Nightly repricing,explained.
The pricing page shows the number; this page shows how the agent gets there — the comps walk, the local events calendar, the demand + lead-time ramp, the decision logic, and the morning brief the host reviews over coffee. Closes with the CTA into /pricing.
Signal inputs
What the agent walks every night.
Four inputs feed the recommended rate — fresh comps in a six-block radius, the local events calendar, the demand + lead-time signal against the host’s base price, and channel parity so one rate lands everywhere at once. Each one runs nightly, not weekly.
Comps walk
The pricing agent scans a six-block radius for comparables — matching bedroom count, similar amenity tier, similar review velocity. It pulls fresh nightly rates (not the ones cached at the last booking) and weighs them by recent booking pace rather than headline price. Comps that have actually booked in the last 14 days outrank the ones that have only been viewed.
Local events calendar
Concerts, conferences, half-marathons, graduations, weather windows — anything that spikes demand for a cluster of nights lands in the demand calendar automatically. The half-marathon that fills four Saturdays does not need a per-property config; it lifts the rate on the affected nights and decays back to baseline on its own.
Demand + lead time
Base price × last 30 days of realized demand is the anchor. Lead time ramps the premium: under seven days is a rush premium, 7–30 sits at baseline, 31–60 takes a small early-bird trim, and 60+ takes a bigger one for the planning horizon you can fill at your leisure. The same four factors the dashboard runs — applied every night, not every week.
Channel parity
One rate gets pushed to Airbnb, Vrbo, and your direct booking site. The channel that drifts loses the booking, so parity is enforced, not aspirational — there is no manual reconciliation step, and no spreadsheet to keep three rate surfaces in lockstep at 02:00.
The four inputs match the four paragraphs in the pillar post — the same vocabulary, the same nightly cadence.
Decision logic
What triggers a nudge — and what stays reviewable.
Three rows cover the routine night, the event-driven surge, and the out-of-band drift the host gets to decide on. The same Auto/Manual switch from the rest of the crew decides the path; the host's floor / ceiling decides what “in band” means.
Routine — the 85% that runs automatically
The nightly walk completes against comps, the events calendar, and the 30-day demand signal. Rate plan refreshes at 02:00; channel parity drains out within seconds. The host sees the audit row in the morning brief at 02:04 and never has to touch the rate surface.
Event-driven surge
A half-marathon lands on the demand calendar; the agent lifts the affected Saturdays and decays back. Same shape for an in-town conference and a school graduation. The lift shows the compset that moved and the event behind it — so the new rate reads as forced by the calendar, not pulled from the air.
Out-of-band drift — proposed rate parked
When the compset pushes the proposed rate above the host’s ceiling or below the floor, the change does not publish. The agent parks it on a tap-or-skip card with the reason inline — the host raises the ceiling, locks the rate as a one-off, or skips the proposal. The Auto/Manual switch decides the path; the floor / ceiling defines the band.
The third row — proposed-rate parked for review — is the bridge into the Auto/Manual surface. For the full tap-or-skip flow, see the Auto/Manual deep-dive.
What the host sees
The morning brief at 06:14.
The host never has to read a rate plan before coffee. The pricing agent writes the audit row into the same tap-or-skip card the inbox agent and the turnover dispatcher use — the same model that decides what reaches the queue decides what the host reviews.
Pricing agent · 02:04
Friday–Sunday 12% lift — City Marathon weekend ahead.
Comps within six blocks jumped $38–$45 over the last 48 hours. Local event: City Marathon. Rate plan refreshed; the affected four nights are now at the new band.
Pricing agent · Tuesday 02:04
90-day rate plan refreshed — 3 of 47 nights flagged out-of-band.
The routine walk resolved 44 of 47 nights automatically. Three pushed the proposed rate above the host’s ceiling or below the floor and are parked for review on the tap-or-skip card.
Manual override · 06:18
Long-stay guest — back half of January held at last week’s rate.
A 21-night confirmed booking triggered a host override on the back-half. The pricing agent stops recomputing those nights; the rate surface stays put across the next nightly run.
Safety rails
The host-set floor and ceiling — and how the rate band sits inside them.
Every workspace starts with an explicit per-property floor and an explicit per-property ceiling. Both are host-set on day one, surfaced in plain English, and reviewed before the pricing agent writes the first rate. Tonight’s walk produces a recommended nightly rate — the band decides whether the recommendation lands in the queue or publishes automatically.
The lowest the host will accept on a quiet Tuesday in February.
The rate below which a booking is worse than an empty night. Anything the comps or the events calendar push below this number gets parked on the tap-or-skip card with the reason inline. The host can raise the floor, lower it, or skip the proposal — the pricing agent respects whatever the host sets.
The highest the market will pay on a peak Saturday in October.
The rate above which demand drops off faster than the rate climbs. A compset that drifts past the ceiling reads as a guardrail trigger, not a price hike — the proposed rate gets parked for the host, not pushed on the next channel. The host can raise the ceiling, lock in the rate as a one-off, or skip the proposal.
The floor and ceiling are the bridge from pricing into the Auto/Manual surface. The band goes wider or tighter based on what the host sets, and the tap-or-skip flow that accepts or skips the proposed rate rides on top. See Auto/Manual — how the surface decides what reaches your queue.
Cross-links
The comp walk in long-form, against the rest of the crew.
Pricing shaped to your roster
See the three Roostr plans.
Three plans, three roster sizes, one Roostr crew. The Boutique Operator tier matches the four-capability 24/7 framing most hosts looking at this page land on — Solo, Boutique, or Portfolio, picked by the size of the roster and the controls the host wants to keep.
Want the full crew teardown? See how Roostr runs the week, the unified comparison with separately managed tools on /compare, or the three plans on /pricing.
Get your nights back
Get your nights back from nightly pricing work.
Join the Roostr waitlist to spend fewer nights checking rates and more time trusting a clear, reviewable pricing rhythm. The crew keeps routine work moving while the decisions that need you stay in your hands.