Close this search box.
Close this search box.

Case study

A platform for segmenting users and advertising messages based on data from mobile apps and personality traits


Advertising in mobile apps is significantly different from advertising in a web browser.

Share this:


In the first case, an incorrectly selected advertisement entails a significant risk for the app owner (e.g. content publisher, e-commerce store or loyalty system).

Developing a mobile app and acquiring a user is a high cost. On the other hand, the exit threshold for the user (i.e. uninstalling an app from the phone) is very low – ‘two clicks’. As a result, app owners must take into account not only the revenues related to the broadcasting of advertisements but also the costs connected to the loss of the user too.

The goal of the owners of apps that earn on advertising is to effectively monetise the advertising space in the app while taking care of user satisfaction. Advertisements broadcast in such an environment should not disturb a user so much that they uninstall the app due to overexposure.

The advertisers’ goal is to achieve the best possible results in their advertising campaigns. The challenge is to limit the number of ads that are displayed.

Organisation profile

Spicy Mobile offers comprehensive support to all entities interested in advertising in the mobile channel – publishers, advertisers and app developers. They are a team of mobile marketing experts – a precursor of the first research on the use of smartphones and tablets on the market.

They are the creator of the largest premium advertising network in Poland in apps (Mobigate) and on mobile websites.


Our experts are currently working with Spicy Mobile to create an innovative automated advertising platform.

The tool will automatically adjust the advertising message to the user’s personality traits based on ML technology. The main goal is to reduce the number of advertisements broadcast on mobile apps by improving their quality. Our experts are responsible for creating algorithms that will allow the software to

automate the data analysis process and detect relationships between variables so that they can be adjusted to the user’s personality profile each time. The project is co financed by Poland’s National Centre for Research and Development (NCBR).


Project successfully completed. The first algorithms have already been productively implemented. They allow traffic to be obtained on competitive terms. In particular, the algorithms have a CPC (cost-per-click) three times lower than the reference model of the algorithm.

You can read more about the implementation of the project results HERE. 

Get in touch

MIM Solutions can help develop AI in your company, please contact us and we’ll talk you through it.