When Hotel SEO Looks Healthy but Revenue Does Not

A hotel can have a good month on the SEO dashboard and a bad month on the revenue-management dashboard, and both reports can be honest. The dashboards measure different things: impressions and clicks describe demand at the top of the funnel, while ADR, RevPAR and net direct-revenue describe money at the bottom. The academic literature has been describing this gap for over a decade: Stangl, Inversini and Schegg’s 2016 study in the International Journal of Hospitality Management, based on 1,014 hotel questionnaires from Austria, Germany and Switzerland, reports an average portfolio of 8.06 offline and online channel categories per hotel and describes the practical consequences of high online-intermediary dependency; Masiero and Law’s 2016 paper in the Journal of Travel & Tourism Marketing on comparing reservation channels for hotel rooms takes a behavioural perspective and shows that reservation-channel choice is not a mechanical function of price alone; Ivanov and Zhechev’s 2012 critical literature review of hotel revenue management in Tourism: An International Interdisciplinary Journal catalogues the elements of the revenue-management system that any hotel dashboard should reflect; and Kimes’s 2011 paper on the future of hotel revenue management in the Journal of Revenue and Pricing Management argues for a revenue-management view that goes beyond room-only ADR. On the platform side, Google Search Console has begun to distinguish AI Overview and AI Mode impressions in the Generative AI performance report (Search) — a rolling rollout — and the GA4 BigQuery export makes raw event data available for SQL-level analysis rather than only pre-aggregated UI reports. This article is written for the GM or marketing director looking at both dashboards and asking why the two stories disagree.

How “successful SEO” and “unsuccessful revenue” both look true at once

The SEO report shows month-on-month impression growth, click growth on brand and non-brand queries, an improved average position for the property name, a rising number of indexed pages carrying valid schema, and a healthy Google Business Profile with a growing review base. The revenue-management report shows a flat or falling RevPAR, a rising OTA share of room-night bookings, a compressed ADR on flexible rates, an increase in commissionable revenue and a fall in net contribution per room-night. Both dashboards can be honest simultaneously. Both dashboards can also both be wrong at the same time — the SEO dashboard because it treats visibility as if it were revenue, and the revenue dashboard because it treats commission-net revenue as if channel mix were a given rather than a lever.

The purpose of this article is to separate the two dashboards’ natural languages, describe the specific mechanisms that make them disagree, and give a weekly reading discipline that reconciles them without demanding that the marketing team learn revenue-management theory or that the revenue team learn Google Search Console.

Short answer

Hotel SEO can look successful while revenue management shows the opposite when SEO growth is concentrated in impressions and non-brand clicks that do not convert to direct-booking room-nights, when brand-search demand is captured by OTAs rather than by the hotel’s own booking flow, when the channel mix shifts toward commissionable channels faster than direct traffic grows, or when AI Overview and AI Mode carry the answer without a click. The reconciliation is a shared weekly report that tracks SEO impressions, clicks and conversions alongside RevPAR, ADR and channel-mix percentages, sourced from the same booking system and read by the same person before either dashboard is escalated.

What “successful SEO” actually means, and why it is necessary but not sufficient

Successful hotel SEO means the pages the reader needs are indexed, ranked, structured and fast enough that the reader can find them. It means the property’s name, address and phone number are consistent across the direct site, the Google Business Profile and the major OTAs. It means the property’s cancellation, rate and inclusion copy is accurate. It means the schema.org markup describes only entities that appear on the visible page. It does not, on its own, mean that a booking will happen through the hotel’s direct-booking flow.

The reader who lands on a well-optimised hotel page can still choose to book on Booking.com, on Expedia, on Agoda, on a regional OTA, on a metasearch aggregator, or on the hotel’s direct site. Masiero and Law (2016) describe reservation-channel choice as a behavioural process influenced by more than price. That behavioural process is where SEO’s contribution to revenue is decided — and it is not a process the SEO team controls alone.

Traffic does not equal revenue: the SEO/revenue-management gap

The most common form of the SEO/revenue-management disagreement is the traffic-vs-revenue gap. A hotel’s SEO dashboard shows a 25% month-on-month rise in Search Console clicks. Its revenue dashboard shows a 3% fall in direct-booking room-nights over the same month. Both are correct. The clicks may have been coming from informational queries that never convert (destination guides, comparison content, “hotels near X” listings), or the clicks may have converted, but into OTA bookings rather than direct bookings. In both cases, the SEO team’s numbers are truthful and the revenue team’s numbers are truthful.

