Skip to main content

Review Velocity and Its Effect on OTA Ranking

Why fresh reviews outrank better-rated hotels with stale ones

Reviewshotel review velocityAnya CortezReviewed Apr 22, 2026

Review Velocity and Its Effect on OTA Ranking

Sources: Booking.com January 2025 review-scoring update (per third-party hospitality coverage), Booking.com Partner Hub officially-named ranking factors + Traveller Review Awards 2026 FAQ (verified April 2026 via Playwright), Cornell Center for Hospitality Research (Anderson 2012 + Anderson & Han response-engagement work), peer-reviewed signaling-and-reputation paper on TripAdvisor reviews (ScienceDirect), Reviewpro Guest Review Index methodology with Cornell revenue coefficients, Shiji Q3 2025 Guest Experience Benchmark (global response-time and response-rate data across 150+ countries), Airbnb October 2025 Professional Host Summit ranking disclosure, and Booking.com Traveller Review Awards 2026 FAQ. Last reviewed: 2026-04-22.

Key takeaways

In January 2025 Booking.com changed how it weights reviews in the Guest Review Score. The old model was a three-year average; the new one is recency-weighted, with recent reviews counting substantially more 1. Third-party coverage (Mara Solutions, Hospitality Net, Shiji Insights, Revenue Hub) describes the tiers as: last 3 months highest, last 12 months high, 12-24 months moderate, 24-36 months minimal, beyond 36 months excluded. Booking.com has not published a primary specification for those tiers; they are reconstructed from third-party coverage and Reviewpro GRI methodology alignment, not from a Booking changelog 12. A property with 50 fresh reviews at 8.5 can score higher than a property with 200 reviews averaging 9.0 spread over three years. Exact weight coefficients are not public.

This matters because Booking names Guest Review Score as one of its five officially-confirmed ranking factors 3. Recency is now baked into how that factor is calculated. Velocity (the rate at which you accumulate new reviews) directly affects how quickly your score reflects current operations rather than history.

Why it moves bookings

Reviews drive bookings, ranking, and pricing power. The Cornell Hospitality Quarterly literature has documented this link for over a decade. Anderson 2012 found that a 1-point increase in TripAdvisor review score (on a 5-point scale) supports an 11.2 percent price increase at constant occupancy 4. Booking explicitly confirms Guest Review Score as one of its five officially-named ranking signals 3. Higher score lifts your sort position; better position drives more impressions; more impressions drive more bookings.

Until January 2025, Booking calculated that score as roughly a three-year average. The result: a hotel that improved meaningfully in 2024 saw little movement, because two years of older reviews dominated the math. Properties got stuck at scores that no longer reflected operational reality. The 2025 update changed that.

The recency weight distribution as described in third-party coverage 1:

  • Reviews from the last 3 months: highest weight
  • Reviews from the last 12 months: high weight (recent feedback significantly influences the score)
  • Reviews from 12 to 24 months: moderate weight
  • Reviews from 24 to 36 months: minimal weight
  • Reviews older than 36 months: not included in the score

These tiers are not in a primary Booking.com publication; they are how Mara Solutions, Hospitality Net, and Shiji Insights summarize the change. The mechanism's design matches the Reviewpro Guest Review Index, which the broader hospitality industry has used for years to provide a recency-weighted alternative to simple-average review scores 2. Treat the tier numbers as directional rather than precise.

A hotel accumulating 30 fresh reviews per quarter will see current quality reflected in its Booking score within 90 days. At 5 reviews per quarter, improvements lag by months.

Booking's own 2025 Traveller Review Awards press release reinforces the recency emphasis at the corporate-PR level: the awards recognized 1.71 million partners based on reviews received during the prior year, not lifetime 5.

The Traveller Review Awards 2026 FAQ (verified 2026-04-22 via Playwright) exposes a non-obvious asymmetry a practitioner should plan around: the live Guest Review Score is recency-weighted, but the Award-eligibility score is a flat average of reviews published 1 December 2022 to 30 November 2025, with an 8-10 minimum on at least three reviews 6. A property managing to both metrics optimizes differently for each. Fresh reviews drive the score that drives ranking; a consistent 3-year average drives the badge that drives the marketing asset. A property that had a bad 2023 can still clear the Award threshold if the 3-year average smooths out the trough, even while its live score (and ranking) reflects a different, recency-dominated reality.

