I’m currently in the middle of a project where we are helping a large company test out an innovation lab based on the “innovation studio” ideas.
We are now in week 4. The beginning of the project was focused on learning, tons of custdev, and ideation. We’ve winnowed down our ideas to 6, and need to choose our first prototype to test (we’ll have up to 4 weeks to do initial validation of an idea). I tried a new exercise yesterday for filtering, borrowing from agile story estimation planning poker.
First, we had chosen and ranked our key filters from this broader list.
On the vertical axis we listed, in order order, our idea filters. On the horizontal axis we listed our ideas. Then we stepped through each idea, one at a time. Each “cell” in our matrix could get a 1, 2, or 3. A one meant that you felt really good about it. Two meant “ok.” Three was bad/worried.
The “sponsor” of the idea kicked things off by reviewing the core elements of the idea, and a few of the most important risks. For each idea filter, we allowed for questions and discussion of key considerations before voting happened. When people felt ready to vote, they held out a closed fist. We counted “1, 2, 3 vote!” and then everyone had to show a number of fingers: 1, 2 or 3.
If someone put out a 3, we started with them explaining their concerns. Then others could optionally share their perspective. If it felt like the group swung, we took a revote, otherwise went with the general consensus. Unlike story estimation, our goal wasn’t to go with the most conservative number but rather a rough consensus. In a few cases, I ended up putting a 1.5 or 2.5 to represent the split worries of the group.
We had one filter “investment capital required to beachhead,” where we had to establish some hard filters. The number 1 equalled under $500K; 2 = “$500K to $2M”; and 3 = “over $2M.” This number was intented to cover product creation, content creation, and initial sales and marketing costs to get to a reasonably proven, but not yet scaled, state. Think of it like money required to get to an A round.
The last item for voting was team passion, and it was good that it was last because it was highly influenced by the discussion from the previous items. Instead of taking a guesstimated consensus number, we actually counted the “votes” from each of our 6 people (1, 2, or 3). A lower score thus would be better. So far we have made it through 5 of our 6 ideas, and have two 8s and 3 at 14 or 15, so two clear winners.
My review of the exercise: it was a great way to focus the team around the ideas and our core priorities, and let to a great discussion. Voting was more useful for how it forced thoughtfulness and a mental judgement, rather than the output of the numbers themselves. In retrospect, rather than estimating a rough consensus number, which drives things to a lot of “2s,” I would probably do sums of the votes as we did for team passion.
All in all, it will probably have taken the team 3 hours to get through all six ideas across all 9 filters.