Alvar Carto is an online web service that lets its users create and order unique custom-made map posters — from any location on the planet. By printing all posters in Finland, they assure exceptional high-quality prints and designs. And there is worldwide free shipping.

They started out about a year ago, and although sales have really lifted off (their stunt on Reddit definitely helped), they wanted to understand if there are certain behaviour factors that their buyers have in common.

Example of Alvar Carto’s online map design tool — www.alvarcarto.com

While Google Analytics certainly gives some insights, their analytics are often limited in showing the underlying reasons why certain behaviour occurs. It is hard to find the correlations from GA.

In comes Aito.

By importing their GA-data into Aito, Alvar Carto was able to uncover surprising correlating factors with their buyers’ behaviour.

The project setup

Alvar Carto tracks an event in GA for each activity a user does on their website (e.g. searching a city, zooming in to a map, zooming out, changing colours). For this project, the data set of 1 month was used.

Example of the analytics data, ready to be imported into Aito

Once the data was imported to Aito, the RELATE-query was used to understand the how users’ behaviour influenced their final purchase.

The query was done separately for each event, and Aito provided a visualisation of the results.

The results

Relate query visualised — the bigger the dots, the higher the correlation of these factors

We picked out the top 5 interesting events:

How the top 5 most interesting events correlate to the basic conversion

IE users are 2x more likely to make a purchase

Conclusions

While certain results were obvious:

• The longer a user fiddles around with a map (= higher eventCount), the more chance that he/she will actually convert.
• Users on a tablet or desktop are more likely to purchase, compared to users on a mobile.

But there were also quite unexpected results:

• Customers who change the poster size selection, are 1.94x more likely to make a purchase
• Users on Internet Explorer are 2.8x more likely to make a purchase

For this article, we only took a couple of highlights.

Aito helped us to look at our analytics data from a different angle and understand the buying behaviour and its correlating factors better.

— Kimmo Brunfeldt, co-founder of Alvar Carto

About Aito

Aito is a predictive database for developers who value quick time to market. It’s familiar as a database, yet provides the intelligence of machine learning. Predictions, recommendations, and explanations are provided in milliseconds over beautiful API, using the entire Aito database as training data.

Aito is a Helsinki-based startup, spin-off from Futurice, established as an independent company in the spring of 2018.

Aito is currently in private beta.

Apply for early access at https://aito.ai