When we launched the booking engine at Onera Wimberley, we had a problem.

The page showed guests a price with a long string of decimals trailing off the end of it.

The problem is, nobody will enter a card number on a page that doesn't tell them what it costs.

The dropoff from that page wasn’t flagged in a report. A guest didn’t write in about it.

We found out only because someone on our team happened to walk through the flow… And there it was.

Here's the part that stuck with me:

Almost every other problem in this business leaves a trace somewhere.

A bad month shows up in the pace report.

A weak reel shows up in the view count.

But friction in your booking flow?

It sits in a part of the funnel that nobody opens, and the only person who ever experiences it is the guest... Who closes the tab.

If you’re like most operators, nobody on your team has tested your booking system in the last 6 months

For a long time we assumed our checkout was fine, because it worked the day we built it.

We also assumed that if something was wrong that…

• Someone on the team would notice
• A guest would tell us, or
• The website would flag it

But this isn't reliable.

For starters, your team uses an internal-facing part of your website. So they wouldn't be seeing the same things as a guest.

Guests also almost never write in about a broken form. (They’d just close the tab and go back to the OTA listing where the checkout does work.)

Finally, your website has no idea anything is wrong, because from the platform side, the page loaded fine.

We found a version of this problem where the email field validation was so strict that it rejected guests’ email addresses.

Guests tried to book, got rejected, tried again... And gave up.

Again, guests don't report friction. They just leave.

Which means you have to…

Be the guest, twice a month

The fix is simple: Make two test bookings a month.

One on your phone, one on a desktop.

Go through the whole flow, all the way to the confirmation screen.

And don’t stop at the payment screen because if you do, this skips the part that's most likely to be broken.

Which defeats the whole point.

Just pay for the booking, and refund it later.

Then put this test on a recurring calendar invite, because this is the first thing that falls off in a busy month.

Now here's 4 choke points to watch for when you do this:

• Field validation. Test with a real email, a plus-addressed email, and a phone number that doesn't start with +1.
• Number rendering. Is every price, tax and fee showing as a clean figure?
• The handoff. Does the domain change when you hit book? And does the page still look like your brand when it does?
• Mobile keyboards. Is the date picker easy to use or does it make you want to tear your hair out?

And make sure to test your site’s speed on cellular, not on your office Wi-Fi.

For the sessions you can't do yourself, Microsoft Clarity gives you heat maps and anonymised recordings of real visitors, so you can see where people stall.

(The reason we use Clarity over GA4 is because GA4 is built for technical people, while Clarity is built for operators.)

One last warning before you start changing things:

Fixing friction and switching platforms are two different jobs (and we learned the difference the expensive way).

We chased better data (and it cost us bookings)

For years we optimised the parts we could see on a dashboard, ignoring the parts of our flow that we couldn’t track.

At one point we switched our original PMS and booking engine for one that promised better tracking.

Yes, the tracking was better.

But the booking engine was worse, in exactly the small ways I mentioned above.

Conversions fell off a cliff.

So we switched back, and they stabilised.

Afterwards, this allowed CartStack, our abandoned cart tool, to recover an additional $200,000 across both Onera properties in the trailing twelve months.

But we needed a working checkout flow first before we could make any downstream optimizations.

Because if your checkout flow is broken, your guest won’t even get far enough to abandon cart in the first place.

You need to find the friction (beyond the dashboard)

​Direct bookings is the one channel you own, so it has to be the smoothest path you offer, and the only way to know involves you going through it yourself.

And you can't just rely on a metrics dashboard. It can tell you people are dropping off at checkout. But it will never tell you why.

So before the end of this week, book your own room. Pay for it, then refund it.

If nothing breaks, you're out five minutes.

If something does, you've found the cheapest bookings you'll recover all year.

— Ben Wolff
Founder, Oasi & Modern Hospitality Accelerator

P.S.. I’ve got two free workshops coming up, and I wanted to invite you guys to both:

• How To Create Viral Instagram Content From Your Phone (Whether You Want To Be On-Camera Or Not), with Isaac French. Friday, August 28 at 1pm EST. Save your seat​
• How Glamping Resorts Can Drive Millions Of Ideal Guests To Their Direct Booking Site (Without Spending Hours On Video Shoots & Edits), with Brian Linton. Thursday, September 3 at 1pm EST. Save your seat​