2025 industry benchmarks: what response volume and speed now look like

Shiji's Q3 2025 Guest Experience Benchmark is the freshest large-sample evidence in the space 7. Global Review Index held at 87.0 percent in September 2025. Median response time fell to 3.0 days globally (down from 4.7 in Q3 2023): 5-star at 2.8 days, by region from Africa at 2.2 days to Oceania at 5.0. Global review volume grew 2.6 percent year over year; Google reviews alone grew 33 percent, a platform-shift signal inside the aggregate.

The counter-trend is instructive. Response rate fell across every star category: 5-star -1.8pp, 4-star -3.8pp, 3-star -4.1pp year over year 7. Hotels are responding faster but to fewer reviews. Revinate 2025 puts global response rate near 70 percent, APAC at 61 percent 8. Under the Cornell 40-percent inflection, that 70 percent mean sits above the diminishing-returns line; the industry is defaulting to a cadence Cornell's regression predicts will hurt scores. The recency-weighted system amplifies both the upside (fast responses compound faster) and the downside (over-responding to positives erodes score inside the 3-month high-weight window).

The revenue math behind the score

Anderson 2012 gives the headline (11.2 percent ADR lift per 1-point TripAdvisor score improvement) 4, but the ReviewPro Guest Review Index, which aligns with Booking's 2025 recency-weighted methodology, carries more granular coefficients. Cornell / ReviewPro work measured a 1-point GRI increase as 0.89 percent ADR, 0.54 percent occupancy, and 1.42 percent RevPAR lift 9. A hotel that raises its trailing-90-day GRI by 3 points with sustained velocity is working toward a ~4.3 percent RevPAR outcome.

Airbnb's October 2025 shift: predicted review outcome as a ranking input

At its October 2025 Professional Host Summit, Airbnb disclosed that listings are ranked on two primary signals inside an 800-plus-signal model: likelihood that a specific guest books, and likelihood that booking leads to a 5-star review 10. Communication responsiveness and active-host "vitality" feed the 5-star calculation. This is the first OTA to name a predicted review outcome as a ranking input; Booking and Expedia rank on observed score, Airbnb ranks on an estimate of the review you haven't yet received.

What "great" looks like

A property whose review-velocity profile aligns with the Booking 2025 system has three traits:

  1. Sustained monthly review intake. Not bursts. The recency-weighted score gives the most weight to the last 3 months. A property that gets 0 reviews in December and 30 in January looks worse on the trailing 3-month window than a property getting 10 per month consistently.
  2. A high response rate on negative reviews specifically. Cornell research established that responding to all negatives lifts review score by 1.65 percent at the property level 11. With recency weighting, that lift now compounds faster (since old reviews matter less, the response-driven score gain shows up quicker). See Responding to Negative Reviews.
  3. Active prompts at checkout to drive review velocity. Properties that prompt guests in the first 24 to 48 hours after checkout get the highest review-completion rates per industry benchmarks. The hotel-controlled lever is the prompting moment, not the review itself.

The 90-day score-recovery window

Under the old three-year-average model, score gains crept at ~0.1-0.2 per year. Under the recency-weighted model, the same operational improvements can show up within 90 days; the last-3-months bucket is heavily weighted and turns over quickly 1. The recovery window now matches a quarterly business review, not a multi-year plan.

Shiji's Jan 2026 case write-up shows the magnitude concretely. One tracked property moved from 8.8 to 9.3 on Booking inside 8 months of 2025 and, in Shiji's own words, recorded "15 score changes, more than the total for the previous two years combined" 7. Pre-2025, scores moved by increments of 0.1 once or twice a year. Post-2025, the score is a live signal that can shift monthly as the 3-month bucket turns over, up or down.

What happens to properties with stale review portfolios

Under the 2025 system, coasting on years-old reviews is penalized; recent operational wins show up faster. A Q1 2026 Barcelona sweep makes the mechanism concrete: OTALift tracked a 40-room independent (8.7 overall on ~1,100 lifetime reviews, 8 reviews in the trailing 90 days) that held position 12 pre-update. Within 90 days of January 2025 it dropped to position 31. Lifetime score barely moved. The recency-weighted component did 12.