The disciplined weekly report describes what fraction of Search Console clicks landed on the pages that carry the direct-booking flow, what the on-page booking-flow conversion rate was for that traffic, and what fraction of the property’s total room-nights sold in the same week came from direct booking. Without those three lines, neither the SEO team nor the revenue team can tell whether SEO is helping.

Channel-mix drift under successful SEO

Stangl, Inversini and Schegg (2016) report an average of 8.06 offline and online channel categories used by hotels in Austria, Germany and Switzerland, based on 1,014 questionnaires — a portfolio that spans direct-site booking, brand.com direct call, GDS, brand-parent hotels-chain sites, wholesalers, OTAs, metasearch, tour operators, corporate contracts and more. Under successful SEO, the total number of room-nights sold can rise while the direct-site share of that total falls, because the OTA share is rising faster. A rising OTA share is not automatically bad — sometimes it is the healthy consequence of a new market being opened up — but a rising OTA share that the hotel did not plan for is a channel-mix drift the SEO dashboard cannot detect on its own.

The disciplined weekly report includes the direct-site share of room-nights sold, the OTA share (broken out at least into the top three OTAs), the metasearch share, the GDS share where applicable, and the corporate-contract share. When the direct-site share is falling week after week, the SEO team’s rising click numbers are being converted by other operators — and the revenue team’s contribution-per-room-night is falling because those channels are commissionable.

Revenue-management metrics that matter

Ivanov and Zhechev (2012) catalogue the elements of a hotel’s revenue-management system: the RM centres (room division, F&B, function rooms, spa and fitness, golf, casino, other), the data and information layer, the pricing tools (price discrimination, dynamic pricing, lowest-price guarantee), the non-pricing tools (overbookings, length-of-stay control, room-availability guarantee), the RM software, and the RM team. The KPIs that flow out of this system — occupancy, ADR, RevPAR, TrevPAR, GOPPAR and channel contribution — are the numbers a revenue manager reads first. Kimes (2011) argues that the future of hotel revenue management is a broader profitability view that looks beyond room-only ADR toward per-guest and per-room-night contribution.

The disciplined weekly report therefore separates gross revenue (which SEO can influence via demand) from net contribution (which channel mix and commission structure decide), and reports both against the SEO dashboard’s traffic and conversion numbers in the same view. When RevPAR is flat but SEO clicks are up, the report should explain where the demand went; when RevPAR falls but SEO is stable, the report should explain which channel picked up the shortfall and at what commission.

Attribution: last-click hides the truth

The default GA4 attribution model records the last non-direct channel that touched the reader before conversion. A reader who found the hotel through a Search Console non-brand keyword, spent five minutes on the direct site, closed the tab, later searched the brand name in Google, clicked a Booking.com listing and booked there — is attributed entirely to Booking.com in GA4. The SEO team’s contribution is invisible in that path, and the revenue team’s channel-mix report treats the same booking as a commissionable OTA sale. Neither dashboard is lying. Neither dashboard is telling the whole story either.

The GA4 BigQuery export makes raw event data available for SQL-level analysis so that assisted paths can be reconstructed and the SEO team’s role in the funnel can be estimated more honestly. The disciplined weekly report should describe the last-click attribution and the assisted-path picture separately, so that the SEO team is not credited with revenue it did not cause and, more importantly, is not penalised for revenue it helped cause but that another channel closed.

Brand-search leakage

A specific and expensive form of the channel-mix problem is brand-search leakage: the reader Googles the hotel’s name, sees a Booking.com or Expedia listing at the top of the results, clicks it and books there. The SEO dashboard records this as brand-search click growth; the revenue dashboard records this as a commissionable booking. The hotel has paid to build the brand recognition that generated the query, and the OTA has captured the booking. In competitive markets this leakage is the single largest hidden cost of “successful SEO”, and it is invisible on either dashboard read alone.

