Added basis to GET /market_report/{market_id}/time_series. Pass basis=stly to get each stay date's values as they stood around the same time last year, under this year's stay_date and with the same fields, so the rows of both bases can be joined on stay_date. It is in early access: accounts without early access get a 403, and a market whose last year figures are not available yet returns a 404. basis=current, the default, returns today's values as before.
Behavior change:GET /preferences/{listing_id}/changelog is no longer limited to the last 90 days. A request returns up to one year, ending at end_date, and older history is read by moving end_date back. A start_date more than 90 days ago used to be moved to 90 days ago; it now returns up to a year, so such a request can return more events than before.
Behavior change:GET /preferences/{listing_id}/changelog leaves out Settings changed events on the end_date day, like every other event type. It used to include them, so paging back with adjacent ranges returned a settings change made on the shared day twice. The default end_date (tomorrow) is unaffected.
Added event to GET /preferences/{listing_id}/changelog to return only some event types, e.g. ?event[]=Settings changed&event[]=Custom rates to leave out automatic price posts. Without it, every type is returned as before.
Documentation fix: the id on a checkin_checkout rule pairs the check-in rule with its check-out rule. It is derived from the rule's contents and was documented as a stable identifier, which it is not: it changes when the rule changes. Requests don't use it to identify a rule (PUT /preferences/{listing_id} replaces checkin_checkout as a whole), so re-read it after a save rather than storing it.
Behavior change: a checkin_checkout rule's id no longer depends on the default priority the server fills in for its type, or on the key order and null fields it was stored with. Saving any preference used to change the id of a rule stored without a priority, because the save filled the priority in. A non-default priority, such as a prioritized time_based rule's, still counts, so that rule keeps a different id from the same rule at the default priority. Existing custom, monthly, time_based and day_of_week rule ids change once, on the listing's next save or read. seasonal and event rule ids are unchanged.
Documentation fix: demand_sensitivity_rules is shown in the Wheelhouse UI as Local events, under Pricing Engine rather than Model Weights, now that day of week is its own setting. The demand_sensitivity note category is labelled Local Events for the same reason. Field names, rule types and behavior are unchanged.
GET /listings/kpis can now rank listings on a calendar month, quarter or year instead of a rolling window. Pass period (monthly, quarterly or yearly) and period_start (the first day of the period, e.g. 2026-07-01) in place of window. Add compare=previous_year to also return the same period one year earlier as compare_value, with change and change_ratio, and rank_by=change or rank_by=change_ratio to rank on the year-over-year change, e.g. to see which listings improved most on last summer. The comparison is the earlier period's final figure, not the figure as it stood at the same point last year. basis=stly works with window only. Existing window calls are unaffected.
Added POST /feedback, which sends feedback about the API to the Wheelhouse team: a message, an optional category (bug, confusing_docs, missing_capability, praise or other), and optionally the endpoint and listing_id it is about. It accepts read-only keys and is not billed. See Feedback.
Added GET, PUT and DELETE /listings/{listing_id}/ota_links to read and manage a listing's links to its pages on Airbnb, Vrbo and Booking.com and to its direct booking website. See Listing OTA Links.
Added basis to GET /listings/{listing_id}/kpis, /kpis/monthly, /kpis/quarterly and /kpis/yearly. Pass basis=stly to get the listing's KPIs as they stood at the same time last year instead of today's, in the same response shape. Rolling windows are measured 52 weeks earlier. Each monthly, quarterly or yearly row holds the same period a year earlier, as it stood 52 weeks ago, under the current month, quarter or year it compares with, so the rows of both bases can be joined on that key. model_date, comp_set_count, the comp-set metrics, revenue_blocked, min_price_occurrence, revenue_score and the occupancy_neighborhood metrics are not available for last year and are null. It is in early access: accounts without early access get a 403. basis=current, the default, returns today's KPIs as before.
Added basis to GET /listings/kpis. Pass basis=stly to rank listings on their values as they stood at the same time last year, measured over the same window 52 weeks earlier, instead of today's. It is in early access: accounts without early access get a 403. min_price_occurrence, revenue_blocked, revenue_score and the occupancy_neighborhood metrics are not available for last year and return 400. Listings with no stats for last year are absent from the ranking. basis=current, the default, keeps the existing ranking.
Documentation fix: endpoints that act on one listing return 406 Not Acceptable when the listing is inactive. Only GET /listings/{listing_id}/neighborhood/pricing and /neighborhood/occupancy documented it. It is now listed on every affected endpoint and explained in Inactive Listings. No behavior change.
Fix:attr_seasonality on GET /listings/{listing_id}/price_recommendations and POST /preferences/{listing_id}/preview with attribution=true is now a whole number, like the other attr_* fields. It could previously have a fractional part or be null.
Documentation fix: the listing returned by GET /listings, GET /listings/{listing_id}, GET /segments/{segment_id}/listings and /sets/{set_id}/associated_listings documents meta, which it has always returned. No behavior change.
Documentation fix: updated_at on GET /market_report/{market_id}/time_series, GET /market_report/{market_id}/distribution, GET /sets/{set_id}/time_series and GET /sets/{set_id}/distribution is a timestamp such as 2026-08-30 04:17:22 UTC, as on GET /market_report/{market_id}/bedroom_groups, not ISO-8601. No behavior change.
