inDriver

Challenge

Going into this project, I understood that we were attempting something ambitious. We wanted to rapidly scale up new user sign-ups for inDriver through a channel we had never tried before – in-app programmatic advertising – and do it across a wide range of countries and cities. Our campaign launched in January 2021 with a clear goal: acquire quality users on a CPA basis (driving app registrations) while hitting specific daily sign-up targets in each region.

This meant not only finding users in major metropolitan areas, but also in the smaller cities where inDriver’s expansion often begins. We knew this wouldn’t be an overnight success; reaching niche local audiences via programmatic would require careful tuning and enough time to gather data. I was also wary of the unknowns, from potential ad fraud to the possibility that in-app inventory might not deliver the volume or quality we needed. It was a challenging setup, but we were ready to tackle it head-on in partnership with Bidease.

Solution

From day one, our execution was a close collaboration between the inDriver growth team and Bidease’s experts. We kicked off by casting a wide net with our ads, deploying campaigns across the largest in-app ad exchanges that Bidease could access. The Bidease team helped configure the targeting to reach even the smallest towns if necessary. To jump-start the campaign, we initially concentrated spend in larger population centers and grouped similar geos together, which allowed us to gather sufficient conversion data quickly for optimization. In parallel, Bidease’s data scientists worked with us to create a custom lookalike audience model using our first-party user data. That way, even in the early stages, we were bidding more intelligently; we leveraged what we already knew about our best users instead of blindly searching for new ones.

Creatives were another area of tight teamwork. We began by testing video ads, since Bidease’s experience indicated that video would engage users more deeply. Sure enough, those video ads grabbed attention. They vividly showcased how inDriver lets you set your own price for a ride, which is our app’s standout feature. After running the video campaigns for a short while, we reviewed the performance metrics together. During these initial video-only tests, roughly 1 in 1,600 ad impressions resulted in an install. Seeing room for improvement, the Bidease creative team quickly iterated. They polished our top-performing videos and simultaneously introduced complementary static native banners. These static ads were shown to users who had watched the video but hadn’t yet installed the app, acting as friendly reminders of our value proposition. This one-two punch in creative execution paid off: soon we were getting about 1 install per 1,250 impressions, a noticeable boost in conversion efficiency after adding the retargeting banners to the mix.

Throughout this process, communication was constant and agile. I was in close contact with Bidease’s customer success team as we tweaked targeting, budgets, and bids based on the latest data streaming in from our AppsFlyer tracker. We set up dashboards to monitor key metrics like cost per install and impression-to-install rate in real time. When something wasn’t working, we caught it early and adjusted on the fly. After a few weeks, once we had collected enough conversion events, the Bidease team activated their optimized bidding algorithms trained on our registration goal. I remember the turning point clearly: as soon as the ML-driven optimization models kicked in, our campaign performance started improving dramatically without the need for manual adjustments. The algorithm began purchasing impressions far more selectively, focusing on users most likely to register. In effect, we handed over the reins to a smart system that was learning and optimizing in real time, and it quickly honed in on the best opportunities.

Results

The results of this hands-on, iterative approach were outstanding. Over the course of January to June 2021, we brought in roughly 200,000 new installs via Bidease. What’s even more impressive is how the pace accelerated: about 120,000 of those installs came in the final two months (May and June) once the predictive models were in fully optimized. By that time, our impression-to-install conversion rate had leapt to around 0.2%, up from about 0.06% in the early test phase, more than tripling our conversion efficiency. In practical terms, our monthly new-user acquisition through in-app ads ended up roughly three times higher than where we began before using Bidease’s programmatic solution.

Crucially, we hit all our key performance targets in the process. Even with the surge in volume, we maintained user quality. The vast majority of those installs turned into actual registrations in the app, and we consistently met the daily signup quotas for each region. My initial fears about in-app traffic being rife with fraud or low-quality users proved to be unfounded. Bidease’s anti-fraud safeguards and the platform’s focus on optimization ensured we were getting real, engaged users. The close collaboration and real-time optimizations were key to this success: the quick creative adjustments, the agile budget reallocations, and the timely deployment of ML optimization all combined to drive performance upward day by day.

Looking back, our partnership with Bidease transformed what started as a risky experiment into a reliable growth engine. Our success was the result of diligent monitoring, open communication, and letting a smart system steadily improve the campaign. We learned that in-app programmatic UA requires patience, but when done right, it can become a powerful driver of growth. Thanks to this project, in-app advertising is now an integral part of our user acquisition strategy, and we have a clear playbook for executing it successfully in every new market we enter.

‍