The disciplined weekly report includes a brand-search leakage estimate: total brand-name Search Console impressions and clicks, direct-site brand-page clicks, and direct-booking conversions on those brand-page sessions. When the brand-page click share is small relative to the brand-impression share, the property is paying for demand that is monetised elsewhere — and the fix is not more SEO, it is a defensive brand-search advertising campaign, a clearer own-site value proposition or a rate parity conversation with the OTAs.

AI Overview and the zero-click brand query

Google Search Console now reports impressions from generative-AI features. The Generative AI performance report (Search) distinguishes impressions from AI Overviews and AI Mode, groups by pages, countries, devices and dates and applies the same 1,000-row limitation as the standard Performance report. The report is rolling out to a subset of website owners. For hotels, the practical consequence is that a share of brand-search impressions may resolve inside the AI Overview without a click at all — the reader gets the room-type, phone number and check-in hours from the AI summary and moves on to book somewhere else, or does not book at this point in the funnel.

The disciplined weekly report starts to separate AI-feature impressions from classic-Search impressions. A hotel that shows healthy AI Overview impressions and falling classic clicks may be losing traffic to zero-click resolution rather than to a competitor. That is a different problem from ranking loss, and the answer is different too: better direct-site value proposition, better in-schema price and cancellation surfacing, and a defensive brand-search advertising presence for the queries where the AI Overview does the answering.

Reservation-channel behaviour is not a mechanical function of price

Masiero and Law (2016) take a behavioural perspective on hotel-room reservation-channel choice and show that the choice is influenced by trust, familiarity with the platform, perceived flexibility and information availability, not only by the headline price. A reader who trusts Booking.com’s cancellation flow will book there even when the direct site shows the same rate, because the reader is buying the platform’s guarantee alongside the room. This is why a hotel with rate parity can still see channel-mix drift — the SEO dashboard has produced the demand, the direct site has met the rate, and the reader has chosen the platform’s guarantee rather than the hotel’s cancellation copy.

The disciplined page-level response is to make the direct-site cancellation, payment and booking flow at least as trustworthy as the OTA’s — same-day confirmation, named local-time cut-off, in-currency charging clarity, direct email and phone contact — so that the reader who is going to book has a reason to book with the hotel that meets the reader’s platform-guarantee expectation. The disciplined SEO/RM report reflects the direct-site conversion rate as evidence for whether that response is working.

OTA dependency literature reads directly onto the current dashboards

Stangl, Inversini and Schegg (2016)‘s three-country study describes hotels’ dependency on online intermediaries as a portfolio decision rather than a single-channel one. A hotel that treats OTAs as an on/off switch will misread the report. A hotel that treats OTAs as one column in a channel-mix table, with a target share and a target contribution per room-night, will read the same numbers with more precision. The academic evidence is not that OTAs are bad; the evidence is that dependency compounds when it is not planned for, and that a portfolio approach — direct + one to two strategic OTAs + metasearch + corporate — is more resilient than a two-channel approach.

The disciplined channel-mix line on the weekly report includes a “planned share” and an “actual share” per channel. When actual is drifting away from planned by more than a small delta for two consecutive weeks, the report escalates — regardless of what the SEO dashboard’s clicks are doing.

Revenue-management literature perspective on SEO

Ivanov and Zhechev (2012) position revenue-management data and information at the centre of the system — it is the layer the RM tools and the RM software both draw from. SEO is not one of the revenue-management centres, but it is a demand-generation input to that data layer. The clean model is that SEO produces qualified demand at low commission cost; revenue management prices, allocates and captures that demand across channels; and the shared report reads both sides at once. When SEO is treated as a separate team with a separate dashboard and no shared table with the revenue-management team, the model breaks quietly.

The organisational fix is not always a new dashboard. Often it is a weekly meeting with a one-page shared report on the table, with SEO impressions and clicks on the top half and RevPAR, ADR and channel mix on the bottom half. When the two halves disagree, the meeting decides which lever to move — content, rate, channel mix, direct-site flow, defensive brand advertising — and next week’s report closes the loop.

