10x Engineers: Fool’s gold or worth their weight in gold?
Charlie Sell
As COO, I lead our EMEA business, who offer global solutions to our clients talent and transformation challenges. Our core practices are in life science, engineering, legal, business transformation and technology.
I’ve noticed a bit of hype around the concept of ‘10x engineers’ recently, and I wanted to share my own opinion, plus some great views from my network.??
Bringing in a true rockstar engineer sounds like a good idea. They can solve engineering problems others can’t, and are supposedly 10 times more productive than their peers.?
But if you’re not careful, you could create risks that outweigh the benefits. However, there is a way to avoid the dangers and get the best out of 10x engineers in the long term.?
Before we get to that, what does 10x really mean?
It’s very hard to quantifiably measure if someone is twice or three times as good as other engineers, let alone 10 times. It’s unlikely someone is 10x faster, or designs systems that are 10x more efficient. For me, it’s a metaphor for someone who's significantly better, who is the whole package. There’s no shortage of good engineers in the world, but you can tell when you meet someone who is operating at a completely different level.?
Theo M. is a Full Stack Software Engineer who gave me his really interesting take for this newsletter:??
“10x developers don’t exist, especially if you’re talking 10x more productive than the average professional developer. Ten is just too big a multiplier when it comes to human ability. Even Messi can’t beat a team of 10 averagely skilled professional footballers, he’d lose 100-0. The tallest man in the world is nowhere close to 10x taller than the average.?
However it may be that the best developer can be 10x better than the worst. The Coding War Games was a famous study from the 1980s which found that the best teams were 11x better than the worst. I’d be interested to know if anyone has more recent data or a larger data set. Also I'd be interested to know how AI assistants used effectively might help to create a developer or team that’s closer to 10x.”?
If there are people who are better even than the elite – let's call them unicorn engineers - is it healthy to try and find them, and can you just drop them straight into your organisation?
Why wouldn’t you hire a 10x engineer?
In my experience, elite engineers often don’t do well in permanent roles. Some know they are the smartest person in the room, and that can disrupt a team culture.?
They may code for themselves instead of looking for the simplest, most elegant solution, so they create a complex code base that’s impenetrable for other engineers who come after them.?
领英推荐
When you have a rockstar, it’s understandable that you rely on them completely, especially for urgent fixes that only they can do. That short-term mindset takes focus away from the strategies needed to meet long-term challenges, such as creating good documentation, cultural alignment and scaling your team.?
A 10x engineer’s code is also the product of one mindset, rather than a collaborative effort that considers different factors. You might end up with architecture designed by one mind, creating risk. This is what some call a high ‘Bus Factor’, where a very high dependency on one engineering genius ends in disaster for your business if they cross the road at the wrong moment.
A better model
For those and other reasons, 10x engineers may not thrive in traditional team structures. But that doesn’t mean their talents can’t be harnessed. They can find a more suitable home in specialist impact-driven consultancies that channel their expertise into building 10x teams, not 10x individuals.?
By helping other engineers build as a team, the unicorn creates a better flow of continuous code delivery and reduces the Bus Factor.?
Engineers generally want two things – autonomy and alignment. They want ownership of their code, and they want to be aligned with business goals. The model I described earlier – the lone genius – has high autonomy but low alignment with long-term business goals, a dangerous combination. However, a team that has been augmented by a true 10x engineer will have high autonomy and high alignment, a recipe for effective performance.?
In a 10x team, there’s a healthy culture of inspection and feedback, as people continuously review each other’s code, and as a result, they are always learning from each other. When a 10x engineer is embedded with a team for a period of time - what we call engineering team augmentation - they excel, sharing knowledge, showing their workings and helping others raise their game.?
Business Agility Coach & Consultant Amanda Colpoys has helped the likes of Moonpig, Gousto OVO Energy and Bumble innovate and scale through agile approaches. She gave me her 2 cents on 10x engineers:?
“I believe team performance is more important than individual performance - particularly in software engineering where great results are achieved through successful cross-functional collaboration toward a shared goal.? For me, a 10x engineer is, firstly, someone who optimises team and company success over individual or function success. Secondly, it’s a person that raises everyone's game - not simply in terms of technical ability, but in role modelling great behaviours which positively influence their peers. 10x teams matter much more than 10x engineers, but a “real" 10x engineer can be a powerful influencer that will increase your chances of building a 10x team.”
Couldn’t have put it better myself.?
Do you have a different view on the 10x engineer debate? Have you had an interesting experience with elite engineers or building 10x teams? Thanks for taking the time to read this piece, and please do share your thoughts in the comments section.?
CTO | Engineer | Co-Founder at Adroit Cloud Consulting | Building Scalable Solutions for Modern Enterprises
1 年A good read! Great points on autonomy and alignment.?Thank you.?
Expeditionary CTO. Co-founder at Refute
1 年In my experience, they do exist. But they are certainly not best measured about how much code they generate/what percentage of the code base they own. Instead, they are best characterised by their ability to make measured decisions and pick the right tradeoffs, informed by an innate problem solving affinity and a thorough understanding of the business and product mission. In some people, this comes naturally with experience. In others, instinct plays a part. But in either case, the retrospective measure is a near-perfect record of choices - big or small - made along the way. Last, but not the least, I can count the people who I have met throughout my career (20+ years) and fit this description on the fingers of one hand.