Documentation fix: POST /listings/{listing_id}/sync, POST /teams/memberships/{master_user_id} and DELETE /teams/memberships/{master_user_id} return an empty JSON object. No behavior change.
Documentation fix: description on a note (GET, POST and PUT /listings/{listing_id}/notes) is null when the note has no text. Creating a note without a description has always been allowed, and the field has always come back null. The schema now marks it nullable. No behavior change.
Documentation fix: currency on GET /listings/{listing_id}/price_calendar and on calendar_snapshots in GET /listings/{listing_id}/calendar_day_history is null when no currency is recorded for the night. Both endpoints have always returned it that way. No behavior change.
Added asking_rate_available, the Asking Rate over available nights only: sum of nightly prices on available nights / number of available nights, excluding fees. Booked and blocked nights count in neither, nor do nights without a price, and it is null rather than 0 when no available night has a price. It is returned wherever asking_rate is: GET /listings/{listing_id}/kpis (every period in both directions), GET /listings/kpis, GET /listings/{listing_id}/kpis/monthly, /kpis/quarterly and /kpis/yearly, GET /sets/{set_id}/aggregated_metrics and GET /segments/{segment_id}/aggregated_metrics. Values fill in as each listing's stats are next regenerated.
Documentation fix: asking_rate averages booked and available nights, not available nights only, with each booked night at the rate it was booked for. Blocked nights are left out. Its description now says so. No behavior change.
GET /listings/{listing_id}/reservations takes date_filter_type=updated_at, which returns the reservations whose updated_at falls in the window from start_date (included) to end_date (excluded). With this filter type, start_date and end_date also take ISO 8601 timestamps such as 2026-09-01T00:00:00Z, so an integration can poll for the reservations that changed since its last request instead of re-reading every reservation.
Extended GET /listings/{listing_id}/kpis/monthly, /kpis/quarterly, and /kpis/yearly with asking_rate_highest and asking_rate_lowest (Asking Rate (Highest) and (Lowest): the highest and lowest nightly asking price on any night in the period), matching the rolling KPIs of the same name.
GET /listings/{listing_id}/price_recommendations returns events on each day: the events in the listing's market that run over that date. It is an empty array on a date that no event covers.
Fix:rates on GET /listings/{listing_id}/recent_changes now moves when a custom rate is removed outright, through DELETE /listings/{listing_id}/custom_rates, DELETE /listings/{listing_id}/bulk_custom_rates or the Wheelhouse app. It used to move only when a rate was added, changed or split, so an integration polling it never saw a removal. A rate expiring, or its dates passing, still does not move it.
Added market_name to the listing returned by GET /listings, GET /listings/{listing_id}, GET /segments/{segment_id}/listings and /sets/{set_id}/associated_listings: the human-readable name of the listing's market_id, so it no longer takes a GET /market_report call to resolve.
POST /preferences/{listing_id}/preview documents the attribution query parameter. With attribution=true, each day carries the same attr_* and historical_anchor_* breakdown as GET /listings/{listing_id}/price_recommendations, so the two responses can be compared to see what a preference change does.
Behavior change:POST /preferences/{listing_id}/preview now returns each day's min_stay, from the listing's min stay settings with the request's changes applied. It was previously always omitted.
Extended GET /listings/{listing_id}/kpis/monthly, /kpis/quarterly, and /kpis/yearly with bookings (Bookings: reservations with at least one night in the period), matching the rolling KPI of the same name. Quarterly and yearly values add up the months, so a stay that spans months counts once in each month it touches.
Added DELETE /segments/{segment_id} endpoint. Deletes a portfolio segment owned by the authenticated user. A segment assigned to a team member cannot be deleted until it is unassigned.
GET /listings/{listing_id}/price_recommendations and POST /preferences/{listing_id}/preview now return min_price on each day: the minimum price applied to that day, in the response currency. It is the higher of the minimum price rule for the date and min_min_price, except on days with a fixed custom rate, where only min_min_price applies. null when no minimum applies.
Documentation: added How minimum stay rules combine, with a worked example. A gap rule lowers the minimum stay that lower-priority rules set, rather than replacing it. Rule prioritization can raise time_based and gap rules above custom rules, and min_min_stay applies last. The min_stay on GET /listings/{listing_id}/price_recommendations is the final value and does not say which rule set it. No behavior change.
Documentation fix: base_price: null on GET /preferences/{listing_id}, GET /listings and GET /listings/{listing_id} means the listing's base price follows the Wheelhouse recommendation, multiplied by base_price_adjustment. It does not mean the listing has no base price. The base price in use is the top-level base_price on GET /listings/{listing_id}/price_recommendations, which is never null. base_price_adjustment is a multiplier (0.9, 1.0 or 1.1 for the presets), not a percentage. No behavior change.
Documentation: added Pricing Tiers, which lists what each tier returned by GET /listings/{listing_id}/pricing_tier allows: the price_recommendations horizon next to the price_calendar range, and which endpoints return 402, 404 or 429 for which tiers. Pro Trial and Pro Onboarding get the same 545-day price_recommendations horizon as the paid Pro plans. No behavior change.
