Why the Funnel model doesn't work for small process innovations

---

I saw a presentation the other day from one company, which included a description of their Ideas process. This process was like a suggestion scheme,  based on a funnel model, with online submission, and a series of filters which refined the idea until it was accepted. The presenter describing the system said that ideas were sought on improvements at all scales, and that to date they had received 35000 ideas and implemented 1000 of them.

You can hear this statistic, and think "Hey, great, 1000 implemented ideas", or you can think "Hmmmm - 34000 disappointed people and wasted opportunities".

This level of filtering seems to be typical in Western companies, with about 5% or 10% making it through the filter.

Toyota, on the other hand, implemented about 80% of the ideas they got from their staff. Not much of a funnel, eh?

Now, I fully support the funnel model for product innovations, or for innovations that require significant investment. You cant afford to do everything, so you need to filter out the best ideas. But for little improvements? For little innovations to process? Why not do them all, unless there is a good reason not to?

Here are several reasons why the funnel model doesn't work for small process innovations.

• The funnel model requires voluntary submission. But why not ask people instead? Regularly ask the teams, and the communities of practice - "how could we do this better, faster, cheaper?" 
• The funnel model is reactive, not proactive. It funnels all ideas, good and bad, rather than focusing  on creaign good ideas.Why not introduce review processes such as After Action review, and ideation processes such as BDAL or quality circles? Focus on creating, rather than filtering.
• The funnel model requires individual written submission. But why not talk about the ideas in the team, or in the community of practice? Have a good discussion around "how could we improve this idea?
• The funnel model is based on filtering and rejecting. Why not look for a model based on implementing? Why not say "How can we make this work?"
• The funnel model is centralised. But why not look for opportunities to implement ideas locally? Only the ones that need cross-organisational change need centralised decisions.
• Filtering decisions are made by a committee. But the committee may not be near enough to the work - why not let the relevant manager decide? Why not let the decision be made by the lowest possible level? If a plant supervisor or factory manager could implement the idea, then there's no need to go to the ideas committee.
• The funnel pre-supposes that only a small percentage of ideas will make it through. That's what a funnel does - it narrows down. So what would you call the Toyota system, where 80% of ideas were implemented? A Freeway, perhaps? A wind tunnel? Unless big resource is needed - unless you are looking at New Product Development - why assume you need to filter at all (unless absolutely necessary of course).

Maybe that is what we need - the Idea Freeway, rather than the Idea Funnel.