Common failure modes

Coasting on a 5-year-old review average. Pre-2025, this worked: the three-year average smoothed out current performance. Post-2025, reviews older than 36 months are not in the score, and the most recent quarter dominates. A hotel that hasn't actively driven new reviews since 2023 is now operating on a score that ignores its best historical performance.

Letting review intake stall during shoulder seasons. A 60-day stretch with zero new reviews creates a stale window in the recency-weighted calculation. The hotel can't influence what guests choose to review, but it can influence the prompting cadence. Under-prompting during low season means the score depends on a small sample, which is more volatile.

Treating burst reviews (after a renovation or PR moment) as a permanent gain. A property gets featured, picks up 50 reviews in a month, then goes back to baseline volume. The 50 fresh reviews dominate the 3-month window briefly, then age out within 24 to 36 months. The score gain is temporary unless the velocity is sustained.

Triggering Booking's velocity fraud detection. Booking actively suppresses unnatural review bursts. Properties that suddenly receive 10x their baseline review volume in a short window can have those reviews flagged or suppressed. The line between "healthy velocity push" and "unnatural burst" is not publicly documented, but as a directional rule: keep month-over-month variation within ~2x your trailing baseline. A property going from 10 reviews per month to 80 in a single month will draw scrutiny. A property going from 10 to 18 will not.

Not responding to recent reviews specifically. Cornell's 1.65 percent score lift from responding to all negatives is now front-loaded under the new system. The lift compounds faster because old unresponded-to reviews matter less, but only on the condition that the hotelier is actually responding to the most recent ones. See Responding to Negative Reviews for the response cadence rule.

Confusing review volume with review velocity. A property with 1,000 reviews accumulated over 5 years and a property with 200 reviews accumulated in 6 months have very different velocity profiles. The 200-review property scores higher on the recency-weighted system because more of its reviews fall into the high-weight buckets.

Step-by-step

The velocity workflow

  1. Pull your review-volume trend by quarter. From Booking Extranet (Reviews tab) and Expedia Partner Central. Plot the last 8 quarters: how many reviews per quarter? Is it trending up or down?
  2. Compute your trailing 3-month and 12-month review counts. These map directly to Booking's high-weight buckets. If your 3-month count is below your 12-month run-rate, your score is at risk.
  3. Identify the prompting gap. What percent of your stayed guests leave a review? Industry baselines vary, but if your conversion is under 10 percent, the prompting cadence is the lever, not the guest behavior.
  4. Implement a 24-hour and 7-day post-checkout prompt. Email or SMS, both within the platform's allowed marketing scope. The 24-hour prompt captures memory-fresh reviews; the 7-day captures the considered ones.
  5. Pair the velocity push with a response push. Cornell's response-rate findings still apply 11; under the recency-weighted system the response-driven score lift just compounds faster. Respond to 100 percent of negatives, sample 30 to 40 percent of positives.
  6. Stay under Booking's velocity-fraud threshold. Do not 10x your monthly review volume in a single push. Booking rewards smooth increases and suppresses bursts.
  7. Monitor the trailing 3-month score weekly. Under the 2025 system, the 3-month bucket is the sensitive one. If it drops, you have a current-quarter operational issue, not a historical drag.
  8. Re-check the trend quarterly. Velocity is a sustained-effort metric, not a sprint. If you see consistent improvement over 2 to 3 quarters, the system is working.

Practical velocity targets

For a 50-room independent at typical occupancy, sustainable velocity targets:

  • Quarterly intake: 20-50 new Booking.com reviews (benchmark against your trailing 12-month run-rate)
  • Response rate: 30-50 percent overall, 100 percent on negatives
  • Time-to-first-response: under 48 hours

Scale with stayed-guest volume for larger properties.