Documentation: added Keeping a local copy fresh, on refreshing a local copy without re-reading every endpoint: webhooks for recommendations, reservations and flags, and GET /listings/{listing_id}/recent_changes for preferences and custom rates. No behavior change.
Extended GET /listings/{listing_id}/kpis/monthly, /kpis/quarterly, and /kpis/yearly with nights_calendar (Nights (Calendar): nights in the period, irrespective of listing status), matching the rolling KPI of the same name.
Documentation fix: POST /preferences/{listing_id}/preview returns 201, not 200. It always has.
Documentation fix: a tag's type is UserTag or ImportedTag, not user.
Documentation fix: exclude_inactive on GET /listings, GET /segments/{segment_id}/listings and GET /listings/kpis also leaves out listings that are not linked to a channel.
POST /listings/{listing_id}/sync returns 503 Service Unavailable with a Retry-After header while syncing with the listing's channel is paused, for example during the channel's scheduled maintenance. A refused sync does not count toward the daily limit.
Add GET /price_models, listing the price models customers can pick from, and price_model_id on GET /listings/{listing_id}/price_recommendations and POST /preferences/{listing_id}/preview to price with one of them, now that several can be available at once.
Behavior change:price_model=opt_in returns 400 when more than one opt-in model is available; pass price_model_id instead. price_model=current prices with the model the listing gets without a pick of its own.
Documentation fix: the GET /market_report description no longer says the response includes a market boundary. It does not. No behavior change.
Bug fix: PUT /listings/{listing_id}/bulk_custom_rates with an empty custom_rates array returns 200 with an empty updated_custom_rates. It previously returned 500.
Documentation fix: days_before and days_after on a time-based calendar rule are null when that side of the window is unset. Preference reads have always returned them that way. No behavior change.
Behavior change:GET /notification_settings and PUT /notification_settings include api_spend_alert, the alert sent when API usage in the billing cycle passes the account's spend alert threshold. The response lists only documented event types. A setting stored under any other event name stays as it is and is left out of the response.
last_minute_discount, far_future_premium and seasonality_adjustment now take smooth. When true, the value eases between rules instead of stepping at each rule's edge. See Smoothing.
Behavior change: a day_of_week setting of CON or AGG sent without rules now stores the matching preset rule. It previously stored the type with no rules and priced like REC. PUT /preferences/{listing_id}/{setting} now accepts day_of_week.
Behavior change:PUT /preferences, PUT /preferences/{listing_id} and POST /preferences/{listing_id}/preview now return 400 naming the field, rather than a 500, when checkin_checkout is missing check_in_rules or check_out_rules, or when occupancy_pacing has no global matrix or a values that is not one array per row with one value per column. None of these payloads could be saved before.
Sending gap_night for a listing with more than one unit now returns a warning with the new reasonnot_supported_for_multi_unit: gap night pricing does not apply to those listings, though the setting is still stored.
Documentation fix: on listings that accept length-of-stay premiums, weekly_discount, monthly_discount and long_term_discounts rule values are signed (-10 is a 10% discount). An empty check_out_rules leaves check-out unrestricted. No behavior change.
Documentation fix: occupancy_pacing.values was documented as accepting a flat list; only one array per row has ever been accepted. The check-in/check-out, gap night and day-of-week descriptions now state the value formats and how overlapping rules resolve.
GET /listings/{listing_id}/kpis/monthly, /kpis/quarterly, and /kpis/yearly now return future periods through the maximum calendar horizon (1.5 years from today), where they previously stopped 12 months ahead. Monthly responses carry up to six more future months, and quarterly and yearly responses can include later quarters and years. The furthest month, quarter and year include only the nights up to the horizon, so their nights and revenue cover part of the period. Past periods and the response shape are unchanged.
Preferences now include historical_anchoring_price_source, which sets which historical price historical anchoring weighs in: observed (the listing's price as Wheelhouse saw it on the channel once the date booked), posted (the price Wheelhouse last posted for the date before it booked) or posted_with_fallback (the posted price where there is one, otherwise the observed price). Every preferences read returns it, and PUT /preferences, PUT /preferences/{listing_id} and POST /preferences/{listing_id}/preview take it. Listings that never chose a source read observed, which is how they have been priced so far.
Documentation fix: the per-day values (sunday through saturday) on custom rates are numbers, not integers, on both reads and writes. Responses can carry decimal values such as 90.25, and the write endpoints accept them. The API's behavior is unchanged.
Documentation fix: an array query parameter takes more than one value only in brackets, for
example ?metric[]=asking_rate_w_fees&metric[]=occupancy. These docs previously showed the key
repeated without brackets (?metric=asking_rate_w_fees&metric=occupancy), which keeps only the
last value: GET /preferences returns one listing and DELETE /sets/{set_id}/associated_listings
removes one. This applies to listing_ids, dates, metric, market_ids, and room_types. A
single value can still be sent without brackets. The API's behavior is unchanged.
