Impact Above All Else: The Two-Question Priority Loop
Two daily questions, driver-mapping, and back-of-envelope math before any project earns your time
- Difficulty
- Moderate
- Time to result
- ~ongoing to results
- Steps
- 5
- Confidence
- 95%
Vlad Loktev's operating system for going from IC PM to GM of Airbnb's core business was a single obsession: impact. He asked two questions every morning — what are the company's top priorities, and what can I do to dent them — then invested dedicated calendar time into learning the underlying business drivers so he could answer the second question credibly. Every project was pressure-tested with back-of-envelope math before it earned resources.
Origin
Developed by Vlad Loktev across his ~10 years at Airbnb, where he went from IC growth PM to GM of the homes business. He credits it as the mindset (not skill) behind his trajectory, and cites the Instant Book project (circa 2014) as the canonical example of driver-led project selection.
Core principles
- 01Priorities are usually easy to identify; figuring out how to make a real dent is the hard part.
- 02If you are not working on a top company priority, you should probably get a different job.
- 03You cannot claim impact without understanding the drivers of the business.
- 04Dependency-thinking ('can the other team do Y so that I can do X') is a trap — focus on what is in your control.
- 05Assumptions can be crazy, but the mechanics of the impact case must be written down.
How to run it
- 1
Ask the two questions every morning
Question one: what are the highest priorities for the company today and over the next ~3 months? Question two: what can I do to make a meaningful dent in those priorities?
Pro tip Keep the horizon short (3 months). It forces you onto priorities that are live rather than aspirational.
- 2
Block calendar time to learn the drivers
Set aside dedicated thinking time away from your day job to understand why users behave the way they do in the product, which levers exist, and what happens if you move each one. Build a mental (or literal) flowchart of the business levers.
Pro tip Loktev physically blocked recurring calendar time purely to think through drivers — treat it as a standing appointment, not a spare-time activity.
- 3
Run back-of-envelope math on every project
Before committing, write the assumption chain: which sub-lever this project moves, by what percentage, and what that translates to for the top-line business metric. State what you have to believe for it to be a success.
Pro tip Present the opportunity as a lever flowchart with a modelled lift, not as a narrative of why it feels important.
Watch out Assumptions being uncertain is fine. Having no assumption chain at all is not — that is when good opportunities get rejected for lack of a lift estimate.
- 4
Drop the dependency framing
Audit how much of your day is spent on 'can team X do Y so that I can do Z'. Convert as much as possible back into things you can do yourself. Let go of the dependencies you cannot move.
Watch out As orgs scale, this is the default failure mode — people quietly stop having agency and start managing other teams' backlogs.
- 5
Position at the intersection of priorities and your spikes
Map a Venn diagram of the company's real priorities against your own strengths/spikes. Work in the intersection; if you are not there yet, gravitate toward it by doubling down on a nearby strength.
Watch out The worst outcome is sitting in a job where you are neither working on a top priority nor using your spike.
In the wild
Loktev points to Airbnb's Instant Book work (around 2014) as driver-led rather than opportunistic. The team had reasoned about the core problem with Airbnb at that moment in time and modelled what impact the change could have on the overall trajectory of the company before committing.
→ Instant Book became one of the defining marketplace changes at Airbnb, and Loktev cites it as the clearest case of the impact-first selection method producing a company-level result.
Lenny Rachitsky brought Loktev a project pitch with strong reasoning but no lift estimate. Loktev drew out a flowchart of all the business levers, located the sub-lever the project actually touched, and showed what moving it 10% would produce for the business.
→ The opportunity was confirmed as genuinely big — but now it was presentable and fundable, because the impact case was explicit rather than intuitive.
Common mistakes
Assuming you are already impact-driven
Most people hear 'work on impact' and assume they do. The tell is where your energy actually goes day to day: if it goes into chasing other teams' dependencies rather than levers you can move, you are not doing this.
Skipping the math because the assumptions feel unknowable
The point of back-of-envelope math is not accuracy, it is forcing you to articulate the mechanics of what must be true. Refusing to model because inputs are fuzzy means you never test whether the project can matter at all.
Is it for you?
Best for
ICs and product leaders at fast-growing companies who feel busy but suspect their work is not moving the business.
Not ideal for
Roles where the job is stewardship, compliance, or maintaining stability rather than moving business metrics.
From the transcript
“my main mindset was impact above everything else”
“what are the highest priorities for Airbnb today and over the next like 3 months?”
“just immerse yourself in understanding why certain behaviors exist on the product that you're working on”
“every single project that we worked on at Airbnb, we did back of the envelope math with assumptions to help us understand the impact that…”
From the episode
Airbnb’s Vlad Loktev on embracing chaos, inquiry over advocacy, poking the bear, and “impact, impact, impact” (Partner at Index Ventures, Airbnb GM/VP Product)