Enabling users to evaluate the comparative effectiveness of different marketing communication channels

Marketing data and drawing conclusions from them is the basis for the effectiveness of marketing campaigns and promotions, regardless of the industry in which we operate. Observing the market, we noticed that many online marketing tools don’t allow for the automation of the analysis of this data. This is a niche in the market that we decided to fill.

That’s how we came up with the idea of Predicty, a SaaS platform for analyzing collective marketing data.

Service

Product Management + Delivery

Technology

Custom

Vue

Scope

Idea to release

Vertical

Data Science

01

Why were we asked to help?

iOS14.4 was a big turning point for many companies. The lack of proper analysis for Apple devices created a big hole in the analytics market, so we decided to fix it. Our first impression was that marketing mix modeling (MMM) might be a solution, but as we later found out, that's just the tip of the iceberg. It turns out, obviously, you can have any algorithm you want, but if you input data sorted in a wrong manner, the results will be useless. We tried breaking down acquisition by channel, but that didn't move a needle. Later on, we monitored how agencies work, and that's where we got the idea of segmenting the data by communication and not by channel. This was a ground-breaking pivot.

Stanisław Świątkiewicz, Project Manager of Predicty

We saw the issue of many agencies and marketers: in order to understand the effectiveness of communication, they had to look through each platform and manually compare data in Excel sheets or other numerical and visual data management tools. The idea itself was also discussed with more than 10 marketing agencies who saw a real need and pain point with gathering and comparing data for their clients.

02

How did we tackle the tech?

The very first minimum viable product was created back in 2021. Our aim was to assist our client, LPP, in determining whether investing in offline stores or allocating the same budget to promote online ones would yield better results. To achieve this, we needed a comprehensive understanding of the impact of each dollar spent both per store and per advertisement. With this in mind, we researched available technologies, discovered MMM, and settled on this technology. In this project, we opted to implement the lean startup methodology.

Predicty’s purpose from the beginning was to allow users to compare various communication channels and their effectiveness with each other. In order to do that, it connects with all major marketing platforms and collects the data in one place. Following this process, trained artificial intelligence (AI) models are used to assist users in matching identical communications across various channels. For instance, this could involve matching promotions such as "50% off for first-time buyers" on platforms like Instagram, Google Ads, and TikTok.

Matched and filtered data are then displayed on "the Timeline" that allows for further analysis. Users can use the primary dashboard to compare different communications with each other. Subsequently, they have the option to delve deeper into a specific communication and analyze its specific components.

Our team consisted of two front-end developers, one back-end developer, a product designer, and a project manager.

03

What did we achieve?

A challenge makes success even more satisfying

The work put into the project paid off. We created a product that was easy to scale while validating the business idea and market need. We also managed to build a community around our open-source project that supports the group involved in the development of the idea.

Once the project was completed, we assembled all the team members to evaluate the project and its implementation. This provided us with valuable insights to aid our future decisions on when and how to implement them in upcoming projects.