Breaking: portfolio segment filter_backend no longer accepts wide
listing_stats.{metric}_{window} numeric range keys (for example
listing_stats.revenue_0_30_gte). Performance filters use scoped keys
listing_stats_rolling:metric:{metric}:{window} instead (for example
listing_stats_rolling:metric:revenue:next_30_gte).
listing_stats.last_booked_at date-range filters are unchanged.
With attribution=true, GET /listings/{listing_id}/price_recommendations now names the
historical price behind each date historical anchoring raised:
historical_anchor_date, historical_anchor_price (in the response currency),
historical_anchor_source (observed or posted) and historical_anchor_pct, the
anchoring percentage in effect for the date. The fields are left out on dates anchoring
did not raise, and attr_historical_anchoring is unchanged.
Behavior change:GET /preferences/{listing_id}/changelog events can now have statuswarning, a partial success. A Streamline price post that posted some dates but was rejected for dates with no PM season set up in Streamline used to return error; it now returns warning, with the same msg saying which dates were rejected and how to fix them. Treat warning as needing the user's attention.
Bug fix: amenities on GET /sets/candidates and GET /sets/{set_id}/listings lists each amenity as its own array element. Since 2026-09-21 some responses returned the whole list as a single element.
Added fallback to GET /listings/{listing_id}/fee_impact_calendar. With fallback=estimate, a listing with no fee adjustment in effect gets a multiplier estimated from the fees on its recent reservations instead of 1.0. Each entry now carries source (setting, estimate or none) saying where its multiplier came from. Without the parameter the multiplier is unchanged.
Documentation fix: GET /listings/{listing_id}/neighborhood_pricing prices have always included the fees guests pay on top of the nightly rate, as in the Wheelhouse pricing chart. The description now says so and how to compare a listing's nightly rate with them. No behavior change.
Preferences now include historical_anchoring_mode, which sets how a historical anchoring percentage applies. blended, the default and how every listing has been priced so far, averages the historical price with the recommendation. anchor_only takes the percentage of the historical price alone, so 50% of a $1,200 historical price sets a $600 floor, below which the recommendation is raised. Every preferences read returns it, and PUT /preferences, PUT /preferences/{listing_id} and POST /preferences/{listing_id}/preview take it. Existing listings stay on blended.
Added category to each flag returned by GET /listings/{listing_id}/flags: engine, info, performance or settings, saying what the flag means for the listing's pricing. It is null for an older flag that Wheelhouse no longer raises. New flags can be added to any category, so decide how to treat a flag from its category rather than from a list of flag names.
Behavior change:GET /listings no longer returns each listing's photo URLs (meta.photos) unless you ask for them with the new include_photos=true query parameter. num_photos is still returned.
Added is_hidden to the listing object returned by GET /listings, GET /listings/{listing_id}, GET /segments/{segment_id}/listings and the /sets/{set_id}/associated_listings endpoints. It is true for a listing that has been hidden in the Wheelhouse app. Hidden listings are still returned and ranked with exclude_inactive=true, so filter on is_hidden to leave them out.
Behavior change: a user with restricted editor access to a shared listing can no longer change its tags with PUT /listings/{listing_id}/tags, nor update it through the batch PUT /preferences. The Wheelhouse app refuses restricted editors tag changes and updates to several listings at once. The batch endpoint is refused however many listings a request names: the listing is reported in errors with status 403, and the rest of the batch still applies. Update a single listing's preferences with PUT /preferences/{listing_id}, which restricted editors can still use.
Behavior change:PUT /sets/{set_id}/associated_listings now requires manage access to each shared listing, as in the Wheelhouse app. A user with viewer access to a shared listing could previously associate it with a set, and now gets 403. Removing a listing from a set with DELETE is unchanged.
Owners, editors and managers are unaffected by either change.
Each Wheelhouse user can now hold up to 5 active RM API keys. Creating another returns an error until the user holds fewer than 5. Users already holding more keep all of them. Team members each have their own limit. The key behind MCP access is not counted. See Direct API integrations.
Added owner_user_id to GET /listings and GET /listings/kpis. It narrows either response to the listings owned by one Wheelhouse account, matching owner_user.id on the listing, so a team member can read or rank one team's portfolio at a time by passing a membership's user_id from GET /teams/memberships. It grants access to no listing you could not already reach, so an account whose listings you cannot access returns an empty list.
Documentation fix: listing_preferences on the listing object (GET /listings, GET /listings/{listing_id}) was documented as an unstructured object. It has always carried a summary of the listing's pricing preferences — automatic rate posting, base price and adjustment, fees, long-term discount percentages, guest counts, nickname, and the global minimum and maximum price and minimum stay where configured — and the schema now lists those fields. The schema also now marks as required the fields that are always present on the listing object (id, channel), on price calendar, last posted price and minimum stay calendar rows, and on the pricing tier, check-in/check-out, min/max price and neighborhood responses, and lists the two custom_type values a price recommendation can carry. No behavior or response-shape change.
Documentation fix: name on GET /listings/{listing_id}/pricing_tier listed a tier the endpoint never returns and omitted two it does. The values are Free, Pro Flex, Pro Flat, Pro Trial and Pro Onboarding. No behavior change.
