Optimaizer:
Integrating a custom Machine Learning advertising model with Display & Video 360 platform
Location:
Poland
Technology:
Robocorp
80
hours of tedious work saved/monthly
0.5
FTE saved
Client
Optimaizer is a technology-first advertising agency from Warsaw, Poland. Their product is a Machine Learning model that collects the data from digital advertising, analyzes them, and optimizes their performance.
Challenge
The client’s challenge was straightforward - to automate passing the digital ads data between their custom AI optimization model and Display & Video 360 advertising platform from Google.
There was no publicly accessible API from the platform, so most of the work was done manually. The tasks were straightforward to deliver by human employees, but when scaled - it took quite a lot of time.
Solution
To solve Optimizer’s problem, we decided to leverage the Robocorp platform. Optimizer had some Robocorp setup already, but it had issues they didn't know how to address.
We have rewritten the already existing PoC of the solution and created a bot that:
- Automatically retrieves raw ad performance data from DV360 for the Machine Learning model
- Waiting for the model to perform calculations
- Extract the calculations from the model and pass the outcomes to Divi 360 daily at 8 AM.
- Creates a post-run report about the progress and displays it in a Robocorp dashboard
In other words, Robocorp has become a bridge between the client’s system and Google’s advertising platform.
We have also built a feature that allows the Client to send the data to the bot directly via e-mail. That allowed it to process even the most extensive files (the size limit was virtually impossible to meet).
Thanks to Robocorp’s user-friendly features, we were also able to quickly build reporting dashboards that can be accessed in real-time by every member of Optimizer’s team.
Outcome
In four days, we have created a solution that 100% solved the Optimizer problem.
Robocorp enabled Optimaizer to shrink the time between calculating optimal advertising strategies and their implementation, and it was the glue that kept their data-driven marketing pipeline together.
It produced real value for Optimaizer’s users - their ad campaigns were adjusting to changing conditions way quicker than before, translating to better performance and higher Return on investment.
It also provided tangible savings for the company - our bot saves around 80 hours of manual labor per week, which equals hiring one half-time employee.
It was a short Robocorp development mission, but it hugely impacted our Client’s business!
80
hours of tedious work saved/monthly
0.5
FTE saved