with two ears and one mouth, and should therefore listen twice as much as we speak. In a word, yes.

I've learnt that having an opinion without expertise is just noise.

A common blind-spot for engineering teams is knowledge and expertise in the problem domain. Engineers know a lot about technology, but often very little about how hotels or lawyers operate. But that doesn't stop them guessing.

There's only one thing that keeps a technology team on the straight-and-narrow, and that's clients. Without a client, there's a risk of making stuff up. That stuff is nearly always the wrong stuff. I've learnt that if you don't have access to someone who knows the problem domain first-hand, you need to find someone quickly.

I now classify people, and how much I listen to their opinion about the problem domain, into three groups:

  1. People who are running a business right now, or have done so in the very recent past, and know the problem domain in detail.
  2. People who have access to speak to (1) and have an inkling about what the problems are, based on what they are told.
  3. Everyone else, who don't have a clue and are just guessing.

In case it's not obvious, (3) have not earnt the right to an opinion. Maybe they have value in rethinking how technology might change things. That's fair enough, but that thinking needs (2) or preferably (1) to be involved β€” otherwise, the (3) people have no idea what they are redesigning.

Modern life gives us a surfeit of opinion. A leaders job is to work out quickly who to listen to and who to ignore. Quality of opinion comes mostly comes from expertise and experience.

πŸ™ You can't manage what you don't understand

If you want to manage something, you have to understand it. Otherwise, you won't know when to believe what you're told, or not. You may not want to, or need to, in the detail all the time. But if you can't get into that detail, you're in trouble. Managing requires that you have some sense of how things should work.

That means a business person can't manage an engineering team. Likewise, a technical person cannot manage a business team. Unless, of course, such people are willing to roll their sleeves up and learn.

Sometimes it's tempting to "outsource" something, either by recruiting an expert to take care of it or by paying an agency to do it for you.

However, I've learnt that if that thing really matters, I have to invest time and energy myself to learn about it β€” regardless of who I get to help me. If I don't understand it, I can't manage either an internal or external team. I may not want or need to always get into the detail, but having the capacity and willingness to do so, is important.

🀫 Some things are better said privately

I've learnt that providing "feedback" with an audience is a bad idea. When someone feels they have been embarrassed, rightly or wrongly, their "fight or flight" response is more likely to kick in. If it's the fight response, you've got an unsightly public argument. If it's the flight response, they might not say things they should. Either way, it's a bad outcome.

Feedback is almost always better had in a 1:1 environment, where the lack of an audience prevents that flight or fight reactions and we're more likely to have the mature debate we need.

πŸ€“ Look for bright people

I once employed someone I suspected wasn't too bright, but whom I thought didn't need to be that bright for the role. I was wrong. Almost the first week I was getting frustrated that I needed to spoon-feed that person too much. I need people who can work stuff out and figure out how to get stuff done. I need them to take initiative. I need them to be smart, so they can tell me what I need to do β€” rather than me tell them what to do.

Steve Jobs used to refer to the "bozo explosion", where b-team players employ c-team players, etc. He was right. Always aim for a-team players.

Always employ smart people. The bozos will only drive you up the wall in frustration. But the smart ones will surprise you in a pleasant way.

🀣 Humour matters

Some people find humour flippant, a sign that a subject isn't taken seriously. I disagree strongly. Humour is a critical part of life. It sustains us through the difficult times. It helps us put things into perspective. It's life's steam valve.

I've learnt that the team who can laugh in the worst of times, is the one who's built resilience. In contrast, the team who takes everything seriously, and who can't see the funny side, are much more liable to become over-stressed at minor challenges.

Humour is important in business. I know that not everyone agrees with me, but that's my opinion honestly held.