The SEO team’s weekly numbers

The SEO team should report the same numbers each week, in the same order, with the same definitions:

• Search Console impressions — brand and non-brand — for the property’s country of primary demand
• Search Console clicks — brand and non-brand — for the same segment
• AI Overview and AI Mode impressions, when the property has access to the Generative AI performance report
• Landing-page traffic to the direct booking flow, from the source above
• Direct booking-flow conversion rate for that traffic
• Direct booking-flow abandonment reasons where trackable
• Same-day-booked room-nights attributed to the direct site by last-click GA4
• Any BigQuery-derived assisted-path picture, when the export is enabled

The revenue manager’s weekly numbers

The revenue manager should report the same numbers each week, in the same order, with the same definitions:

• Total room-nights sold
• Occupancy
• ADR — total and by rate tier (flexible, non-refundable, corporate, promo)
• RevPAR
• TrevPAR / GOPPAR where the property tracks them
• Channel mix: direct, brand.com direct call, top three OTAs by name, metasearch, GDS, corporate, wholesale, groups — as a share of room-nights and a share of net revenue
• Contribution per room-night by channel, net of commission
• Planned vs actual channel share by channel

Bridging the two reports into a shared weekly view

The two lists above become useful when a single page carries both and the same person reads them at the same time. The next table shows what each pattern of divergence means and where to look first.

SEO signal	Revenue signal	What it usually means	Where to look first
Clicks up, impressions up	RevPAR flat, OTA share up	Channel-mix drift — demand converted by OTA rather than direct	Direct-site conversion rate; brand-search leakage; rate parity
Clicks up, impressions up	RevPAR up, OTA share stable	SEO is contributing to revenue at the planned mix	Nothing to fix; keep publishing
Non-brand clicks up, brand clicks flat	Direct bookings flat	Non-brand traffic is informational, not booking-ready	Landing-page value proposition; content-to-conversion mapping
Brand impressions up, brand clicks down	Direct bookings down	AI Overview or zero-click resolution capturing brand answer	Generative AI performance report; direct-site brand landing page
Clicks flat, impressions flat	RevPAR down, OTA share up	Demand is stable, but reader is choosing OTA at rate parity	Direct-site cancellation and payment surface trust
Clicks down, impressions down	RevPAR down, OTA share up	SEO ranking loss AND channel-mix drift together	Ranking diagnostic; channel-mix planned vs actual

Working example — a hypothetical mid-market hotel reads both reports together

Consider a hypothetical mid-market hotel with 118 rooms, an international guest base and a mixed direct-booking and OTA channel split. It has run SEO consistently for two years and has a healthy Google Business Profile. Over the last four weeks, the SEO dashboard shows Search Console clicks up double-digits and impressions up by a similar amount. The revenue-management dashboard shows RevPAR flat, direct-share falling from a planned target down to a materially lower actual, and OTA share up. This example is illustrative; no percentages, no real property name and no real result are asserted.

Diagnostic questions the shared report answers

The GM and revenue manager ask, in order, the four questions the report is designed to answer.

Question	Signal to check	Likely lever
Are the new clicks landing on booking-ready pages?	Landing-page distribution; direct-booking-flow entry rate	Content-to-conversion mapping; internal linking
Is the direct-site booking flow converting the traffic it does receive?	Direct-booking-flow conversion rate	Cancellation copy, payment surface, mobile parity
Is brand-search demand being captured by an OTA?	Brand impression vs brand click ratio; brand landing-page share	Defensive brand advertising; own-site value proposition
Is the reader choosing OTA at rate parity because the OTA’s trust surface is stronger?	Direct-site cancellation, payment and confirmation trust signals	Direct-site booking flow discipline

What the shared report is not for

The shared report is not a ranking-guarantee. It is not a channel-share guarantee. It is not a substitute for revenue-management strategy or for a competent digital marketing plan. It is a weekly disagreement-resolution surface: when the two dashboards disagree, the report is where the disagreement is read, understood and turned into a lever the team will actually move.

Sources and evidence

Academic sources used:

• Stangl, B., Inversini, A. and Schegg, R. (2016). Hotels’ dependency on online intermediaries and their chosen distribution channel portfolios: Three country insights. International Journal of Hospitality Management, 52, 87–96. doi.org/10.1016/j.ijhm.2015.09.015.
• Masiero, L. and Law, R. (2016). Comparing reservation channels for hotel rooms: A behavioral perspective. Journal of Travel & Tourism Marketing, 33(1), 1–13. doi.org/10.1080/10548408.2014.997960.
• Ivanov, S. and Zhechev, V. (2012). Hotel Revenue Management: A Critical Literature Review. Tourism: An International Interdisciplinary Journal, 60(2), 175–197. hrcak.srce.hr/file/125946.
• Kimes, S. E. (2011). The future of hotel revenue management. Journal of Revenue and Pricing Management, 10(1), 62–72. doi.org/10.1057/rpm.2010.47.

