Custom Solutions for AdTech Company

About the Client

MediaMath is an AdTech company that offers advanced digital solutions for thousands of advertisers, including the most prominent world-famous brands. The platforms and tools that the company provides efficiently address the needs of businesses of different types and sizes. Among the available products are solutions for targeting, audience segmentation, ad campaign management, etc.

Goal

MediaMath is one of our regular customers, and we have been working on several of its successful projects already. 

It was decided to build a row of additional supportive solutions to make the development process smoother and eliminate the necessity to perform repeated tasks manually. These products can be defined as helpers that include some set of resources for future projects on which our team or our client’s team will work.

Moreover, our client asked us to build separate solutions that expand the existing products’ functionality.

Work Description

Apart from the main projects we built for this client (a scheduler system for ad campaigns and strategy scheduling integrated into the client’s DSP and an ad management system for large advertisers), we also created a row of supportive solutions.

The duration of the development of each of these projects varied from 2 to 5 months, depending on the specificity and complexity of each product. The project realization process in each case was organized following the principles of Agile methodology, which allowed us to stay flexible and deliver solutions of the highest quality.

It is quite natural that within the development process, a client may introduce new requirements and fresh ideas. Our task was to find the best approach to seamlessly implement them into the existing solutions and parts of the work done. Such situations are usually quite challenging, but thanks to our specialists’ expertise, we achieved the set goals.

We worked in close cooperation with the customer’s team. From our side, we have a team of 4 members: 

• 2 full-stack developers;
• 2 QA specialists.

The communication was built in the following way. We communicated daily within our team and Zoom calls with a tech lead once a week. We also conducted planning and review online meetings at the beginning and the end of each sprint, which helped us stay fully tuned with all the changes and updates that could appear from our customer’s side.

Here is a range of solutions that we created for MediaMath.

Graphic representation of the results of ad campaigns

The solution can analyze the outcomes of ad campaigns and visualize them in graphs and diagrams, making it easier for advertisers to see the effectiveness of their efforts.

This tool can be viewed as a separate stand-alone product that can be combined with other AdTech solutions.

To cope with the set tasks, it was necessary to work with graphics libraries, which allowed us to deliver a solution that successfully met our customers’ expectations and requirements.

AdTech entity validator

This system allows users to validate and verify data related to ad campaigns, strategies, etc. Thanks to it, the MediaMath team can see whether the data provided by its clients is correct or if it was supplied and transmitted with any errors. This solution was packed as a separate component and was shared with other teams via a Python repository.

Thanks to it, it is possible to avoid creating the functionality for validating data manually each time it is required for the next solution.

Automated test package

While working with MediaMath, all our products had to be well-tested, and we paid much attention to this requirement. Given the specificity of the projects, it was essential to automate the testing process. That’s why it was decided to create a separate package that would automatically test the systems we were building.

Thanks to this solution, it was possible to start running tests with the help of writing just one line instead of doing all the required work manually.

This package of tests was also placed in a private repository. It means that other development teams can also work with it and reduce the time and effort typically required for testing.

Website for feedback gathering

Another sub-project that we worked on was creating a centralized website for collecting feedback from users of all the AdTech products we had previously built for MediaMath.

When users open a website, they can choose a product from the list and share opinions, and we can see all their comments in our email box. The support team can make decisions and take the required measures based on the received information.

Boilerplate (project template)

On our projects for MediaMath, we used practically the same tech stack. That’s why building a so-called boilerplate – a project template was sensible. We didn’t need to write the solution from scratch for future projects as we already had some ready-made parts.

For example, we could use an API template with the authorization form and some essential tools for the backend. For the front end, we also had a template that had a login form and a core set of React components. In other words,  we had some minimal functionality for our new projects that we could use to save time.

The existing set of libraries requires regular updates to stay relevant and bring real value to developers.

Work Stages

1. Requirement gathering
2. Analysis of the collected data, planning
3. Agile development cycles: development- testing- production
4. Release

Results

Our team has been successfully collaborating with MediaMath for years; these projects were not an exception. We already have well-established workflows and communication formats, making it easier for us to understand the requirements and specificity of the tasks.

These projects were aimed at facilitating the development of other products and the enhancement of the functionality of the existing ones. And we managed to ensure the desired results.

According to our estimates, thanks to the offered solutions, it has become possible to reduce the development time for new products by 20%-30%. It is important to mention that our and other teams can leverage the new possibilities in development thanks to these products.