Why was the result measured year-over-year instead of before/after?

Hotel bookings are strongly seasonal. Comparing the month before the change with the month after would mostly measure the season, not the fix. Comparing the same period year-over-year filters the seasonality out.

Does instant online booking suit every hotel?

Most, yes — customers used to e-commerce expect immediate confirmation. Exceptions are properties with highly individual arrangements (weddings, full-property rentals), where an inquiry makes sense as a complement, not the only path.

Isn't booking without a card a no-show risk?

It depends on how often no-shows actually happen. At Hotel Silva they were nearly nonexistent, so the mandatory card protected against a rare problem while deterring customers daily. Where no-shows are real, softer options exist — e.g. a card only for non-refundable rates.

What is a direct booking and why does it matter?

A booking made on the hotel's own website rather than through an OTA portal. The hotel pays no commission on it, so the margin per stay is significantly higher — and the hotel owns the customer relationship.

Case study: how Hotel Silva grew direct bookings by 23 %

Hotel Silva had decent website traffic. People arrived, browsed the rooms, spent time on the pages. Direct bookings were nonetheless minimal — and every booking that happened on a portal instead meant commission paid and a customer relationship lost.

The data was unambiguous: analytics showed a high exit rate on the page with the contact form. Direct bookings were a fraction of what the traffic suggested.

A conversion audit found the common denominator: the website didn’t let customers book now. And a customer who can’t take the step now usually takes no step at all.

How we went about it

The Silva case is a textbook example of data-driven conversion work, which is why we break it down in detail. Four steps:

1. Data before opinions. Before looking at the site “with a designer’s eye”, we went through the analytics: where visitors come from, where they go, where the journey ends. The high exit rate on the form page was the strongest signal — that’s where the most people who had already shown interest were being lost.

2. Walking the customer’s path. We went through the booking journey the way a guest experiences it: from picking a date to submitting the request. Not as a design check, but as a test of a single question — what does a customer have to overcome for the hotel to get a booking? Every step that takes effort or creates uncertainty is a candidate barrier.

3. Hypotheses prioritized by impact. An audit always finds more than you should fix at once — fix everything simultaneously and you won’t know what worked. We ranked the findings by how many customers pass through each spot and how strong the barrier is. Two clearly led: the inquiry form standing in for booking, and the mandatory credit card.

4. Change and measurement. Both barriers were removed, and the result was measured in a way that filters out seasonality — more below.

What was blocking direct bookings

1. An inquiry form instead of a booking

The site had a classic inquiry form: name, email, dates, message. You hit Send and wait for a reply. No confirmation, no certainty — just a non-binding inquiry with no guarantee the dates are free.

A customer at the moment of decision doesn’t want to wait. They want confirmation now — the same as any e-shop or online service gives them. Every extra step and every second of waiting is an opportunity to reconsider, close the tab, “do it later”. Except “later” rarely comes in hotel bookings: the guest finds another hotel in the meantime, or completes the booking on a portal that confirms instantly. The hotel then pays commission for a customer it already had on its own website.

Behavioral data confirmed it: the form’s completion rate was very low, and most visitors left the page within 30 seconds without filling in anything. That detail matters — they weren’t leaving after deliberation, they were leaving immediately. At first glance, the page failed to convince them a booking could actually be completed there.

2. A mandatory credit card

The site required adding a payment card as a condition of booking. The reason: no-show protection. In reality, the hotel sees a no-show about once a year — minimal risk, maximal friction.

The customer didn’t know the reason. They just saw: “Enter your card number.” That’s precisely where a decision process stalls — an unexpected demand, no explanation, and instead of continuing, the customer breaks off.

There’s a general principle worth pausing on: rules that protect the business carry a hidden cost on the customer’s side. A mandatory card, strict cancellation terms, extra verification — each costs something that never shows up on an invoice. It shows up as bookings that never happened. The right question isn’t “does this protect us?” but “does it protect us from something that actually happens — and is that protection worth the customers it deters?” At Hotel Silva, the answer was clear.

What changed

A booking engine with instant confirmation. The inquiry form was replaced by an online booking system. The customer picks dates and a room, fills in details, and receives confirmation by email immediately.

Booking without a card. The card requirement was removed entirely.

Equally important is what didn’t change: the design, the photos, the copy, the site structure. Only the two specific barriers in the booking path. That’s what makes the result cleanly interpretable — it doesn’t compete for credit with a redesign or a new campaign. A targeted change to two spots, each with a clear rationale, is cheaper, faster and measurable.

How the result was measured

For a seasonal business like a hotel, the measurement method matters as much as the change. A “month before vs. month after” comparison would be worthless — the difference between May and July is the season itself.

Direct bookings were therefore compared for the same period year-over-year. The YoY comparison filters out seasonal swings and shows whether customer behavior changed — not the weather. It’s the more conservative metric, which is exactly why it’s credible.

The result

Direct bookings grew 23 % year-over-year.

Every additional direct booking saves the hotel the OTA commission — a significantly higher margin per stay. And with a direct booking, the hotel knows its guest: it has the contact, can communicate before arrival and after departure, and can reach out directly next season. That makes direct bookings a double win — higher margin today, and a better starting position for repeat business tomorrow.

The takeaway

Customers arriving at your website are already interested. The problem usually isn’t that they can’t find you — it’s what they experience once they’re there.

Three things decide:

Let the customer act now. Every “we’ll get back to you” form is an opportunity to lose them.
Question every step in the process. Not every rule that protects you is worth the friction it creates. If the risk almost never materializes, the rule probably deters more customers than it protects you from.
Measure so the result can be trusted. A targeted change plus a season-proof comparison gives an answer. A broad change plus “before/after” gives an impression.

If people come to your site but don’t convert, the barriers are on the site — and they’re the thing you fully control. Kaen watches for exactly these barriers continuously: it reads your analytics, finds where customers drop off, and proposes the test — like at Hotel Silva, just without waiting for a one-off audit.