Behavior change:amenities on GET /sets/candidates and GET /sets/{set_id}/listings is now an array of amenity names. It was previously a single comma-separated string.
Documentation fix: optional fields on those listing responses (including guests_included) have always been null when unknown. Custom rate responses have always returned currency as null for adjustment rates, and a weekday value as null when that day is unset. The schema now matches. No other response-shape change.
Documentation fix: several response fields have always been able to be null or to carry a shape the schema omitted. The schema now matches what the endpoints already return. No behavior or response-shape change. In particular: optional reservation fields (including comments and monetary amounts); tag description; reservation_id on the price calendar and calendar-day history (a string, the channel reservation ID); custom-rate periods, which do not include id or listing_id on read; custom calendar rules, which may send day_of_week_values instead of value and may include an id; calendar-rule priority, which can be fractional; and legacy days_after / days_one_sided_gap on adjacency rules.
Preferences now document and validate fee_adjustments: the estimated share of fees guests pay on top of the nightly rate (total_fee_percentage, -100 or more) and whether it is refreshed weekly from the listing's recent reservations (auto_update). Every preferences read returns it, null when nothing is configured, and PUT /preferences and PUT /preferences/{listing_id} take it. The object replaces the stored one as a whole, and null clears it. Behavior change: the object was previously stored as sent, without validation; a fee_adjustments that is not an object, or a total_fee_percentage below -100, is now rejected with 400. GET /listings/{listing_id}/fee_impact_calendar continues to report the setting as a multiplier.
Documentation fix: custom_type on GET /listings/{listing_id}/price_recommendations and the preferences preview response has always been null when the night is a Wheelhouse recommendation, not an empty string. The schema now marks the field nullable and the description matches. No behavior or response-shape change.
The bedrooms parameter on GET /market_report/{market_id}/time_series and /distribution no longer takes a fixed set of five groups. Which bedroom groups a market is broken down by is derived from the listings in it, so a market with many larger listings is broken down further (0, 1, 2, 3, 4, 5, 6+) than one without them (0, 1, 2+). Read a market's groups from the new GET /market_report/{market_id}/bedroom_groups, which also reports how many listings sit behind each one. A group the market has no data for is rejected with 400, and the error names the groups it does have. Existing calls keep working where the market is broken down past the group they name: bedrooms=4+ is answered from the 4 bedroom group in a market broken down into 4, 5 and 6+.
Breaking: the geometry fields have been removed from the GET /market_report response — both the market's own geometry (the polygons describing its boundary) and the per-entry geometry under postal_codes (each postal code's boundary). Anything drawing market or postal code outlines from this endpoint will find both absent; the rest of the response is unchanged, postal_codes[].latlong included, so a centroid is still available per postal code. Set boundaries are unaffected — boundary on the dynamic set endpoints still returns MultiPolygon coordinates.
Documented installing the Wheelhouse Plugin from https://github.com/pricemethod/wheelhouse-plugin for Claude Code, Codex, and Cursor.
GET /sets now takes name and kind filters, so an account with many sets can find one without paging through the whole list. name is a case-insensitive pattern in which * stands for any run of characters — *Palm Springs* matches the phrase anywhere in the name, and a value without a * has to match the name in full. kind takes one value or several comma-separated, and kind=other also matches sets that carry no kind, which is what those sets already report. Sending both filters returns the sets matching each of them, and either combines with the existing pagination parameters.
Documentation fix: GET /preferences/{listing_id}/changelog was titled as the changelog for a listing's preferences, which undersold it — it has always returned every change recorded against the listing, including prices posted, failed posts and calendar syncs, custom rates added, split or removed, a change of pricing engine, and reservations imported. The title now says so, the description lists what is covered, and GET /listings/{listing_id}/custom_rates now points here for custom rate history rather than leaving it to be inferred that it carries none. No behavior or response-shape change.
Behavior change: preference settings the listing's channel cannot act on are no longer stored. Previously a setting the channel had no support for was accepted, saved and logged to the changelog — so the write looked successful and the changelog confirmed it, but the value was never posted to the channel. Such settings are now discarded instead, matching what the Wheelhouse app has always done. The rest of the request still applies and the response is still a success.
Preference write responses now carry warnings, naming each setting that was not applied, why (not_supported_by_channel, or partially_supported_by_channel when only part of it was too narrow for the channel), and the attributes it covered. The array is present and empty when everything applied, so it can be read unconditionally. Applies to PUT /preferences, PUT /preferences/{listing_id}, PUT /preferences/{listing_id}/copy and POST /preferences/{listing_id}/preview.
Listings and preferences now report supported_settings — whether checkin_checkout, minimum stays and long-term discounts can be set, and which form of long-term discount the channel takes. Check it before sending a setting rather than after. It reflects per-account early access, so two listings on the same channel can differ. See the Listing Preferences section.
Every response now carries X-RateLimit-Limit, X-RateLimit-Remaining and X-RateLimit-Reset, and a 429 additionally carries Retry-After. Your current standing arrives with work you were already doing, so pacing an integration no longer costs a request — and the 429 itself now tells you when to retry. See Rate Limits.
