Strategy Candy: Strategy | OKR | Product Management

The episode contains a handful of genuinely useful structural concepts—the Rolling Four cadence, L1-L4 levels, CTB vs RTB separation, and the exec review lightning-round format—but roughly half the runtime is consumed by personal backstory, Trello-for-knitting anecdotes, podcast recommendations, and Australian small-talk that carry zero operator value.

“every quarter we look back and we review results, then we look forward and we plan the next four quarters, the next quarter or two in detail and the further out quarters at a higher level”

“we talk about okrs that change the business and following metrics as run the business”

The 'Rolling Four' label and the explicit decoupling of OKR laddering from universal coverage are mildly fresh framings, but most ideas—focus over volume, shorter cycles, objectives as inspiration, outcomes over outputs—are standard OKR doctrine found in any introductory book. Nothing is argued from first principles or contradicts received wisdom.

“we've got a company objective. Win an AI. Probably most companies have that as a suggested objective right now”

“practice makes perfect. So I've just talked about how right now our company defaults more towards 12 month versus 3 months. If you're starting out, only do 3 month chaos”

Amber is a genuine practitioner who owns the OKR process end-to-end at a 15,000-person public tech company and has prior hands-on rollout experience at two smaller firms; that is real operational credibility. However, she is a Principal Program Manager, not a C-suite decision-maker, and her visibility into strategy outcomes is acknowledged to be limited in areas like the monthly business review.

“I lead the OKR rhythms across the whole company including the planning process, refresh process, the monthly review with the executive team”

“Two companies that I worked at, Wikia, which is now Fandom, and I actually was asked to introduce okay out there. So this was 10 years ago”

There are concrete structural details—four company objectives, four KRs each, L1-L4 taxonomy, 3/6/9/12-month KR durations, cameras-off 20-minute pre-reads—but the episode contains no outcome data, no named KR examples with numbers, no before/after metrics, and no evidence of what the OKR practice actually changed at Atlassian.

“we do okrs, krs that are three, six, nine and 12 months”

“we do the cameras off, pre read the page, come back. We usually have curated what we call a lightning round of people”

The host asks some reasonable structural questions about exec involvement and product-team fit, but consistently affirms rather than probes, allows extended personal tangents (Gippsland upbringings, Loom product pitches, book recommendations) that eat substantial runtime, and never challenges a single claim or asks about failure cases or tradeoffs.

“That's amazing. No, that totally makes sense.”

“I'm super conscious. I could keep going for hours about this, but I've only got 10 more minutes of your time”