Anonymised client results

How a healthy booking business leaked for eight months

The tours were running and the reviews were excellent. Revenue arrived every week. For eight months, a broken piece of tracking quietly fed the ad platform false data, while everyone spent against it. This is how we found the leak, how we fixed it, and how the same business went on to take 70% of its turnover through its own website.

Broken tracking found and fixed Ad budget cut, returns up Direct share: 60% to 70% Every change measured

The situation

Everything looked healthy.

This operator had what most booking businesses want: tours sold out in peak weeks, positive reviews from guests, real revenue through their own website, and the online platforms brought in bookings on top of that. On every surface that mattered, the business was working.

Despite these successes, the owner remained with questions that were not fully resolved yet. A meaningful slice of every platform booking left as commission. The ad account spent money every month, and nobody could say with confidence what exactly came back. Opinions on what to do about this differed from advisor to advisor. One would suggest increasing the budget, whereas another would recommend cutting it. Both had reasonable arguments, which was exactly the problem. When the numbers cannot settle an argument, every decision is a guess.

However, the problem was not the budget, but something much simpler and worse. The numbers on which everyone relied were no longer accurate.

  • Project type Ongoing technical management of a direct booking channel
  • Client type An established tour operator with serious online sales volume, shared anonymously
  • The core problem Channel and budget decisions were resting on corrupted tracking data
  • TheTechDodo's role Measurement, advertising, structured data, booking-flow work and monthly reporting

The hidden problem

Nothing looked broken, and that was the problem.

At some point, a small change to the site's conversion tracking corrupted the data flowing to the ad platform. This happened quietly, as no alerts or errors appeared in the system. The dashboards kept filling with plausible numbers, and the reports kept being read. For eight months, the platform optimised campaigns against inaccurate values, and everyone involved made decisions based on them. The operator was told the answer was to spend more. They did, but nothing changed.

That is what a booking leak actually looks like from the inside. Rather than a problem that is ignored, it is a problem that nobody can see. On top of this, it bills you twice for every month it survives: once in ad budget spent training an algorithm on fiction, and once in commission paid to platforms for bookings your own website should have gotten.

Could this be you?

Three questions to ask your own numbers today.

No tools or login needed. If the honest answer to any of these is "not sure", you are standing where this story started.

01

Do you know your real return on ad spend, from your own booking data?

Platform dashboards grade their own homework. The number that matters reconciles what you spent against what your booking system actually recorded. If nobody has done that reconciliation, the real figure is unknown.

02

Is your website's share of turnover rising, or is the platforms' share rising?

One number answers who owns your growth. If it is not tracked, the answer is usually the platforms, and the commission bill compounds quietly.

03

When did a human last verify the numbers your ad budget optimises on?

Tracking breaks silently, and ad platforms keep spending either way. Eight months is what unverified looks like. If the answer is "never", the check is overdue by definition.

Technical appendix

The full diagnosis, for those who want it.

The story above skips the technicalities on purpose. This section is for technical readers, and for owners who want to know what checking actually means.

Stack and environment

A WordPress-based marketing site with a commercial booking engine behind it, analytics and tag management in front, and paid search running on top. Structured data is generated server-side, unaffected by routine content and plugin updates.

How conversion tracking breaks without anyone noticing

A configuration change upstream of the conversion tag altered the values passed to the ad platform. Every report kept rendering, because reporting systems do not know the difference between wrong values and right ones. Detection required tracing individual real bookings end to end. This involved tracking every step, from the click through to the value recorded in the platform. This was then compared against the booking engine's own records.

Measurement and privacy

All tracking runs consent-aware. Results are verified against the booking system's own records rather than analytics alone, and reported to the client monthly in plain language.

Known limitations

Attribution never captures the full picture. Bookings made by phone, by email or on a second device are systematically under-counted. As a result, the tracked improvements are floors rather than ceilings. In additon, the work is not finished, because the platforms never stop changing.

FAQ

The questions this story usually raises.

How do you know if a booking website is leaking direct bookings?

Common signs: commission paid on guests who already knew the operator, ad spend without a reconciled return, a website share of turnover that stays flat while platform share grows, and tracking nobody has verified end to end. To reveal these leaks, a technical audit compares what the systems report against what actually gets booked.

Why didn't the ad platform notice the data was broken?

Ad platforms optimise toward whatever conversion data they receive. Corrupted values are still values, meaning that campaigns keep running and budgets keep spending. There is no alarm for 'your numbers are wrong'. These checks have to be done by a person.

Would the same results apply to any operator?

Absolutely not. Results depend on traffic, offer, pricing, seasonality and operations. What transfers is the order of operations. We verify the measurement first, then fix the highest-impact problems one attributable change at a time.

Find out what your own numbers actually say.

The Direct Booking Leak Audit checks measurement, advertising, structured data and the booking path. The results are communicated in plain language. TheTechDodo personally replies within 2 business days. If an audit is not the right next step for you, we will let you know.

Request the Direct Booking Leak Audit

See all results →