The hotel website that converts: what actually loses a booking

A hotel spends three months and a serious budget rebuilding its website: a full-width photograph of the façade at golden hour, cleaner typography, quicker transitions. Three months after launch, direct bookings haven't moved and the OTA share of the calendar sits exactly where it did before the invoice arrived. Nothing on the new site is broken; it simply never answers the questions a guest has in the minute before they pay or close the tab, and a beautiful site that skips them loses the booking as reliably as the plain one it replaced.

TL;DR: A hotel website converts when it removes a guest's reasons to hesitate, not when it looks better. The room rate needs to sit in view from the first screen, not behind a date picker, so it can be compared honestly against the platforms (rate parity: keeping the same room at the same price across every channel it sells on). The photography needs to show the actual room a guest can book, not the building around it. The mobile flow has to work with one thumb, the cancellation terms need to be visible before the guest reaches the booking engine (the third party checkout that takes payment and holds the room), and the site needs to be fast enough not to lose any of it first. This article covers each of those in turn, the redesign that hides the price, and when a redesign is the wrong fix.

This is written for the owner or general manager of an independent hotel of roughly 50 to 200 rooms weighing up a redesign, or wondering why the current one underperforms: a diagnostic to run before the design brief gets written, not the brief itself.

What "converts" means on a hotel website

Conversion, on a hotel's own site, is one thing: a completed booking through its own booking engine, at a rate the hotel would rather take than the commission-bearing version on an OTA. It is not sessions, time on page or the bounce rate a redesign proposal likes to quote as success. A site that loads fast and looks sharp on every device can still send a visitor back to a platform tab if it fails to answer what they came to ask.

In the accounts we run, the sites that convert well are rarely the most visually ambitious; they're the ones where the decisions below got treated as core, not administrative detail bolted on after design sign-off. One pattern from the hotel work specifically: keeping the property's reassurances and reviews visible at every step of the journey, from the ad through to the booking engine, rather than parked on a page of their own, did more for completed bookings than any single design choice. If the wider problem is direct bookings generally, our guide to increasing direct bookings covers the full list; this stays narrowly on what the site has to get right.

Rate parity, visible before the guest goes looking for it

Most hotel sites make a guest work for the price: pick two dates, wait for a spinner, sometimes create an account, before a number appears. A guest who has to do that opens a second tab to check Booking.com or Expedia while they wait, purely to compare, and that tab is where a lot of direct bookings quietly go to die, even when the hotel's own rate matches or beats the platform's.

The fix is not complicated. Show an indicative rate, or a real one for the dates already in the URL, on the first screen, before the guest has done anything except arrive. If your price matches the OTA's, that first screen is doing marketing work no banner copy can do alone.

Whether a parity clause obliges you to keep that price identical across every channel varies by market and by the agreement each OTA has you sign; courts in different European countries have not landed on the same answer. That is a matter for whoever manages your commercial agreements, not a website decision.

Photography that answers the question in the guest's head

A guest browsing a hotel site isn't deciding whether the building is attractive; they already clicked through because something appealed to them. What they're trying to work out, in the seconds before they book, is what room they'll get for the money, and a hero photograph of the lobby or the pool at sunset doesn't answer that.

The room type being booked needs its own accurate, well lit photography, taken from roughly where a guest stands on walking in. A single aspirational suite photographed for every room category, while the guest books the cheapest room elsewhere in the building, sets up a gap between what was shown and what arrives at check-in. That gap doesn't just cost the booking; it shows up again at the front desk as a complaint, then again as a review, a far more expensive place to lose the same argument.

If photography budget is limited, the room categories on sale are worth more than the public spaces every OTA listing already shows.

The mobile flow, where most of the deciding happens

On the hotel accounts we run, most booking engine sessions now start on a phone, often mid research, in bed or on a train, not at a desk. A flow built and tested on desktop assumes things a phone doesn't have: room to compare several rate plans side by side, a mouse precise enough for a small calendar cell, the patience to scroll sideways for the total price.

None of that patience exists on a phone at eleven at night. A date picker needing pinch zooming, a wrong keyboard on a form field, or a payment step that reloads and loses the selected dates all cost the booking the same way: the guest closes the tab and finishes on whichever app remembers their card details and login, often the platform, not the hotel. Walking the flow yourself on mobile data finds problems an automated audit tool typically misses.

