Why innovation culture dies after the launch event

Innovation culture usually dies after the launch event because the event creates expectations the organisation has no system to meet. Hundreds of ideas arrive, nobody has the time, budget or authority to act on them, and staff watch their suggestions disappear. Within months, the event has taught people the opposite of what it intended: that speaking up changes nothing. What works is starting smaller, with the follow-through built before the announcement.

The belief, stated fairly

The reasoning behind a big launch makes sense. Culture change needs visible commitment from the top. A launch event, whether an ideas campaign, an innovation day or a hackathon, signals that leadership is serious, gets people talking and gives everyone a chance to take part. It creates energy, and energy seems like the hardest thing to manufacture.

John Kotter's well-known 1995 Harvard Business Review article on why transformation efforts fail lists "not establishing a great enough sense of urgency" as the first error. A launch event looks like a good answer to that.

The problem is what Kotter also lists: "declaring victory too soon" and "not anchoring changes in the corporation's culture". A launch event is often treated as the change itself, rather than the start of it.

Why it fails in practice

The event produces more ideas than the organisation can process. An open call for ideas across a few thousand staff can easily produce hundreds of submissions. Someone has to read them, assess them, reply to them and do something with the best. Usually that's a small team with no extra capacity. The backlog builds, replies slow down, and most ideas get no response at all.

There was no route from idea to action. Even the good ideas need an owner, a budget line, some time and permission to test. If those weren't agreed before the launch, every promising idea has to fight for resources individually against existing priorities, and most lose.

The ideas weren't aimed at anything. "Share your ideas to make us better" invites everything from car parking to strategy. Without named problems, ideas scatter, quality varies wildly, and it's hard to justify acting on any one of them.

Silence is read as rejection. Staff who submitted an idea and heard nothing don't assume the process is slow. They assume leadership didn't mean it. This is the most damaging part, because it confirms what sceptics already believed.

Leadership attention moves on. The launch had the CEO on stage. Three months later, the CEO is focused on budgets, and the innovation team's updates are item eleven on the agenda. Staff notice where leaders spend their attention.

The next launch is harder. Each failed campaign makes the next one less credible. Organisations that have run two or three find that the most capable people no longer take part.

What works instead

Build the follow-through first, then announce it modestly.

1. Name a small number of problems before you ask for ideas. Three to five, specific and owned by a senior leader each. "How might we cut repeat visits for document collection?" gets better ideas than "share your ideas".
2. Agree capacity before launch. Decide how many ideas you can genuinely test in the first quarter (it might be five), who will test them, how much time they get and who signs off. Don't invite more than you can handle.
3. Promise a reply, and keep the promise. Every person who contributes gets a response within a fixed time, even if it's "not now, and here's why". Use a stakeholder communication plan to decide who hears what and when.
4. Triage openly. Sort ideas with a simple, visible method, such as an impact–effort matrix, and publish the result. People accept "not selected" much better when they can see the reasoning.
5. Show the first results fast. Aim to show something real, a test result or a changed process, within eight weeks. Small and real beats large and promised.
6. Put innovation on the standing agenda. A short slot in the regular leadership meeting, not a separate committee that meets when convenient. Review what was tested, what was learned and what was stopped.
7. Plan the second and third cycles before the first ends. Use a light roadmap so people can see that this is a rhythm, not a one-off. Run a retrospective at the end of each cycle and change one thing.

If you've already had a launch that fizzled, don't launch again. Go back to the people who submitted ideas, acknowledge honestly what happened, tell them what will be different, and act on two or three of the original ideas visibly. Repairing trust is slower than building it, but it's possible.

An example

Illustrative. A national library service in Canada runs an "innovation week" with keynote speakers, workshops and an online ideas portal. More than four hundred ideas come in. The two-person innovation office acknowledges each one automatically, then takes four months to review them. By the time ten are shortlisted, the sponsoring executive has moved roles.

A year later, a new director tries again, differently. She names three problems (one is "people who come for printing never discover the other services") and asks branch managers which they want to work on. Six branches volunteer. Each gets one day a week of a staff member's time for six weeks, and the director commits to reviewing results at her monthly management meeting.

There's no event. The first message to staff is a short note naming the three problems, the six branches and the date results will be shared. Eight weeks later, she shares four test results, including two that didn't work, and names the staff involved. The second cycle gets twelve volunteer branches without any campaign.

Common mistakes

Do

• Build capacity and decision routes before the announcement.
• Ask for ideas against named problems, each with a senior owner.
• Reply to every contributor within a set time.
• Show real results, including failed tests, within two months.
• Keep innovation on the regular leadership agenda.

Don't

• Launch with a big event and an open ideas portal you can't staff.
• Treat the number of ideas submitted as a success measure. (See counting ideas is the worst innovation KPI.)
• Leave contributors in silence while you review.
• Hand everything to a small innovation team and move on.
• Run another launch to fix a failed one. Repair first.

Questions people ask

Why do innovation initiatives fail? Usually not for lack of ideas. They fail because there's no capacity, budget or decision route to act on ideas, because the ideas weren't aimed at specific problems, and because leadership attention moves on after the launch.

How do you sustain innovation after a launch? Build a regular rhythm: a small number of named problems, a fixed cycle (six to twelve weeks), visible results, and a standing slot on the leadership agenda. Sustaining it is about routine, not events. The pillar on innovation culture covers the full set of habits.

Should we still run hackathons or innovation days? Yes, if they're connected to a system that carries the results forward. A hackathon is useful when owners, budgets and next steps are agreed in advance. See why most hackathon winners never launch and, for leadership behaviours that keep things moving, 5 things great innovation leaders do.

Where to go next

To build the follow-through: the impact–effort matrix for open triage, the stakeholder communication plan for keeping contributors informed, and the retrospective for improving each cycle. More in the method guides.