Use Case

Driving Product Innovation with A/B Testing and BI Insights

The Business: Digital advertising

Adikteev is a digital advertising platform with a strong focus on ad innovation and quality, but also visibility, interaction and content match.

A technological expert in the mobile and web realms, they cater to the needs of demanding publishers and advertisers. Their B2B services focus on both efficiency and aesthetics to satisfy the advertisers’ need for performance whilst the publishers’ requirement for a good user experience is met.
Focusing on the experience of internet users is key to high performing targeting, both specific and context-appropriate. Adikteev pays the utmost attention to internet and mobile users’ behaviours to serve the right ads at the right moment.

Frédéric Leroy, COO, told us how helps such a data-driven company accelerate insights on their technological service:

The Challenge: Alerting on business KPIs to nurture high company performance

We initially had technical alerts to watch over our system and wanted to go one step further by easily integrating business metrics as well. A straightforward access to tech and business metrics would indeed help us focusing on our core business. Simple access to metrics such as number of ads served, click-through rates, loading latency or display availabilities was necessary for us to be free to focus on the intensive growth we were and still are experiencing.

adikteev tool

Considering the industry we are in, dealing with massive amounts of data in real-time, we needed a powerful tool to gather and analyze all the metrics we were planning to pour into it.
We started looking into a couple of open source solutions and were about to install them on our servers when we discovered, which seemed both easier to set up and more BI oriented.

1. The solution: Full visibility on tech & business metrics

In only a matter of days was ingesting our data and we saw graphs appearing: we were able to visualize our business metrics with very little effort!

“In 2-3 days we already had our metrics and I got to understand the power, but also the simplicity of the tool. I became its ambassador – they call me Mr. Logmatic”, says Frédéric Leroy, COO Adikteev allows us to easily get full visibility on our service. Seeing the data, slicing and dicing it has become much easier. So did understanding correlations and what was happening underneath. Its simple troubleshooting and alerting possibilities took much of the hassle out of our tech team’s day to day. All the time that has thus been freed is dedicated to our own core tech know-how, and developing new features.

“If it took us 20min to understand a bug, now it only takes us 30sec. Visibility has been greatly facilitated.” Frédéric Leroy, COO Adikteev

Previously, we could easily see aggregated numbers but filtering information was more complicated. So for example the number ad served / ad inventory was good. But as we wanted to keep on improving performance, we pushed to filter the information and see if the ratio was still as good for some specific websites. This proved to take up internal resources. Now our BI team analyzes logs and events and easily sees the overall tendencies. They also get the specifics right in no time.

2. Beyond the solution: Insights to accelerate operations’ development

As allowed for smoother correlation of tech and business metrics, we were pleased to see it facilitated our day to day decisions on operations.

Previously, getting the insight of why business metrics were in the red, when all tech signals were in the green was cumbersome. With, as everything is centralized, we can easily correlate business metrics and tech events.

One of the issues specific to our industry is getting a “all good” tech signal when in reality, things are not working part of the time. For example, a server needs to answer requests to serve ads. This server could sometimes answer and sometimes not: from a global monitoring point of view, as there is data, everything would seem to be working nicely. So all tech metrics would look good, but business metrics such as impression rate would be taking bad turn.

Now, our ability to deep dive into granular data, such as which user agent line is dropping, which event is crashing, has sped up. easily shows us everything – the exact logs, browser, country, or format – so we get to understand what is happening in no time.

“ improved our vision.” says Frédéric Leroy, COO Adikteev

We use the visibility acquired with to accelerate decision making for both technical and sales operations. Going back to the example above, once we realize there is an unusual ratio number ad served / ad inventory on a specific website, we investigate further and might realize there is an incident. Or if Click-Through Rate is not good as expected on a dashboard specific website, it could be that the ad placement is below the fold. Now that we visualise in a simple way what is happening behind a phenomenon, taking the proper actions from a technical or sales perspective has also become easier. It helps us properly back up the growth we’re experiencing, when everything is moving very fast. We’ve always been very keen on constantly refining our operational activities and helping publishers monetise their audience even better than they already did, and simply helps us doing it.

3. Beyond the solution: Supporting product innovation through A/B Testing

As we started to see our business metrics in focus more readily, understanding each release impact on business became simpler. It freed us in turn to go one step further and launch A/B testing on new features being released. We now proactively drive business performance by doing A/B tests on new released code and only deploy the most effective features.

“We are systematically A/B testing code performance for new features and analyzing their results on” says Frédéric Leroy, COO Adikteev
adikteev ad serving
We’ve always been very keen on driving our business performance and customer satisfaction based on data. helped us strengthen this very ability, it is exciting! We would publish, say, 2-3 versions of ad serving and let them run for 1 or 2 hours. Real-time data thus flows into the dedicated field we created in The BI team first analyzes the results and checks users’ behaviour. Then they decide in all awareness which version, A, B or C works best and should definitely be in production. We know for certain that our tool performance is getting better and better for our advertisers and publishers at each release.

4. Beyond the solution: Facilitating traffickers’ and publishers’ work

Speed forward a couple of months after setting up and many teams were using it: from Tech to Publishing, Trafficking and BI. Once they got their hands on the tool, everyone discovered thatvaluable data was accessible straightforwardly thanks to its simplicity of use. As we said, we’re a data focused company, so our business teams were very interested in easy access to data, and they were keen to build their own customized dashboards to fit their needs.

publish ad adikteev
The trafficking team builds and launches campaigns online while the publishing team takes care of publishers. Working with real-time data accelerates both teams ability to get insights, and they can focus on their real know-how. They monitor campaigns with great details more easily and take actions to optimize performance very rapidly.

“We’re able to see delivery rate, impression rate, CTR… Was the iframe reduced? How long did the user stay on the publisher’s website? All events are visible in real-time” Frédéric Leroy, COO Adikteev

Traffickers directly fine-tune campaigns to improve delivery or optimize performance while looking at the best performing ads. They access granular data and know exactly how campaigns perform at 9, 11am or 9pm. They gather data such as interaction rates or length of user interaction with the ad, and analyze it in a report for our advertisers.

Adikteev using showcases an example of great innovation to drive product performance as well as steering tech & business operations with machine data.

Discover more stories from our clients and how they use to troubleshoot, to enable the Support’s team work or to improve user experience.

« Previous Next »

"We’re able to see delivery rate, impression rate, CTR… Was the iframe reduced? How long did the user stay on the publisher's website? All events are visible."


Frédéric Leroy
COO, Adikteev

Let's get started for free !

Start your Free Two Week Trial Now. Get setup, and your first insights in minutes.