Cancellation terms: the reassurance that decides high-consideration bookings

A weekend booked six weeks out and paid in full is a different decision to a lunch reservation, and gets treated with more caution. Before paying, that guest wants to know what happens if plans change, and on most OTA listings that answer sits right next to the price, at every step, not buried in a separate page.

On plenty of hotel websites, the same information exists only as a link to a terms and conditions page at the bottom of the site, written for a lawyer rather than a guest mid-booking. A guest who can't find the cancellation policy in the few seconds they'll spend looking assumes the worst, books nothing, and returns to the platform where the answer was already visible.

Surfacing the actual policy in plain terms, next to the room and the price rather than three clicks away, is one of the cheapest changes available on most hotel sites, and one of the most consistently skipped, because it reads as administrative detail rather than the copy that decides high value bookings.

Speed, and what "fast enough" actually means

Site speed gets treated as a technical concern, but the evidence puts it in the same conversation as everything above. A Google-commissioned study run with Deloitte, testing real mobile sites across retail and travel brands, found that shaving 0.1 seconds off load time lifted travel booking rates by around 10% and increased checkout completion by 2.2%. That isn't a slow site being fixed; it's an already fast site getting a fraction of a second faster.

For a hotel site, "fast enough" means the first screen, carrying the price, loads before the guest finishes the thought that brought them there. A four second hero video autoplaying above the rate isn't a design flourish, it's friction with a soundtrack. Everything after matters less, because most guests who bounce never get that far.

The booking engine handoff, where sites genuinely leak

Whichever category of booking engine a hotel uses, the moment a guest clicks "book" and leaves the hotel's own pages is where a lot of otherwise good site work gets undone. Checkout research across ecommerce, aggregated by the Baymard Institute across fifty studies, puts average documented cart abandonment at just over 70%. A booking engine handoff is functionally the same step: the point an interested visitor has to commit, where friction costs the most because there's the most to lose.

The way it usually goes wrong on hotel sites is a mismatch: the branding, currency or room description doesn't quite match what appears the instant the guest arrives in the booking engine, and a guest who notices pauses to wonder if they've landed somewhere legitimate. Extra fields never asked for on the site, a differently displayed rate, or a growing step count all read to the guest as the same signal: this is not the site they were just looking at.

Getting the site right and getting this handoff right are two separate jobs. The second, choosing and evaluating the booking engine itself, is covered in how to choose a booking engine; worth knowing before a redesign brief assumes one automatically fixes the other.

The failure mode: a redesign that photographs the building and hides the price

The clearest version of everything above is a redesign commissioned to fix the brand that ends up making the booking mechanics worse, because nobody owned them in the brief. The building gets photographed beautifully. The rate sits behind an extra click it didn't before. The cancellation policy, previously a visible line on the old site, becomes a PDF linked from the footer. This is not deliberate. It is what happens when "redesign" gets scoped as a look-and-feel project and the booking mechanics are assumed to carry over unchanged.

To size what's at stake: a hotel getting 10,000 monthly visits to its booking pages at a 1.5% conversion rate takes 150 direct bookings a month. Lifting that to 2%, purely by fixing the basics above, produces 200 bookings from the same traffic: 50 more a month for no extra spend. That's illustrative arithmetic, not a promised outcome, but it's why fixing the mechanics tends to pay back faster than getting more people to see a site that loses them at the price, the phone or the handoff.

When a redesign is the wrong fix

None of this argues for redesigning every hotel site. If the site already converts well and traffic simply isn't there, a new website changes very little, because there's nobody arriving to convert. That's a discovery problem, not a conversion one, and needs the work covered in our hotel SEO guide and our hotel PPC guide, not a new template.

The way to tell the two apart before committing budget: look at sessions to the booking pages against completed bookings, separately. Low sessions with a fine conversion rate is a traffic problem; reasonable sessions with a low conversion rate is what everything above is about. Redesigning because occupancy is soft, without checking which it is, is one of the easiest ways to spend a redesign budget on the wrong problem.

The quickest way to know which problem you have is to look at the real numbers before a redesign gets briefed. If you want a second pair of eyes on where your own site is losing guests, get a free audit for a straight answer on whether the issue is traffic, conversion, or something upstream of both.