Uncover allows diners to make last minute bookings at over 350 of London’s most desirable restaurants. Unlike other top restaurant booking apps, Uncover provides a “curation” service, hand-picking a shortlist of venues with guaranteed tables.
Uncover was in hyper-growth mode as well as fundraising, which meant their primary goal was continuing user growth. In order to achieve this, Uncover turned to the Appsumer team, who were tasked with tackling four main challenges:
Defining optimisation KPI
On average, the gap between a user downloading the app and making a first time booking is 30 days, rendering it an impossible event to optimise towards. We therefore set out to identify a metric within 3 days that correlated with bookings made after 30 days.
We worked closely with Uncover’s BI team to identity this correlation. A total of 60 behaviours were taken into account to produce a predictive conversion rate.
Using our tool, we were able to aggregate all ad spend and conversion data at the most granular level. Our unification system meant we were able to marry predictive conversion rates with ad spend in order to assess performance based on predictive CPAs, within 3 days.
Automating the data workflow process meant eliminating the headache and wasted hours spent attempting to achieve this in excel. We could also provide access to real time ROAS through automating each supplier Uncover was working with.
Optimise to predictive KPIs
We identified the best performing pockets of inventory and optimised towards predictive CPAs.
Optimising to predictive metrics allowed us to reduce actual CPAs by 72% within 3 months, as well as increasing volume by 4x. Not only did this prove profitable unit metrics for the business but also met the need to scale incrementally, and ultimately led to the acquisition of Uncover.
“Appsumer were able to significantly improve our ROI within just a few months. This was made possible thanks to their platform and its ability to unify our 1st party data from our BI Team with all our paid media channels, allowing us to optimise for quality users at a granular level.”