Being opinionated at the top

OKCupid (a dating site from the 2000s) once published a blog post claiming that, according to their data, the way to find higher quality matches was to include more polarizing information on your dating profile, not less.¹

It’s a strategy of shrinking the number of people who might want to talk to you at all, in exchange for having a greater match with the (relatively fewer) people who still do.

I was thinking about this yesterday, in regards to quarterly planning.

Quarterly planning is also a multi-step process, where the CEO and executive team make decisions on strategy and key priorities, then each of the departments need to figure out what they can do to help achieve them.

The more opinionated and focused the exec team, the fewer options each of the departments need to consider and choose between.²

The more vague or open-ended the exec team is, the greater options that need to be considered and selected.

The problem is, the difference between these two is exponential. For vaguely defined strategies, the number of options to consider at the department level might grow from 5 to 10, multiplied by the number of departments who each need to do this work, plus all of the cross-functional meetings that now need to take place to assess those options.

Same number of company-level strategies, but double the amount of quarterly planning work for the departments.

Execs: being opinionated (while open to bottoms-up data, ideas, and feedback) is one way to really help your teams be more aligned and move faster.

The Workshop

This is a newsletter-only section where I share a half-baked idea in hopes that y’all who are smarter than me can work it out with me.

I got some feedback that Tuesday’s post on the dangers of results-oriented thinking might have implied the opposite conclusion. My intent is not to say that we shouldn’t care about outcomes. We care the most about outcomes, in aggregate. But what we shouldn’t do is interpret or extrapolate any one single result any further than is warranted.

To give an example, let’s say we have a hypothesis for what problem to solve, and a hypothesis for what would be the best solution to that problem. We then run a test (i.e. A/B test, or user test with a design prototype) and the test fails. It could be the case that we have the right problem but wrong solution, or we have the wrong problem. What we shouldn’t do is assume we have the wrong problem and give up on the direction all together just because of one failed test, unless there is something in the qualitative data that tells us explicitly that we should.

Same goes for strategy. Strategy is a hypothesis of what we think is the best path to achieving our goals. If, after executing some number of tactics, we’re not closer to our goal, is it the fault of the tactics or the strategy? It could be either; my point on probabilistic thinking is that we shouldn’t immediately assume the strategy is to blame, it could be just the tactics we chose are wrong.

Instead of a proper Workshop post today, I wanted to share a post I published yesterday in the Everything Marketplaces community, in case it’s valuable to you. If you’re not actively working on a marketplace but are new to Product, I hope you find some value in a basic tutorial on how to use user interviews to develop a design hypothesis and use user testing to validate it.

You can find it here.

1  Ah, the 2000s, how quaint. I have no idea what dating profiles look like these days, but if the state of our domestic politics is any indicator, I would guess showing what “team” you’re on across a bunch of different intersectionalities is more common now.

2  For example, at this moment in time, do we care more about growing new users or retaining existing users? Are we trying to expand broader or deeper? The trap some execs run into is they celebrate having broken down “grow revenue” into its constituent KPIs, and then stop there. “Our strategy is to grow these metrics, because if we do, we’ll grow revenue.” No. Your actual job is to choose which of those metrics to grow first, and to have hypotheses for the best ways to go about doing that.