HEDNA’s State of Distribution 2026: What It Tells Us – and What No Survey Can Yet

By /
Distribution StrategyAI & the Future of Booking·3 October 2026·12 min read
HEDNA’s State of Distribution 2026: What It Tells Us – and What No Survey Can Yet
HEDNA’s State of Distribution 2026 is the most complete benchmark hotel distribution has – more than 270 commercial professionals across 53 countries, representing over 58,000 properties. Its headline is clear: OTAs still take about a third of reservations, AI-influenced search already accounts for 4%, and commercial teams are better aligned than their technology. The more interesting question is what happens when a careful survey meets a market that is changing faster than any survey can measure.

Let me say this up front. HEDNA, with NYU’s Tisch Center and its Hospitality Innovation Hub, has given the industry something it badly lacked three years ago – a global, function-by-function view of how hotels actually distribute, report and invest. Anyone who has tried to build a commercial strategy from vendor case studies and conference anecdotes knows how rare that is. The State of Distribution 2026 report deserves to be read in full, not skimmed for a headline.

And it does more than benchmark – it opens a conversation worth having. The authors are careful to note that each edition is an independent sample, that comparisons are directional, and that hotels can see where a booking lands but not what influenced it. That candour is one of the report’s real strengths. So let’s build on it: set the findings beside other data published this year, and follow the question the report itself points towards – in a year when AI is redrawing discovery, how much certainty can any of us claim?


What does the HEDNA 2026 report actually find?

In short: hotels have built the transactional machine, but not yet the intelligence layer above it. The report finds property management systems in 97% of respondents’ stacks and booking engines in 91%, yet customer data platforms and media attribution tools sit at just 22%. Hotels are far better at processing a reservation than explaining it.

The channel picture is just as instructive. OTAs account for 34% of reservations in 2026, against 35% in the 2024 edition. Direct digital has grown from 22% to 27%, which now includes hotel websites and apps (18%), direct metasearch (5%) and AI-influenced search (4%). Independents remain the most exposed, with OTAs at 42% of their mix.

Investment is rising – 62% of respondents report bigger technology budgets, up from 38% in 2025 – but the priority is optimisation, not expansion. Making existing systems work harder ranks first; adding new ones ranks third. And despite all the talk of automation, nearly a third of teams still report entirely by hand, with distribution the least automated function of all.

Perhaps the most human finding is the one about people. Four in five hotels kept or grew their commercial teams, and revenue management teams grew fastest. Automation, the report concludes, is changing the work faster than it is reducing the need for people. Having watched the revenue role absorb reporting, channel analysis and budgeting over the years – a shift we explored in how revenue management is evolving into commercial strategy – that rings true.


Is OTA dependence really standing still?

Possibly – but it depends on the lens. HEDNA’s 34% is a picture of its respondents’ estimated mix, and on that measure the needle has barely moved since 2024. Other long-run datasets suggest a slower, steadier drift towards intermediaries that a two-year comparison may not catch.

HOTREC’s European hotel distribution study, covering 2,713 properties, puts OTA share at 29.9% in 2025, up from 19% in 2013, with Booking Holdings handling almost 69% of OTA bookings, as reported by PhocusWire. In the US, Phocuswright puts OTAs at 52% of online lodging bookings and expects them to outpace every other channel in 2026 and 2027.

These numbers are not in conflict with HEDNA’s. They measure different things – a European hotel census, a US online market sizing, a global survey of commercial leaders. But set side by side they raise a fair question. When two editions draw on different respondents, can a stable average tell us whether dependence is holding steady or quietly creeping up? HEDNA is candid that it can’t, and it says so. The rest of us should be equally careful before quoting “OTA share is flat” as settled.

What HEDNA captures brilliantly is the sentiment: 79% of distribution respondents call OTAs necessary but over-indexed. That is the industry telling itself the truth. The harder part, as we argued in our look at rate parity, is turning that awareness into a channel strategy with real economics behind it.


How much hotel demand does AI really influence?

Somewhere between under 1% and 4% – and the gap between those two numbers is the most important finding of the year. HEDNA reports that AI-influenced search accounts for 4% of reservations, rising to 6% among mid-sized chains.

Now look at the measured data. On its second-quarter 2026 earnings call, Booking Holdings said traffic from large language models made up significantly less than 1% of its room nights, according to the call transcript. Lighthouse data published in July found AI referrals at just under 1% of visits to hotel websites, though growing fast and roughly 3% relative to organic search.

Who is right? Probably both. Booking and Lighthouse measure the last referring click. HEDNA asks commercial leaders to estimate influence – the traveller who shortlisted three hotels in ChatGPT, checked reviews and then booked on the brand website. That booking shows up as direct in a PMS report. It shows up as AI in a respondent’s judgement.

And this is exactly the point HEDNA itself makes so well. The report argues that AI is an influence layer sitting above search, metasearch and OTAs, and that its commercial impact may be greater than the final 4% suggests. Don’t you think that is the real headline? We are trying to size a channel that, by its nature, leaves few fingerprints. Any figure – 1%, 4% or 10% – is an estimate wearing the clothes of a measurement. We made a similar case in what the data really says about AI and hotel bookings.


Are AI's productivity gains really ‘incremental’?

HEDNA’s respondents say yes, and they may well be right – but self-reported productivity is one of the least reliable measures we have, in either direction. Only 7.8% of hotels in the report see AI cutting manual work by more than 30%; 18.5% see no impact at all.

Two pieces of research outside hospitality are worth keeping in mind. In a randomised trial, the research group METR found that experienced software developers took 19% longer to complete tasks with AI tools – while believing afterwards that AI had made them 20% faster. And MIT’s NANDA initiative reported in 2025 that around 95% of enterprise generative AI pilots had delivered no measurable profit-and-loss return.

