2.5 years — the average life of an innovation team

Organisations are variously good or bad at innovation. Whether they fit into the former or the latter category, all seem to feel the need from time to time to set up an innovation group that sits outside the normal brand teams or business categories.

These days that group has a sharper-edged name – an "AI Lab," a "Transformation Office," a "Venture Studio," a "Future of X" team. But the playbook for setting one up hasn't changed much in twenty years. The arguments still run along one or other of the following lines: (a) we're getting lots of innovation, but it's all incremental – we need fewer, bigger, better ideas, (b) we're getting good innovation in our current categories, but we're not opening up new areas and our people are too busy doing their day jobs, or (c) our innovation success rates are not good enough, we need to develop better capabilities and put some real resource behind it.

Hence the decision is made. A dedicated team gets created, usually 6–8 people, sitting outside the normal business. They get objectives related to one or more of the three areas above.

In principle this seems sensible. Innovation is a specialist discipline. Marketing teams are usually too busy delivering this year's plan. And it really does take a dedicated effort to move an established business into new territory.

Despite this apparent common sense, it remains rare to find examples of the approach delivering its intended results. More likely the team will follow the course below – which I've now watched, in one form or another, in well over a dozen cases.

Year 1 — Setting up the space

The team sets about recruiting, defining its vision, and most importantly creating its own "space." The space is the corner of the office set aside for the new team, repainted in something the office services manager describes as "energising," desks replaced with standing desks and writeable walls, and a selection of clever-looking creative tools dotted around – the modern equivalent of the wacky play toys we used to install in plastic containers in the corner. There may even be a table football machine, on the grounds that people from other departments will drop in and create a "one team" feel.

The team also invites every innovation agency and AI consultancy under the sun to come in and present. The hope is to find some inspiration that will help them create their "own" process for ideation, plus the obligatory funnel-and-gate system, plus an "innovation operating system" rendered in Notion or Miro that gets shown to anyone who'll watch.

Another first-year favourite is the global trend tour: Tokyo, Singapore, Austin, New York, São Paulo, Tel Aviv. Each city visit produces a deck and a "what this means for us" slide. The trend tour has the added benefit of being expensible, building team cohesion, and giving people something interesting to put on their LinkedIn.

That's year one pretty much over – doesn't time fly. At this point the leader of the team needs to provide a progress report to the sponsor. They will report that the team is now in place, that they've developed a systematic process for creating and assessing ideas, that there are some interesting insights coming out of the trend work (these will look more interesting than they really are because they will be shown in a beautifully designed Notion workspace) and that they are well placed to start creating significant opportunities next year. The sponsor will already be a little worried that nothing very tangible has been delivered, but will hold back on the basis that this is a "creative" team that needs time and space to perform.

Year 2 — The promising work

The team starts to get down to what is often some genuinely promising work. Key opportunity areas get identified and projects convened around them. Ideas get developed, new technologies explored – these days that means LLMs, computer vision, agentic workflows, IoT – and concept boards created. Rough business cases get produced. The smarter teams work closely with brand, commercial and R&D teams throughout to mitigate the famous Not Invented Here syndrome.

About halfway through the second year the team's original sponsor leaves the business. Their replacement, naturally, wants a progress update.

Looking at the figures she can see the cost of 6-8 headcount, plus the budget for external agencies, plus AI tooling subscriptions, plus the trend tours. In total this has almost certainly exceeded £1m (the cost of the standing desks and Miro Enterprise licenses is hidden in another budget). Looking at the new concepts the team has developed she is quite impressed — there are some interesting ideas – but, new in post, she needs more. Specifically, she needs an early marketplace success to justify the spend.

With six months to go before the budget review, the team sets about deciding which of their concepts will produce the quickest wins. Spreading their bets, they pick three to commercialise. This will require additional budget and dedicated commercial resource. Their request is only partially met. They get half the budget and part-time resource from finance and operations, but they'll need to do the bulk of the development work themselves.

By this point it's already too late. They've hit the classic trap of spending all their time and money on the "fuzzy front end." The capabilities they built were about generating concepts – assuming they would hand the ideas off at concept stage to "someone from a brand or commercial team." Unfortunately, those people are still more interested in doing the things they get paid and bonused for than taking on new high-risk projects mid-year that aren't on their official objectives.

By the end of year two none of the projects have been delivered. All of the team bar two have been "redeployed." The remaining two — usually the leader and one senior manager — are given a small budget for the first half of year three to consolidate the learning and create a training programme in their "way of innovating" to transfer the capability back to the brand teams.

Six months later the training programme has been built and piloted, but the leader has resigned to start his own innovation agency (or these days, to become an "AI Transformation Advisor" on LinkedIn). The remaining manager stays in post another year to be a kind of "internal consultant" on innovation, facilitate workshops, run the occasional hackathon, before eventually being absorbed into the strategy team to focus on "foresight" work.

Total elapsed time, sponsor to sunset: about two and a half years.

So what's actually going on?

This is a slightly tongue-in-cheek look at the corporate innovation team – and there are plenty of teams out there doing excellent work. But the pattern above is real, and it's been remarkably stable for two decades. The AI era has changed the vocabulary and the tools, but it hasn't changed the underlying mechanics. Innovation teams still fail for the same reasons.

The common mistakes:

The team's mandate is fuzzy. "Innovate" isn't a brief. Dedicated teams are usually best focused on the bigger bets — stretching the existing business model and superseding it altogether — not on the incremental work the brand teams should be doing anyway.

The capabilities are skewed to the front end. Ideation, foresight and insight are usually over-resourced. Commercial, finance and operational delivery skills are usually under-resourced. The team needs at least as much "make it land" capability as "make it up" capability. And much of the front-end work can be outsourced more efficiently than it can be built in-house.

Sponsorship is too junior or too unstable. Innovation work runs on a 2–5 year cycle. Sponsors tend to last 18 months. Without genuine board-level sponsorship and a realistic investment horizon, the team is dead the moment its champion moves jobs.

The team doesn't check what's already in the building. The winning ideas are almost always hiding inside the business already — abandoned proposals, half-built prototypes, things the R&D team has been trying to get funded for years. Finding them, commercialising one or two quickly, and getting some early wins on the table buys the credibility everything else depends on.

Acquisitions, JVs, partnerships and licensing are excluded from scope. They shouldn't be. They are often the fastest path to genuine new capability.

No one plans the reintegration. The hardest part isn't generating ideas or even commercialising them. It's getting the ideas back into the operational business at scale. That requires redeploying some of the innovation team into operational roles, and — critically — aligning the reward and incentive plans of the operational team with the success of the new initiatives.

The headline statistic in this piece — 2.5 years, the average life of an innovation team — was true when I first wrote it nearly twenty years ago. It's still true today. The reason isn't that companies have got worse at innovation. It's that the way most of them set up innovation teams hasn't really changed.

If your organisation is thinking about setting one up, or restarting one, or quietly winding one down — there is a better way.

David is the founder of redebuter.com, a consulting practice focused on helping leaders reboot, rethink, reinvent and reignite their businesses. Before redebuter he co-founded and led Happen GP, the Innovation and Analytics Group now part of Accenture. Earlier in his career he ran the Brand Planning team at Coca-Cola GB and was a Managing Director at Synectics. He has a PhD in Consumer Behaviour from Cranfield School of Management.

Web: www.redebuter.com · LinkedIn: www.linkedin.com/in/drdavidwalker

#Innovation#Leadership#AI#Redebuter