Added GET /usage, so an integration can inspect its own API usage rather than reading the audit log in the Wheelhouse app. It reports request volumes, error rates (including 429s) and response times across every RM API key on the account, broken down by day, by endpoint and by key, over a window of up to 7 days. Endpoint breakdowns are grouped by route pattern rather than exact path, so all calls to GET /listings/{listing_id}/notes count as one entry. Results are cached for a few minutes — see generated_at — and a window too busy to read in full is reported as window.truncated with the shorter window it actually covers.
Extended GET /listings/{listing_id}/kpis/monthly, /kpis/quarterly, and /kpis/yearly with additional metrics already persisted on the periodic stats tables: adr_fees, asking_rate, nights_available, nights_blocked, nights_bookable, nights_booked, revenue_available, revenue_blocked, revenue_fees, and revenue_fees_taxes. Names match the rolling KPI lexicon.
Added GET /sets/{set_id}/price_calendar, the API equivalent of the set calendar in the app. booking_probability_percent is bounded to 10–90 for unavailable nights, mirroring the app, which renders those extremes as <10% and >90% rather than asserting certainty about a listing it only scrapes; an available night reports 0. The rest: a night-by-night calendar for each active member listing of a set, with prices converted to one currency so comps priced differently are comparable. Paid sets only, and only for sets with at most 25 member listings; a larger set responds with 422. Both limits track where the data comes from: Wheelhouse requests the detailed per-night scraping that backs this endpoint only for paid sets at or under that size, so beyond it there would be little to return. The night object follows GET /listings/{listing_id}/price_calendar except that it carries no is_booked flag and no reservation_id, neither of which applies to a listing that is not yours. In their place, booking_probability_percent reports Wheelhouse's assessment that a night is booked rather than merely held, as a whole percentage.
Added PUT and DELETE /sets/{set_id}/listings, which add and remove a set's member listings — the comparison listings the set is built from. Previously members could only be chosen when the set was created with POST /sets, with no way to adjust them afterwards, so a set could not be refined once built. Identify listings by the internal listing_id values from GET /sets/candidates; only eligible comparison candidates can be added, the same restriction POST /sets applies. Both return the set's members grouped by status, matching GET /sets/{set_id}/listings, and both work on free and paid sets. Not to be confused with the associated_listings endpoints, which manage which of your own listings the set is compared against.
Fix:GET /sets/{set_id}/listings and GET, PUT, DELETE /sets/{set_id}/associated_listings returned 404 Not Found ("Dynamic set not found") for a free set. Building a set was therefore impossible before paying for it: you could not review which listings were in it, nor associate it with one of your own listings. All four now work on free and paid sets alike. The per-listing metrics returned by /listings were already available ungated through GET /sets/candidates, which returns the same object. The paid gate stays on the set-level analytics — aggregated_metrics, time_series, distribution and changelog.
Added canceled_at to the reservation object returned by GET /listings/{listing_id}/reservations. It carries the timestamp a reservation was cancelled, or null if it stands, and is set whenever status is Canceled — previously the status told you a reservation had been cancelled but not when. Note the single-l spelling, which matches the Canceled status value rather than the cancelled field of the PUT request body.
Fix:GET /segments/{segment_id}/listings and GET /segments/{segment_id}/aggregated_metrics evaluated a segment's filter only against listings the user owns, so a user whose portfolio is entirely made up of listings shared with them to manage got an empty result from every segment. Both endpoints now match owned and managed listings, consistent with GET /listings, and both accept include_managed_listings to control it (default true for RM API keys, false for channel integration keys).
Listings can now be identified by their Wheelhouse listing ID as well as by the channel's own listing ID. Pass channel=wheelhouse and listing_id is read as a Wheelhouse ID (the wheelhouse_id field of GET /listings, also returned as listing_id by GET /listings/kpis) rather than as a channel listing ID. Wheelhouse IDs are unique across channels, so no channel needs naming alongside them. This works everywhere a listing is identified, including the batched listing_ids parameters and copy_preferences_from. Under channel=wheelhouse a listing you cannot access returns 404, not 403, so the sequential ID space cannot be walked to discover other accounts' listings; the channel-listing-ID form keeps its existing 403. See Identifying Listings. Existing calls are unaffected — omitting channel=wheelhouse keeps the channel listing ID behavior.
Added the Webhooks endpoints: GET /webhook_events, GET /webhooks, POST /webhooks, GET /webhooks/{id}, PUT /webhooks/{id}, DELETE /webhooks/{id}, POST /webhooks/{id}/rotate_secret, and GET /webhooks/{id}/deliveries. Register an HTTPS endpoint to be notified when price recommendations change (recommendations.updated), when new reservations are ingested (reservations.ingested), or when new listing flags are detected (flags.detected), instead of polling for changes. Deliveries are batched over a five-minute window, name the affected listing_ids rather than carrying the changed values, and are signed with a per-subscription HMAC-SHA256 Wh-Signature header. See the Webhooks section for verification and retry semantics.