Neither study is about hotels, and neither proves HEDNA’s respondents wrong. What they show is how hard this is to judge from the inside. A revenue manager who drafts a commentary in ChatGPT feels faster. Whether the team is actually doing less manual work – or simply producing more reports – is a different question. HEDNA spots this risk itself, warning that saved time can be swallowed by another reconciliation or another dashboard.

So when the report says AI is incremental rather than structural, read it as “this is how it feels from the desk in 2025”. That is valuable. It is not the same as a stopwatch.


What happens when the market moves faster than the fieldwork?

The report covers December 2024 to November 2025. Since then, the ground has shifted in ways no respondent could have answered for. That is not a criticism of HEDNA – it is the unavoidable lag of any serious research – but it changes how we should use the findings.

Consider what has happened since the survey window closed. In late August 2026, Google began letting US travellers book hotels without leaving AI Mode, with Booking.com, Expedia, Marriott, Hilton, IHG, Wyndham, Choice and others on board and the hotel or platform remaining merchant of record, as Hospitality Net reported. Booking.com says its voice AI now handles the majority of eligible inbound calls. A booking-channel question asked a year ago is now being answered in a different market.

Zero-click search shows the same pattern. In HEDNA’s data, 27% of marketers saw zero-click behaviour reduce traffic to their channels, 20% saw no change – and a third were simply unsure. Outside data suggests the effect is real: Pew Research Center found Google users clicked a traditional result on 8% of visits when an AI summary appeared, against 15% when it did not. Google disputes the method. Similarweb data showed zero-click news searches rising from 56% to almost 69% in a year.

Frankly, “unsure” may be the most honest answer in the whole report. It reflects exactly the visibility gap HEDNA describes, and we explored in why direct distribution needs a new playbook. When a third of your peers can’t tell whether a structural change is hurting them, certainty is not on the menu.


How should hotel leaders read a benchmark in a year like this?

As a set of well-founded signals, not a scoreboard. HEDNA’s methodology note is admirably transparent: not every respondent answered every question, percentages are calculated on valid responses, and function and size cuts reflect subgroups rather than the full sample.

Working back from the published percentages, some of the most quotable functional findings appear to rest on modest numbers of answers – the distribution team’s view of OTAs, the visibility of unmanaged business travel, the marketing response to zero-click. That is normal for specialist survey paths and takes nothing away from the report. It simply means the direction of travel deserves more weight than the decimal point.

I remember the same discipline from my years in commercial leadership and hotel distribution at Wyndham and Starwood: a channel report was only as good as the definitions behind it, and the most useful conversations started with “what are we actually measuring?” HEDNA’s own conclusion makes the same point – that shared definitions, identifiers and ownership matter more than another tool. When its revenue managers and its distribution specialists see business travel so differently, that is not noise. It is the finding.

So the question for every commercial leader is not “is HEDNA right?” It is “which of these signals would I bet on, and how would I know within six months if I were wrong?”


The commercial take

Use HEDNA as your agenda, not your answer. Take its four big signals – OTA over-indexing, the attribution gap, manual reporting and governed AI – and put each on your next commercial meeting with an owner.

Build your own baseline before you need it. Tag AI referrals in analytics now, add a “how did you hear about us” field to the booking path, and track branded search and direct share monthly. In two years, the hotels that can answer “what did AI do to our mix?” will be the ones that started measuring today.

Align definitions across functions. If revenue management and distribution disagree on what counts as business travel or direct revenue, fix that before you buy anything. It is the cheapest integration project you will ever run.

Treat agentic booking as a channel decision, not a tech experiment. Google’s AI Mode booking is already live in the US. Decide your content, rate and connectivity position deliberately – rather than finding out from a parity report.

Hold your certainty lightly. The best benchmark in the industry says the picture is not linear. That is a reason to run small, measurable tests every quarter – not to wait for the 2027 edition.


Frequently asked questions

What is the HEDNA State of Distribution report?

It is an annual global benchmark of hotel commercial practice from HEDNA (the Hospitality Electronic Distribution Network Association) and NYU SPS Jonathan M. Tisch Center of Hospitality’s Hospitality Innovation Hub, powered by RateGain. The 2026 edition draws on more than 270 professionals in 53 countries, covering booking mix, technology, reporting, AI, teams and governance.

What share of hotel bookings come from OTAs in 2026?

HEDNA’s 2026 report puts OTAs at 34% of reservations across its respondents, almost unchanged from 35% in 2024, and 42% for independent hotels. Other sources measure differently: HOTREC finds 29.9% in Europe in 2025, while Phocuswright puts OTAs at 52% of US online lodging bookings.

How many hotel bookings come from AI search?

HEDNA respondents estimate AI-influenced search at 4% of reservations. Measured referral data is lower: Booking Holdings reports LLM traffic at significantly below 1% of room nights, and Lighthouse finds AI referrals just under 1% of hotel website visits. The difference reflects influence versus last-click measurement.

Is AI reducing manual work for hotel commercial teams?

Modestly, according to HEDNA. Only 7.8% of respondents report AI cutting manual work by more than 30%, and 18.5% see no impact. Most use is in content, reporting and communication. Because these are self-reported estimates, hotels should measure time saved directly before drawing firm conclusions.

HEDNA has done the industry a real service by asking the questions most of us only debate over coffee. The best way to honour that work is to keep asking them – of our own data, every quarter. Are you ready to measure what your next booking was influenced by, rather than just where it landed?


Sources

Scroll to Top