Pricing cockpit & multi-channel rate tester: five flavors
Mockups for the Airbnb Q4 program’s two generic surfaces. Sample data uses real medians from the Oct 15 research (rents $183–$292, 23-day median lead, 77-night window). All pricing surfaces are channel-agnostic with direct included. October 2026, internal draft.
Recommendation up front. Ship Cockpit B (Concierge) as the default entry for most hosts, with Cockpit A (War Room) as the portfolio view for PMs, and Cockpit C (Ribbon+) as the embedded layer in the calendar hosts already live in. For the tester, lead with Tester B (Parity Radar) because mismatch is concentrated (144 accounts carry the skew; 78.6% of dual-adjuster accounts are already equal), and keep Tester A (Quote Compare) as the drill-in. Reasoning at the bottom.
Who are we designing for?
| Avatar | Scale in our data | Job to be done | Failure mode to design against |
|---|---|---|---|
| Solo host | Median cohort account: 1 listing | “Tell me the 1–3 things costing me bookings, in plain English, fixable from my phone.” | Dashboard blindness; jargon (LOS, adjusters); settings archaeology |
| Portfolio PM / revenue manager | Top 200 accounts = 45% of cohort listings | “Show me exceptions by dollars across 118 listings; bulk-fix; prove pacing to owners.” | Per-listing click work; no $ ranking; owner-consent paper trail missing |
| Dynamic-pricing power user | 65% of cohort (PriceLabs 8,638 listings) | “Don’t re-price me. Show me rules my tool can’t fix: orphans, blocks, parity drift, layering.” | Opaque stacking (our adjuster on their spot rate); duplicate min-stay logic |
| Channel mixer / direct-booker | Direct + Vrbo + Airbnb + BDC mix is the norm | “Keep my guest price sane everywhere without favoring a channel.” | Airbnb-branded tooling; hidden adjuster/fees drift |
Pricing cockpit: three flavors
A. War Room
Portfolio PMPower userDesktopTrading-desk density: pacing vs last year, dollars at risk, and lever cards priced like bets (cost to pull, expected nights and dollars). Portfolio table sorted by unbookable dollars. Best when someone runs revenue as a job.
B. Concierge Feed
Solo hostMobileWeekly digest = same feed“Three things need you this week.” One plain-language card per problem, one button per card, evidence in one line. No charts to interpret. The Monday email is this feed, unchanged.
C. Ribbon+
Everyone, in-flowBuilds on shipped ribbonNo new destination: orphan nights render loud in the multi-property ribbon (the day model already carries GapType), a problems-only toggle, and an inspector that explains why a night cannot book and fixes it in place. Zero new navigation to learn.
Multi-channel rate tester: two flavors
A. Quote Compare
Revenue managerAd-hoc checksPick property, dates, guests. Four columns (Airbnb, Vrbo, Booking.com, Direct) break the guest total and host net into nightly, discounts, cleaning, channel fee, and taxes, with the culprit line highlighted. The honest diagnostic.
B. Parity Radar
PMChannel mixerAlways-on scanNightly scan across the portfolio: only listings whose all-in guest price drifts >5% between channels surface, each with the named culprit (adjuster, fee, tax) and an inline fix. Matches the data: most accounts are already at parity, so silence is the correct default state.
Why this recommendation
- Entry friction decides adoption. The median account has one listing and a phone. Concierge (B) is the only flavor they will finish. War Room (A) would be a beautiful ghost town for them, and the digest email needs B’s card shape anyway, so B does double duty.
- Dollars live in portfolios. 45% of cohort listings sit in 200 accounts. They need A’s exception table and bulk actions; giving PMs the concierge feed alone would read as a toy. Same data model, two densities.
- Meet hosts in the calendar. Ribbon+ (C) costs the least new surface and puts the fix at the moment of seeing. The API already ships GapType; this is styling plus an inspector, shippable inside the quarter.
- Parity is an exception problem, not a homework problem. Radar (B) respects that 78.6% are already equal and hunts the 144-account skew. Quote Compare (A) stays as the drill-in and the Dianna-facing proof tool.
- Shared spine: one warehouse-driven model (pacing, orphans, Rent, rules) feeds all five surfaces. Build order: C (ribbon emphasis) + B feed for Q4, A portfolio table next, Radar scan nightly from day one, Quote Compare as the drill-in.
All five pages are static mockups with sample data. Nothing here is wired to live data.