Added GET /listings/kpis, which returns one rolling-window KPI for every listing the authenticated user can access, ranked by value and paginated. It answers portfolio-wide questions — best and worst performers on a metric — in one call rather than one per listing. Choose the metric and window with metric and window; pass currency to convert monetary metrics before ranking so that portfolios spanning currencies order correctly. Each row carries listing_id (Wheelhouse's id), partner_listing_id (the channel's), value, currency, and updated_at. Comp-set metrics are not covered yet.
Documentation fix: GET /listings/{listing_id}/kpis/monthly, /kpis/quarterly, and /kpis/yearly were described as returning only history. They have always returned future periods as well — monthly spans up to 15 months back and up to 12 months forward, quarterly and yearly likewise extend up to 12 months ahead. The word "historical" has been dropped from their titles and each endpoint now documents its full window, with a note that current and future periods reflect bookings on the books plus asking rates for open nights and will therefore change as bookings come in. No behavior or response-shape change.
Behavior change:include_managed_listings on GET /listings now defaults to true. GET /listings returns every listing the authenticated user can access — owned listings plus listings shared with them to manage — where it previously returned owned listings only unless the parameter was passed explicitly. Pass include_managed_listings=false to keep the previous behavior. Use each listing's access_level and owner_user to tell the two apart.
Added owner_user to the listing response. It identifies the Wheelhouse account a listing belongs to (id, email, first_name, last_name), which makes it possible to tell whose portfolio each listing sits in when a response mixes owned and managed listings — for example GET /listings?include_managed_listings=true. Note that neither owner_name (the property owner name supplied by the channel) nor source_user_id (the host ID in the channel's own namespace) identifies a Wheelhouse account; their descriptions have been clarified accordingly.
Extended GET /listings/{listing_id}/kpis to cover the full Wheelhouse metrics lexicon, now served from a dedicated rolling-stats store. New metrics: adr_fees, asking_rate_fees, nights_bookable, nights_booked, nights_calendar, nights_percent_open, pickup_bookings, revenue_fees, revenue_fees_taxes, last_booked_days, lead_time, length_of_stay. nights_available, nights_blocked, revenue_available, revenue_blocked, and min_price_occurrence now also return backward-looking (N_0) periods; asking_rate_lowest and asking_rate_highest now return all periods in both directions; pickup now covers all backward-looking periods (7/14/21/30/60/90/180/365). Breaking:revenue_score is now keyed by cumulative forward-looking periods (0_7 … 0_365) instead of the non-overlapping 0_30/31_60/61_90 windows; the scalar last_booked_at field has been removed in favor of the last_booked_days metric; and each metric field is null until rolling stats have been generated for the listing.
Renamed the performance metrics returned by GET /sets/candidates and GET /sets/{set_id}/listings to the public Wheelhouse metrics lexicon, matching the other RM API KPI endpoints: anr → adr, apr → asking_rate, nrevenue → revenue, nrevpar → revpar, nrevpar_open → revpar_adjusted, open_occupancy → occupancy_adjusted, openness → nights_percent_open, los → length_of_stay, and *_nights → nights_* (each keeps its _365_0/_90_0 period suffix).
Added dynamic-set creation endpoints: GET /sets/candidates (search market listings with metrics near a location to build a set from), POST /sets (create a free set from selected listings), and POST /sets/{set_id}/upgrade (purchase the paid Dynamic Sets plan for a set to unlock its KPI endpoints). GET /sets and GET /sets/{set_id} now also return free sets, each carrying an is_paid flag.
Added GET /segments/{segment_id}/aggregated_metrics endpoint. Returns monthly performance metrics aggregated across the listings matched by a segment's filter, mirroring GET /sets/{set_id}/aggregated_metrics. Monetary values default to the currency of the segment's most common market.
Added GET /notification_settings and PUT /notification_settings endpoints. Users can now read which alerts they receive and enable or disable individual alert event/channel (in-app or email) settings via the API.
Added the Team endpoints: GET /teams/members, POST /teams/members/invite, DELETE /teams/members/{sub_user_id}, POST /teams/members/{sub_user_id}/auto_managed_listing_level, POST /teams/members/{sub_user_id}/assigned_segments, POST /teams/members/{sub_user_id}/refresh_auto_managed_listings, GET /teams/memberships, POST /teams/memberships/{master_user_id}, and DELETE /teams/memberships/{master_user_id} to manage team members and memberships.
Added POST /segments endpoint. Creates a new portfolio segment owned by the authenticated user. Accepts filter_backend to define the listing filter criteria.
Added PUT /segments/{segment_id} endpoint. Updates the name, description, filter_backend, and/or default status of an existing segment.
Added POST /listings/{listing_id}/sync endpoint. Triggers a manual sync that pushes the latest Wheelhouse price recommendations to the connected channel and refreshes reservation data — equivalent to the Sync button in the UI. Available on paid plans only; subject to a per-day rate limit and a 60-second debounce.
Added GET /listings/{listing_id}/base_price_history endpoint. Returns a daily log of Wheelhouse base price recommendations and effective base prices for a listing over the last 30 days (or a custom date range).
Added GET /listings/{listing_id}/calendar_day_history endpoint. Returns the history of price postings and calendar state snapshots for a specific stay date. Requires the Historical Price Changes feature on the listing's plan.
