A debate at this week’s wonderful Names Not Numbers event has crystallised something that has been bouncing around my head for a while. I’m done with the idea of industry-specific software platforms. Here’s why:
1. Platforms automate out humanity
Part of yesterday’s debate was agreeing where the boundaries of machine and human lie in our business world. Alongside a tech investor we had two online logistics companies- the CEO of a generalist logistics and delivery company, and that of a company that is providing a platform to provide aggregated ordering and delivery from local shops.
Putting aside the issue that my good friend Chris Weston raised (“how depressing that both of these companies are based around the idea that people have to spend too long at work these days to shop.”), both are ultimately about building a scalable service that can be applied into many different locations. Scalable is great if you want a return on investment, or if you want a uniform service, but not so if you want any individualism. The loss of individualism leads, in my take, to a loss of humanity. McDonald’s might provide a consistent service and product, but time to chat with customers is managed out of the system in the way it isn’t in my local Fish & Chip shop. When the service becomes computer-centric there is even greater risk that staff (if they are actually staff) and customers alike become merely the fleshy robot arms of the machine.
2. Platforms are parasitic
Notice the rash of restaurant delivery services that have sprung up recently? Or the apps to hail taxis? Or those to book a table? Or those to obtain beauty salon services?
No doubt the founder stories here will be some romantic tale of a botched date or a last minute emergency that led to the lightbulb moment of “What if we just… ?”
But there are too many of these platforms around run by too many people in formerly City-type jobs to not imagine that the stories more commonly start with “what is a service or retail segment that is currently dominated by independent small companies that doesn’t yet have an industry specific online platform?”. Has anyone launched funer.al for the “recently deceased yet still time poor?” customer segment yet?
The risks are low (certainly in comparison to the investment in staff, stock and bricks and mortar required by all of the serviced businesses) and the potential rewards massively unequally distributed for the successful platforms. It’s difficult to see this as much other than parasitic.
3. Platforms stifle industry innovation
And here’s the thing. Generalist platforms (a generic CRM, a website host, that kind of thing) do something that is generic. An industry-specific platform however runs the significant risk of seriously halting anything but incremental tinkering. If the platform is specific to an industry it will compel companies in that industry to operate in certain ways to conform to the platform. It’s something I’m seeing in the legal industry with Case & Matter Management systems. Where’s opportunity for service differentiation in the world of personal transport in an industry dominated by Uber? Everything, over time becomes top down, outside-in.
4. Platforms are there to serve tech investors first
Which of course brings us to the real reason why platforms are great: they are great, first and foremost, for tech investors. They don’t involve the nasty business of building real infrastructure. They do offer the promise of hockey stick returns.
Yesterday’s tech investor panel member cited that the three top skills in demand in the startup scene at the moment are developers, data scientists and growth hackers. Those latter two are the twin horsemen of incremental twiddling. Of marginal gain platforms. Of not very human interaction between individuals and small businesses beholden to the machines…
A quick thought on this… Platforms automate out humanity – yes but only to the extent that is designed in. You could buy a car built in a 50s production environment with ‘Friday afternoon’ jobs and what is sometimes described as ‘character’ – i.e. the car doesn’t work all the time. Automation in manufacturing has removed humanity. Others may feel that’s a bad thing, I’m happy to know the damn thing will start in the morning! If you build a platform that is designed to remove all human interaction you will, unless the scope is *tiny* build a platform that will fail. Human intelligence is still a vital component of most systems and the challenge is to strike the balance between automation and ‘intuition’.
Laughed out loud at the ‘recently deceased but still time poor’ though 😀
Agree, Chris. It’s just when automation approaches are applied to human to human interactions it tends to go off the rails. Like in Customer Service…