Self-audit checklist

  • I know my trailing 3-month review count on Booking.com
  • I know my trailing 3-month review count on Expedia
  • My trailing 3-month count is substantially above the floor (at least 50% of my trailing 12-month run-rate; well above that is the goal)
  • My oldest reviews influencing the Booking score are within the last 36 months
  • I have an automated post-checkout review prompt at 24 hours
  • I have a follow-up review prompt at 7 days
  • My response rate on negative reviews in the last 90 days is at or near 100 percent
  • My overall response rate sits between 30 and 50 percent
  • My median time-to-first-response is under 48 hours
  • I monitor the trailing 3-month review score weekly, not just the all-time average
  • My month-over-month review-volume variation stays within ~2x baseline (avoids fraud-detection scrutiny)

How OTALift surfaces this

Two distinct surfaces share this article's slug. ReviewVelocityValidator runs inside the quarterly review report (review-quarterly). It measures the recency-tier distribution and per-month intake rate aligned to Booking's 2025 model. ReviewFreshnessRule runs inside Property Health. It measures the age of your latest review on each OTA and surfaces per-platform "Sync reviews" recommendations when freshness slips. Velocity and freshness are related but distinct: a property can have fresh reviews on every OTA (latest-review age within 14 days) and still have stalled velocity (only 1-2 new reviews per month). Both are tracked here.

In addition to those, EngagementValidator tracks review response rate, negative response rate, and time-to-first-response (anchored in Cornell's response-rate inflection point and the 48-hour response window). It cites this article alongside responding-to-negative-reviews and reviews-plus-answers-combined-effect.

The velocity validator emits four action cards driven by the research in this article:

  • Review intake below healthy pace. A MEDIUM-priority card when trailing 3-month intake is under 5 reviews/month, with HEALTHY_MONTHLY_INTAKE set to 10/month.
  • Score stuck on stale data. A HIGH-priority card when trailing 3-month intake is slow AND the 24-36 month tier carries >=15% of the corpus (the "any operational improvement you've made in the last year is being diluted" diagnosis).
  • Review velocity slowing. A MEDIUM-priority card when trailing 3-month rate sits at or below 50% of the trailing 12-month run-rate (i.e. substantially below run-rate, not just below). The 50% boundary is the operator-defensible noise floor; normal posting irregularity routinely pushes the ratio into the 0.5-1.5x band without signaling anything operational.
  • Review velocity spiking, risk of Booking fraud flag. A HIGH-priority card when trailing 3-month rate is at least 2x the trailing 12-month rate AND there are at least 6 reviews in the under-12-month window. The 2x guard matches the directional rule in Common failure modes paragraph 4: a property going from ~10/month to 20+/month draws scrutiny, even when the volume is genuine.

These are the implemented surfaces today; further validator extensions surfaced by this article's research are tracked in research/report-improvements.md alongside the article.

Related articles

Sources and methodology

Authored by Anya Cortez · Reviewed by Anya Cortez · Last reviewed: 2026-04-22

Anya Cortez is OTALift's hospitality researcher and writes The Labs.

Footnotes

  1. Booking.com January 2025 review-score update, recency-weighted calculation aligned with Reviewpro's Guest Review Index. Coverage and analysis: Mara Solutions ("Booking.com Review Score Update 2025: What Hotels Need to Know"), Hospitality Net ("Booking.com's new review scoring explained"), Shiji Insights ("Booking.com's new review scoring: What it means for hotels"), Revenue Hub ("What is the Booking.com 2025 Review Score Update?"). Booking.com's primary announcement is not publicly linkable; tier structure is reconstructed from this third-party coverage. https://www.mara-solutions.com/post/booking-review-score-update-2025 and https://insights.shijigroup.com/booking-coms-new-review-scoring-explained-what-hoteliers-need-to-know-in-2026/ ↩ ↩2 ↩3 ↩4

  2. Reviewpro Guest Review Index (GRI) methodology. Industry-standard recency-weighted review-scoring algorithm that Booking.com aligned with in its January 2025 update. Background: Shiji Insights coverage cited in 1. ↩ ↩2

  3. Booking.com Partner Hub, "Search results, ranking, and visibility." Guest Review Score is one of Booking's five officially-confirmed ranking factors. Verified 2026-04-18 via Playwright. https://partner.booking.com/en-us/help/growing-your-business/analytics-reports/search-results-ranking-and-visibility ↩ ↩2

  4. Anderson, C. K. (2012). Search, OTAs, and Online Booking: An Expanded Analysis of the Billboard Effect. Cornell Hospitality Report Vol. 11 No. 9. Source for the 11.2% price-supporting lift per 1-point review score improvement. Travel Weekly summary: https://www.travelweekly.com/Travel-News/Hotel-News/Cornell-study-links-hotel-reviews-and-room-revenue. (Distinct from the later Anderson & Han "Engaged Consumers" paper which provides the 1.65% / 40% findings cited elsewhere in The Labs.) ↩ ↩2