Extended GET /listings/{listing_id}/kpis with additional metrics: revenue_available, revenue_blocked, min_price_occurrence, occupancy_neighborhood, occupancy_neighborhood_pp, occupancy_neighborhood_ratio, occupancy_neighborhood_adjusted, occupancy_neighborhood_adjusted_pp, occupancy_neighborhood_adjusted_ratio (all forward-only, periods 7/14/21/30/60/90/180/365); comp_set_occupancy, comp_set_occupancy_adjusted, comp_set_revenue (forward-only, periods 7/30/60); pickup (backward-only, periods 7/14/30); asking_rate_lowest and asking_rate_highest (backward-only, period 365); revenue_score (non-overlapping windows 0–30, 31–60, 61–90 days); and scalar comp_set_count.
Split GET /listings/{listing_id}/kpis into four dedicated endpoints: GET /listings/{listing_id}/kpis (rolling-window metrics — bidirectional periods 7/14/21/30/60/90/180/365), GET /listings/{listing_id}/kpis/monthly (up to 15 months back incl. default dynamic-set metrics), GET /listings/{listing_id}/kpis/quarterly (up to 5 recent quarters), and GET /listings/{listing_id}/kpis/yearly (up to 5 years). Each response includes currency. The old days parameter is no longer accepted.
Added GET /notifications and POST /notifications/dismiss endpoints. Users can now list their active in-app notifications and dismiss specific ones by ID via the API.
Added optional expires_at field to PUT /listings/{listing_id}/custom_rates and PUT /listings/{listing_id}/bulk_custom_rates. Custom rates can now be set with an expiry timestamp after which they are no longer applied.
Added GET /listings/{listing_id}/fee_impact_calendar endpoint. Returns a per-date fee multiplier derived from the listing's estimated fee setting. Multiply any nightly price by the multiplier to get the fee-inclusive price.
Added GET /listings/{listing_id}/min_stay_calendar endpoint. Returns the per-date minimum stay derived from the listing's minimum stay rules for a given date range.
Added the Notes endpoints: GET /listings/{listing_id}/notes, POST /listings/{listing_id}/notes, PUT /listings/{listing_id}/notes/{note_id}, and DELETE /listings/{listing_id}/notes/{note_id} to manage the notes of a listing.
Added the Dynamic Sets endpoints: GET /sets, GET /sets/{set_id}, GET /sets/{set_id}/aggregated_metrics, GET /sets/{set_id}/listings, GET /sets/{set_id}/associated_listings, PUT /sets/{set_id}/associated_listings, DELETE /sets/{set_id}/associated_listings, GET /sets/{set_id}/changelog, GET /sets/{set_id}/time_series, and GET /sets/{set_id}/distribution. RM API users can access their paid dynamic sets.
Added GET /listings/{listing_id}/neighborhood/pricing endpoint to fetch daily neighborhood price data (median, p25, p75, listing count) for the listing's local cluster.
Added GET /listings/{listing_id}/neighborhood/occupancy endpoint to fetch daily neighborhood occupancy and booking model data for the listing's local cluster.
Added GET /listings/{listing_id}/last_posted_prices endpoint. Returns the last price posted to the channel per stay date. For available nights this is the most recently posted price; for booked nights it is the last price posted at or before the time the booking was recorded.
Added GET /market_report, GET /market_report/{market_id}/time_series, and GET /market_report/{market_id}/distribution endpoints. RM API users can access market-level data for markets in which they have listings.
All PUT endpoints now return 409 Conflict when a concurrent request for the same resource is already in progress. Retry the request after the concurrent request completes.
Added GET /listings/{listing_id}/price_calendar endpoint to fetch the price calendar for a single listing.
Initial release of the Wheelhouse Revenue Management API.
Endpoints: GET /listings, GET /listings/{listing_id}, GET /listings/{listing_id}/pricing_tier, GET /listings/{listing_id}/recent_changes, GET /listings/{listing_id}/kpis.
Endpoints: GET /listings/{listing_id}/price_recommendations, GET /listings/{listing_id}/base_price_recommendation, GET /listings/{listing_id}/checkin_checkout, GET /listings/{listing_id}/min_max_prices, GET /listings/{listing_id}/monthly_seasonality.
Endpoints: GET /preferences, PUT /preferences, GET /preferences/{listing_id}, PUT /preferences/{listing_id}, PUT /preferences/{listing_id}/copy, PUT /preferences/{listing_id}/{setting}, GET /preferences/{listing_id}/long_term_discounts, GET /preferences/{listing_id}/changelog, POST /preferences/{listing_id}/preview.
Endpoints: PUT /listings/{listing_id}/custom_rates, DELETE /listings/{listing_id}/custom_rates, PUT /listings/{listing_id}/bulk_custom_rates, DELETE /listings/{listing_id}/bulk_custom_rates.
Endpoints: GET /listings/{listing_id}/reservations (with date_filter_type parameter).
Endpoints: GET /listings/{listing_id}/tags, PUT /listings/{listing_id}/tags, GET /listings/{listing_id}/flags.
Endpoints: GET /segments, GET /segments/{segment_id}/listings.