Nickey Skarstad on Vision-to-Execution, Product Quality, and Second-Order Thinking

Nickey Skarstad is Director of Product at Duolingo. She previously worked at Etsy (seven years, director of product), Airbnb (experiences team), The Wing, and Shopify. She has a reputation for being unusually strong at both the strategic and executional ends of PM work — setting compelling vision and following through to shipping.

Key ideas

Vision-mission-strategy-objectives pyramid: work top-down. Vision is the 10-year aspiration; mission is how you make the vision real; strategy is the specific choices about what to do now; objectives (OKRs) are what you deliver this quarter. Each layer makes the next more concrete.
Quality metrics must balance growth metrics: a north star that only measures growth misses the question of whether what you are growing is good. At Airbnb Experiences, review rate was the north star precisely because it measured quality; at Etsy, first-sale-in-seven-days predicted long-term seller retention more accurately than shop count.
One-way vs two-way door decisions: give teams autonomy on reversible decisions; slow down and broaden consultation on irreversible ones. The discipline of making this distinction before a decision is most of what good decision process looks like.
Second-order thinking in product: every design decision creates downstream decisions. The larger and older a system, the higher the cost of not thinking ahead. First principles documents operationalise second-order thinking by forcing teams to agree on foundations before entering design.
Red/yellow/green calendar audit: a personal diagnostic tool. Colour-code every meeting after it occurs. If the palette is mostly red, the role or company is wrong.

Episode content

Vision-mission-strategy pyramid in practice

Nickey runs her visioning exercises in Miro or FigJam, not Google Docs. The reason: vision-setting is a creative exercise, not a writing exercise. Whiteboards, sticky notes, and spatial grouping produce different outputs than prose.

Her process: pre-fill the board with prompts (competitive landscape, user problems, open questions), invite the team into a timed session, synthesise together, then take the synthesis away for further refinement. The team does not need to leave the session with a finished vision; they need to leave with raw material the leader can work into a coherent statement.

The key discipline: don’t conflate the visioning process with the strategy decision. Visioning is generative; strategy is selective. Teams that try to do both simultaneously produce bloated documents that please everyone and commit to nothing.

After the vision exists, the question is whether OKRs connect to it. Nickey’s test: can every key result be traced to a specific strategic bet, which can be traced to the mission, which can be traced to the vision? If the trace breaks at any point, the OKR is not strategic — it is maintenance.

Quality as north star

The Airbnb Experiences team used review rate as its primary metric. Revenue and booking volume were tracked, but review rate was the signal the whole team optimised for. This meant that when a feature drove more bookings but worse reviews, the team knew not to ship it. The metric encoded the team’s actual ambition — not just growth, but experiences worth having.

At Etsy, Nickey observed a different version of the same principle: sellers who made their first sale within seven days retained at dramatically higher rates than those who did not. By adding friction to the onboarding flow (making sellers more thoughtful about their listings rather than rushing them open), the team slowed shop creation but dramatically improved first-sale rates. Fewer shops, but more successful sellers.

Both examples illustrate the same structure: identify the downstream quality signal (reviews, first sale) that predicts long-term health, make it a first-class metric, and let it constrain the growth metrics rather than just add to them.

One-way vs two-way doors

Nickey attributes the framework to Amazon but has applied it across every company she has been at. The practical question is: how reversible is this decision in six months?

At Airbnb Experiences, defining the quality standards for an “experience” was a one-way door — the standards would be embedded in coaching, policy, and product and would be very hard to change. The team spent months on it. On the same project, deciding what data to collect on the host application was a two-way door — easy to revise. Those decisions were made quickly.

The failure mode is treating two-way doors as one-way (paralysis by perfectionism) or one-way doors as two-way (speed that locks in the wrong structure). Teams that are good at this distinction move fast on what is reversible and slow on what is not.

Second-order thinking

The concept appears directly in Nickey’s newsletter and was reinforced by her reading of Donella Meadows’s Thinking in Systems. The core idea: every decision you make today shapes the space of decisions available tomorrow.

For product, the implication is about system architecture as much as feature design. A choice to build the listing title field as a 200-character free-text field determines what is possible downstream: how it displays on mobile, how it is indexed for search, how it can be translated. Changing the field later requires migrating millions of listings. The second-order cost of the original decision — small at the time — compounds to millions of dollars of engineering work.

Operationalising second-order thinking: ask “what becomes harder after this decision?” before making it. Write it into the spec. Do it as a team exercise so the knowledge is distributed, not just in the PM’s head.

Shopify diagnostic: red/yellow/green

The calendar colouring exercise emerged from Nickey’s time at Shopify. She had taken a platform product role and was unsure whether the difficulty was normal ramp-up time or a deeper mismatch.

She went back through weeks of meetings and coloured them: green for meetings that gave her energy, yellow for neutral, red for meetings that drained or bored her. The results were almost all red and yellow. That data gave her a concrete, rather than purely emotional, basis for the decision to leave.

She now recommends the exercise to anyone who is unsure whether a role is right. The act of colouring forces honest evaluation meeting-by-meeting, which is more accurate than the overall feeling at the end of a week.

Remote product management

Nickey worked fully remote through the COVID era and has developed specific practices. The most useful tools: Slack for asynchronous updates (send a Loom video rather than scheduling a meeting for status updates), Slack huddles for quick audio-only syncs, and Miro/FigJam for collaborative thinking sessions. The most important principle: do not rely on Zoom meetings to fill the communication gap that in-person spontaneity used to fill. Meeting fatigue destroys remote-team morale faster than any other factor.