  5. Booking.com news, "Booking.com's Traveller Review Awards 2025 Recognize a Record 1.71 Million Partners." Booking's own press release, reinforcing the prior-year recency frame. https://news.booking.com/bookingcoms-traveller-review-awards-2025-recognize-a-record-171-million-partners-and-unveil-the-years-most-welcoming-destinations/ ↩

  6. Booking.com Partner Hub, Traveller Review Awards 2026: FAQs. Page banner "Updated 1 month ago" as of Playwright verification 2026-04-22. Direct quote: "To win a Traveller Review Award, on 1 December 2025 you need to have had: An average review score of between 8 and 10... At least three reviews left by Booking.com travellers." Award score defined as "the average of all your guest reviews published on our website and app between 1 December 2022 and 30 November 2025." https://partner.booking.com/en-gb/help/guest-reviews/award/traveller-review-awards-2026-faqs ↩

  7. Shiji Q3 2025 Guest Experience Benchmark, Defying the Peak: How Global Hotels Sustained Guest Satisfaction in Q3 2025 (October 2025). Sample: "millions of verified guest reviews" across 150+ countries. Global Review Index 87.0 percent (September 2025). Response time 3.0 days globally (down from 4.7 in Q3 2023); 5-star 2.8 days. Response-rate declines: 5-star -1.8pp, 4-star -3.8pp, 3-star -4.1pp YoY. Global review volume +2.6 percent YoY; Google review volume +33 percent YoY. Shiji Insights also published a Jan 2026 Booking-scoring case study (8.8 to 9.3 in 8 months, 15 score changes in a single year). Reports: https://hospitalitynet.org/news/4129390.html and https://insights.shijigroup.com/booking-coms-new-review-scoring-explained-what-hoteliers-need-to-know-in-2026/ ↩ ↩2 ↩3

  8. Revinate, 2025 Hospitality Benchmark Report (April 2025 press release). Dataset: 2.4 billion emails, 23 million texts, 24 million reviews, 5.9 million calls across 2024. Direct quote: global management response rate "climbed to nearly 70 percent, with average reply times dropping to just over three days in early 2025." APAC response rate trailed at 61 percent. https://www.revinate.com/press-releases/2025-hospitality-benchmark-report-release/ ↩

  9. Cornell Center for Hospitality Research / ReviewPro Global Review Index coefficients, as summarized by Hospitality Net. A 1-point increase in the 100-point GRI: 0.89 percent ADR lift, 0.54 percent occupancy lift, 1.42 percent RevPAR lift. The GRI is the recency-weighted review-scoring methodology Booking.com aligned with in its January 2025 update. https://www.hospitalitynet.org/news/4058547.html ↩

  10. Rental Scale-Up coverage of Airbnb's October 2025 Professional Host Summit. Direct quote: "listings are ranked based on their likelihood of being booked by the specific guest and leading to a 5-star review." Model evaluates 800-plus signals including communication responsiveness and active-host "vitality." First public statement from a major OTA naming predicted review outcome as a ranking input. https://www.rentalscaleup.com/how-to-rank-higher-on-airbnb-booking-probability-and-guest-satisfaction-now-drive-visibility/ ↩

  11. Anderson, C. K., and Han, S. Hotel Performance Impact of Socially Engaging with Consumers. Cornell Center for Hospitality Research. ~10,000 quarterly hotel observations across NYC and Orlando. Source for the 1.65 percent score lift from responding to all negatives and the 40-percent response-rate inflection. Cited in detail in the Responding to Negative Reviews sources block. https://sha.cornell.edu/wp-content/uploads/sites/4/2019/03/anderson-engaged-consumers.pdf ↩ ↩2

  12. OTALift field observation, Barcelona Playwright sweep, Q1 2026. 40-room independent (identity withheld pending publication consent) tracked across the January 2025 scoring change: 8.7 overall score on ~1,100 lifetime reviews, 8 reviews in the trailing 90 days, popularity-sort position 12 pre-update, position 31 within 90 days post-update. Same methodology used across the Barcelona cohort surfaces in sibling articles in The Labs. ↩

Related articles

Want OTALift to apply this to your property?

Every recommendation in our reports links back to one of these articles.

Book audit