Team size as an indicator of success

Jan 6 2017

I’m really interested in the discussion about what is probably just around the bend for us as a civilisation; our next “Industrial Revolution”. However, instead of this revolution being about labour (~1900s) or information (~1990s); it looks like it’ll be about automation, machine-learning and ultimately A.I.

I doubt this will be news to anyone. People have already been discussing this for years, but I think the issue is going to more pressing as we start to see real, concrete evidence of automation in our everyday lives. A good example being Tesla’s Autopilot 2.0.

Where does that leave you and me?

Just as CGP Grey outlines in the above video, it won’t be long before the machines come for me and my job - yours too probably! How long before a computer is better able to build a site than me?

Fortunately, ‘The Grid ‘ has shown we’re not there quite yet, but it can’t be long. I’ve been following their development for a few years, and I have no doubt that they’re going to nail it soon. If not them, then someone else!

With that said, I’m not too worried. Just as the Industrial Revolution brought us labour-saving machines, and the Information Age brought us improved communication, I know that the Automation Age will bring a lot of benefits to my work. I already automate whatever parts of my workflow that I can, but there’s still a lot left!

Crucially, all the bits that are likely to be automated, at least in my lifetime, are the bits that I would love to automate today if I could. People may not realise, but the work of a web developer is still a largely manual process, where lines of code are written by hand. All the frameworks, code-generators and time-saving aids are just that; aids. As good as some of them can be (they can also be terrible), they aren’t capable of actually doing the work. No framework can take the abstract idea in your head and present it, fully formed, ready to go!

Don’t get me wrong, I love opening up my text editor, grabbing a piping hot mug of black coffee and combining the blank screen in front of me with the idea in my head and turning it into something real. That said, I’d also love if much of the repetitive code could be automated. A good codebase follows structure, is well documented, has tests to maintain correct functionality. That’s all a lot of manual work, that is much less exciting than the actual building.

Instead of building all of this with explicit instructions in what’s called an imperative language; I’ve love to use a declarative approach where I use code to explain the outcome I want, not the steps to get there. Automation and a degree of A.I. could allow for that type of work environment!

The future of startups durable business

This has implications for business as much as it does for our own, individual work. The successful businesses of the future will need to fully embrace automation. Our current startup tactics probably aren’t going to work in this sort of environment, and large, entrenched corporations probably won’t fair too well either!


Automation is going to allow smaller teams - maybe even solo business owners - to compete at scale; just as they did at the start of the Information Age. We saw individuals with great ideas use software as a force-multiplier to overthrow big business. Whilst big business was caught up with the processes and red-tape it had enshrined as a method to replicate their past success; individuals with no doctrine found new ways of thinking and problem solving.

We’re witnessing it at the moment too. Tesla (and in the future Apple & Google too if the rumours are to be believed) are preparing to pull the road out from underneath traditional car makers. Given how reliant cars of the future are going to be on incredible software, rather than building hardware (engines) who is more likely to do that well? Apple or Volkswagen? Throw in the fact that technology companies have been tackling battery technology for years, and they have a significant head start!

What to takeaway from this

My hypothesis, finger in the air, is that team size will become an indicator of success as automation becomes more prevalent. Being small won’t mean playing small if machine-learning and automation is embraced. However, it will allow a small company to avoid the “systems” and “process” millstones that slow down larger companies, and remain flexible and responsive to changing technologies. By the time that big business is forced to take note and follow suit, they’ll have a lot of catching up to do.

If you’re around 25 years old, and starting a business today, and want it to be your life’s work, the way I see it at least; you’ve got two choices.

  1. Start a business that absolutely, 100%, unequivocally relies on humans to produce value that a computer could never recreate. (Do remember though, we haven’t even scratched the surface of what computers will shortly be capable of - so be careful!)

  2. Start a business that has automation, a small but focussed team and a great idea at it’s core. Leverage whatever automation is available to you, and keep doing it! Keep your business small (in terms of team size and running costs) and stay flexible. Don’t be afraid to make bold leaps into new emerging markets and technologies. “Durable business” tactics will be probably prove to be more successful that “startup” tactics.

So what am I doing to ensure I’ve got a place once the machine’s steal everyone's job?

I’m choosing to stay small, and choosing tools and technologies that will allow me to work at a scale far beyond what individual developers have previously been capable of.

On the other hand, I could have got all of this completely wrong; and A.I. will be our ultimate downfall, just like every film you've seen!

Back