Marketplace Whack-a-Mole: Managing Winners and Losers
Most consequential marketplace changes create winners and losers; manage the trade, don't chase a metric.
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 4
- Confidence
- 90%
Because a marketplace has finite attention and inventory, moving it toward one group takes it from another. Many high-impact changes cannot show a pie-expanding win in the short run; they are reallocations and often bets on future value. Managing a marketplace means consciously deciding whether the winners you create matter more to the business than the losers.
Origin
Ramesh Johari; the winners/losers framing credited to Sachin Salgaonkar (referenced as 'SAS Salen'), a CFO at upwork who later went to Thumbtack.
Core principles
- 01A lot of marketplace management is just moving attention and inventory around
- 02Consequential changes usually create winners and losers rather than expanding the pie
- 03Short-term metrics may stay flat or dip because you're reallocating; the payoff is a longer-run bet
- 04Rolling with change means judging whether your created winners outweigh your created losers
How to run it
- 1
Expect reallocation, not free lunches
Assume a meaningful change will shift attention/inventory from one cohort to another rather than lift everyone.
Watch out When metrics 'keep moving around' month to month, you may be playing whack-a-mole — fixing one side's experience degrades the other's.
- 2
Name the winners and losers explicitly
For any change, identify which side/cohort gains attention and which loses it, even though naming who gets hurt is uncomfortable.
Watch out Nobody likes articulating that a feature hurts some participants, but refusing to name it hides the real trade-off.
- 3
Judge whether the winners matter more
Decide, against your business objective, whether the winners you've created are more valuable than the losers you've created.
Pro tip Define a short-run metric that actually captures the trade-off (e.g. retention of the badged cohort), not just aggregate bookings.
- 4
Treat some changes as reallocation bets
Accept that some right moves won't show observable pie-expanding wins now; they're bets that the reallocation pays off later.
In the wild
Working with one company, the team saw a struggling new supply side, so built bespoke features routing them to more experienced counterparties. Those metrics improved — then the established counterparties' experience worsened, so the team re-matched them, and the next month the numbers moved yet again.
→ Demonstrated that marketplace fixes reallocate rather than expand: helping one cohort predictably hurt another until leaders judged which winners mattered most.
Common mistakes
Expecting every change to lift all cohorts
Finite attention means most changes trade one group's gain for another's loss; chasing an all-boats-rise outcome misreads the system.
Killing a reallocation change because the short-run metric is flat
Reallocation bets (like a status badge) may not move aggregate metrics immediately yet still be right; a flat A/B test isn't disproof.
Is it for you?
Best for
Marketplace PMs and leaders evaluating features that shift attention/inventory between supply and demand cohorts
Not ideal for
Single-sided products where a change doesn't trade one participant group off against another
From the transcript
“a lot of marketplace management is moving attention and (51:30) inventory around”
“many of the changes that are most consequential create winners and losers and rolling with those changes is about recognizing whether the winners you've created…”
“nobody likes to articulate the idea that a feature change is hurting some of the people in your Marketplace”
From the episode
Marketplace lessons from Uber, Airbnb, Bumble, and more
Ramesh Johari (Stanford professor, startup advisor)