The news that Coca-Cola has closed down its’ Founders startup incubator has led to the usual raft of “I Told You So’s” saying that it’s because corporate startup incubators don’t work – but are they right?
I have been involved with three big companies launching internal startup “incubators”/”Labs” teams so far…and what have I learnt?
Firstly I think it is done for good reasons.
Startups normally have good people, energy, good ideas, leading edge tech, a low cost approach, a high risk appetite and a willingness to challenge orthodoxies. They just want to get stuff done.
Big companies normally have deep pockets, millions of customers, a great brand and experience of operating at scale.
Surely the best of both worlds would be for big companies to launch startups? What could go wrong?
Well, quite a lot as it happens.
Mistake #1 – Not defining what you want the startup to achieve
In other words what will success look like?
It sounds simple but there is spectrum of “innovation” ranging from far horizon R&D to tactical (incremental) improvements to existing products and services.
What is wanted/needed? Success needs to be defined and solved for.
In general the startup should not be given special treatment…”success” is to create a growing business that attracts and retains customers.
Mistake #2 – Not getting total organisational buy-in and executive sponsorship
Sooner or later the startup will be stopped dead in its tracks by a blocker if the whole of the organisation isn’t behind it – from the top downwards. With the best will in the world active and passive resistance will be met and the startup needs to be able to wheel in very senior people to call bulls*t and unblock things.
Mistake #3 – Integrating the startup so tightly that it effectively becomes another corporate department
The whole point of a startup is to do something different so if you insist it uses existing people, existing corporate assets and the current approaches you will get what you got. You need to achieve a loose coupling whereby the startup has the freedom to chose between doing something new and leveraging corporate infrastructure case-by-case based on what is best for the startup.
Mistake #4 – Getting the wrong people to “be” the startup
If you are going to create a startup you need to do what startups do. This is for a team to self-select great people passionate about what the startup wants to do. If you ask one of the big system integrators/consultants to create a startup within your company you might end up back in the worst-of-both-worlds territory. The people in the startup don’t need to be existing or new employees either necessarily – they just need the talent and drive to deliver the goals. Organisations like gro.team can supply high impact interim talent at short notice into these kind of situations very successfully. Possibly the best loved UK retail brand will soon be launching a new customer service (that is very different to its’ current business) using this approach and the gro.team person in that team is part of a startup that stands comparison (in terms of talent and effectiveness) with any of the many startups I’ve seen.
Mistake #5 – Not integrating the new product/services back into the Mothership
We created a Labs operation when I was CTO of Yell and we did nearly everything right. We hired great people, they developed a great micro-culture, they produced some really innovative products and services (including a really cool augmented reality app way back in 2010) but we never really managed to integrate the new things back into the existing business and deliver customer impact. For that reason it has to marked down as a failure overall.
Corporate startup incubators can be a graveyard for ambition (and there are a lot of traps for the unwary) but it can be done successfully – Telefonica’s launch of giffgaff and British Gas’s Hive are just two examples that come to mind.
As with a lot of things, in this area experience is a great teacher and nothing beats getting people involved who have been there, done it and have the scars to prove it.