Eigenquestions as Answers
Occasionally I will run across a great framework that sounds really useful, but then struggle to find an appropriate way to apply it in practice. One such framework in my back pocket was Eigenquestions (see link in further reading section). Basically an Eigenquestion is a domino question, the answer to which leads to a cascading series of decisions, both now and in the future. It sounds magical, I know, but I had yet to apply it even though I’d discovered it several years ago.
Fast forward to last week, when I was having a mentoring session with a PM in my network. This PM was struggling with an experiment, which hadn’t reached stat sig yet - further, he didn’t have hope that it would reach stat sig, because there was an opt in step to try out the new UX. He also had a stakeholder who was convinced (without data) that the change could negatively affect a tension metric, so now the team was stuck in a stalemate. We went in circles for a bit, debating reasons why the opt in had to be in place, where the stakeholder intuition came from, and whether to expand the target audience pool.
I asked whether the org cared more about confirming experiments had a positive lift, or avoiding decision deadlock. Because this 1 experiment debate was probably a microcosm of how decision making was dragged out in this org. And there it was, the Eigenquestion: do you value validation of impact or speed of iteration? This helped give the PM I was chatting with a framing for how to go have the conversation with his leadership team and unblock several other people in the org who were dealing with similar stalemates.
All this time I had been looking to apply the concept because it was intellectually interesting - but what I realized from this discussion was that an Eigenquestion makes sense when you’re trying to establish a principle. It’s a framing device for uncovering a latent principle, and principles are how you accelerate decision making.
When do you need an evergreen principle vs just a one-off answer?
when there is no data, conflicting data, or anecdata
when you are having a debate for the nth time
when the decision will have a cascading effect on follow-on decisions
when you’re starting an org from scratch
when you’re revisiting a long-standing ritual that’s atrophied
As always, I’d also love to hear from readers about their experiences with Eigenquestions, decision making, and principle setting - please chime in via comments👇 or join the chat via the Substack app.
And if you enjoyed this post, please consider subscribing.
further reading / references
this is a good read on stat sig 101, along with some common pitfalls (exactly like the ones my mentee was facing)
I’ve mentioned Eigenquestions before in my list of mindbending models
if decision making frameworks are of interest, check out
childish drawing / interpretation