Secret Escapes runs geo experiments through AdWords to understand true cost of search marketing in driving new leads

Aug 2017

Secret Escapes is an exclusive members-only travel club offering members discounts and deals on hand-picked luxury hotels and holidays.

About Secret Escapes
Members-only travel club
Founded in 2010
Headquarters in London
The Goals
Set cost per lead and cost per acquisition targets based on customer lifetime value.
Validate econometric models through incrementality testing.
The Approach
Conducted controlled experiment using geo methodology.
Tested cost per lead in generic search in German market.
The Results
Proved generic search drove 10% of total incremental leads with a cost per lead that was 50% lower than econometrics suggested
Increased budgets by 25% across five markets to capture more profitable demand

New prospects sign up for membership via the brand’s website. Through retargeting and CRM activity, Secret Escapes then aims to convert prospects into customers by encouraging them to complete a holiday booking.

To analyse new signups, measure leads and set last-click cost per lead targets, the team used last-click attribution. The targets were calculated via a combination of econometric modelling and cohort analysis, or customer lifetime value. With econometric modelling providing relatively high cost per lead estimates, the team decided to conduct a geo experiment through AdWords to provide new insights about incremental signups.

“Implementation was straightforward”, says Head of Digital Marketing David Nefs. “We were able to deploy our geo experiment with ease.” Because the robust measurement framework was already integrated with AdWords, Secret Escapes could take advantage of tools to aid experiment design, tracking and analysis.

The first step was to divide the country into geographical regions and then randomly assign each region to either a test or control group. Media spend was then downweighted for the test group using geo targeting in AdWords while media spend for the control group was kept at the normal levels.


"We weren't comfortable that econometrics was able to fully account for the incrementality of last-click PPC signups. The geo lift methodology allowed us to estimate the incremental cost per lead with far more confidence, suggesting it was around 50% of the cost per lead estimated by econometrics."

- David Nefs, Head of Digital Marketing, Secret Escapes


Secret Escapes’ experiment focused on incremental cost per lead. “We tracked the spend deployed in the geo region and the overall leads generated in the experiment”, David explains. “Google was able to use those together to model the incremental leads driven by generic search and calculate a cost per incremental lead. This was compared to both our last-click cost per lead and incremental cost per lead from our own econometric modelling to give us the full picture.”

The results showed that the non-brand cost per lead was 33% higher than last-click attribution, but 50% lower than econometrics had previously suggested. “Using geo experiments enabled us to spend more than we would have taking econometrics at face value”, David says. Secret Escapes has now increased search budgets by 25% across five markets in order to capture more profitable demand.

Looking forward, Secret Escapes has plans to roll out further geo experiments. “It’s a useful measurement tool”, David affirms, “so we’d like to extend it to other markets and to our member targeting activity, which aims to generate bookings. We intend to test our other marketing efforts via geo experiments and other incrementality frameworks in partnership with Google.”