JTBD: Falafel

Google’s Liz Reid, VP of Search:

“One of the interesting things about the evolution of AI is that people stop talking just in keyword ease as much, and they start expressing more of what they want, and then that becomes much easier for us to give an answer. If you say, tell me about someone versus what’s new with someone, that’s actually much easier for us to figure out how to give better results than if all we just say is someone. We used to talk about an example query in a different way is falafel.

What do you want to know with falafel? Some people don’t know what falafel is, they want a definition. Some people want recipes.

Some people want to find where to eat. Some people want nutritional information. They all just use the word falafel, and that’s just harder to figure out how across all of them do that.”

From Odd Lots: Google’s Liz Reid on Who Will Own Search in a World of AI, Apr 23, 2026

That’s jobs to be done! Take something that someone said, and figure out the why and progress they want to make in their life. This whole section of the podcast is good because it’s about the idea of translation.

In the case of Google, it’s about a computer translating what progress a person wants to make. But that’s the whole idea of JTBD.

May the fourth be with you.

Measure What Matters (book review)

Measure what matters book review

There are two aspects – contents and context – to John Doerr’s 2018 book, Measure What Matters, a book about OKRs (Objectives and Key Results).

Content.

Objectives are “what is to be achieved, no more and no less.” Grow the blog, lose weight, or strengthen important relationships.

Key Results are ways to “benchmark and monitor how we get the objective.” List ways to grow the blog, lose weight or build relationships. “It’s not a key result,” Marissa Mayer would say, “unless it has a number.”

Straightforward enough. Is this a book that could have been a blog post?

Maybe, but Doerr offers a trio of cautions.

Warning 1: OKRs are not a way to show activity, they are to focus attention and weigh the opportunity cost. Organizational achievements, not ego appeasements.

Warning 2: Sometimes incentives hijack the Key Results (Goodhart’s Law). An antidote is paired counterparts. In the Wells Fargo cross-selling scandal the Key Result of open accounts could have been paired with monthly active accounts.

Warning 3: OKRs are a tool to use not a dogma to follow. If objectives change then OKRs change too.

Context.

Context is a Bob Moesta word encompassing who, what, when, where, why, and how? Steak and hot dogs are ‘good for dinner’ within the right context. The same goes for OKRs.

Doerr is a venture capitalist at Kleiner Perkins. OKR adopters include Intel – where Doerr learned from Andy Grove – and Google among other technology firms.

The OKR system, Doerr wrote, “was a great impedance match, a seamless gene transcription into Google’s messenger RNA. OKRs were an elastic, data-driven apparatus for a freewheeling, data-worshipping enterprise.”

Google was a perfect match. But your business may not be.

OKRs, as Doerr presents them, requires a certain culture. Part of their effect is to argue well. Andy Grove set the Intel culture for OKRs because Andy Grove was at Intel. Once he left the culture changed from bettering to bullying.

Doerr has many examples, one of which is Zume Pizza, but they’ve gone out of business. What’s the right lesson in that? What’s the context?

OKRs are lightweight, malleable tools. But their usefulness varies. Will OKRs be OK for you?