Official and platform sources used:

• Google Search Console Help. Generative AI performance report (Search) — impressions from AI Overviews and AI Mode; page, country, device and date dimensions. support.google.com/webmasters/answer/16984139.
• Google Developers. BigQuery export for Google Analytics — raw GA4 events available for SQL analysis. developers.google.com/analytics/bigquery/overview.
• Google Business Profile Help. About the guide for hotels. support.google.com/business/answer/9177814.
• Google Search Central. Introduction to structured data markup in Google Search. developers.google.com/search/docs/appearance/structured-data/intro-structured-data.

FAQ

Does more SEO always mean more revenue?

No. SEO produces impressions and clicks; revenue is a downstream function of channel mix, direct-booking-flow conversion, rate parity and the reader’s reservation-channel behaviour. Masiero and Law (2016) show that reservation-channel choice is not a mechanical function of price alone. A hotel can have healthy SEO and falling direct revenue simultaneously — the question is where the demand went, not whether the SEO was working.

What is brand-search leakage and how do we measure it?

Brand-search leakage is the pattern where a reader Googles the hotel’s name, sees an OTA listing at the top of the result page, clicks and books there. Measure it as the ratio of brand-page direct-site clicks and direct-booking conversions to total brand-name Search Console impressions and clicks. When the ratio is small, the property is paying for demand that is monetised on another channel — and the response is typically defensive brand-search advertising and a stronger own-site value proposition, not more SEO.

What does AI Overview do to hotel search results?

Google now reports AI Overview and AI Mode impressions in the Generative AI performance report (Search). A share of brand-search impressions may resolve inside the AI Overview without a click — the reader gets the answer and moves on. That is not a ranking loss, and the fix is not more SEO. The fix is a stronger direct-site value proposition, in-schema price and cancellation surfacing that matches the AI-Overview summary, and a defensive brand-search advertising presence.

Why do our GA4 numbers make SEO look worse than it is?

GA4’s default last-click attribution assigns credit to the last non-direct channel that touched the reader before conversion, which usually favours OTAs and paid channels over organic SEO. The GA4 BigQuery export makes raw events available for SQL-level analysis so that assisted paths can be reconstructed. Read last-click attribution and assisted-path attribution side by side to see the SEO team’s real role in the funnel.

How many channels should a hotel actually run?

Stangl, Inversini and Schegg (2016) report an average portfolio of 8.06 offline and online channel categories per hotel across 1,014 questionnaires in Austria, Germany and Switzerland. The relevant question is not the number but the planned-vs-actual mix. Treat channels as a portfolio with a target share per channel and a target contribution per room-night; escalate when actual drifts from planned for two consecutive weeks.

Which single number reconciles the SEO and revenue reports?

There is no single number. The reconciliation is a weekly shared report with SEO impressions, clicks and direct-booking-flow conversions on one half and RevPAR, ADR and channel mix on the other, sourced from the same booking system and read by the same person. Ivanov and Zhechev (2012) position data and information at the centre of the revenue-management system; the reconciliation is where SEO’s contribution to that layer becomes visible.

Read hotel SEO reports and revenue management reports on the same weekly page or accept that one of them is misleading you

The two dashboards do not disagree because either team is wrong. They disagree because they measure adjacent, non-identical things: SEO measures the demand visible in search results, revenue management measures the money captured in room-nights. Reading them apart is what turns a good SEO month into a bad revenue month without warning. Reading them together — same weekly page, same person, SEO impressions and clicks and direct-booking conversions on one half, RevPAR and ADR and channel mix on the other — is what turns the disagreement into a diagnostic. When the shared report is on the table, “successful SEO” and “unsuccessful revenue” stop being independent stories and become adjacent evidence about the same funnel; the lever the team should move next becomes an editorial decision rather than a debate; and the academic literature on OTA dependency, channel behaviour and revenue-management systems finally has a place to be read out into practice.