We use cookies to ensure that we give you the best experience on our website. By continuing to use the website you agree for the use of cookies for better website performance and a personalized experience.

How to increase email campaign performance 10x - Deep.BI & SARE at Data Commercialization Summit

Pawel Blazejewski
.
October 5, 2017
How to increase email campaign performance 10x - Deep.BI & SARE at Data Commercialization Summit
Pawel Blazejewski
October 5, 2017
.
X MIN Read
October 5, 2017
.
X MIN Read
October 5, 2017
.
X MIN Read

On the heels of the IX DATA COMMERCIALIZATION Summit happening on October 4-5th in Warsaw, Poland, we're proud to announce that our COO Jarosław Góra and SARE CEO Dariusz Piekarski, one of our top clients, have spoken together about Big Data & Data Science's application in e-mail marketing campaign optimisation during the VI FORUM CUSTOMER OMNIDATA.

They've presented a real world example of using the Deep.BI platform to maximise a marketing campaign's profitability, which in this instance is increased by 1000%. Yup, the extra 0 is not a typo.

The project's biggest challenges were SARE's organisation strategy, resource engagement, deep understanding of 1st party data, and the trusting of AI findings and the testing of models in the real world, not to mention the scale, which was enormous. We optimized over 2 billion messages and various campaign models fueled by separated data sources.

Most Big Data projects do not bring an expected concrete, easily quantifiable ROI, which we are aware of at Deep BI. In light of this, we have designed a flexible and iterative process to drastically minimise financial risks along with the expertise we bring on Big Data, data science, and business intelligence. Actionable results are our primary goal.

An example of this is the possibility to form conclusions right off the bat by building aggregators flexibly on Deep.BI and understanding feature values and reliances. Feature engineering is roughly around 80% of the workload and is a key factor of AI algorithm success. Acting on raw data, we modeled reality by choosing a proper set of behavioral, demographic, and context features that fit well. We prepared a scoring system that included dozens of features for each user individually and sourced campaign target groups automatically. We tested out four different machine learning algorithms and finally, our Gradient Boosting Machine totally nailed it.

The comparative results were stunning. The CTR indicator for a batch sent e-mail campaign was between 300% to 1000% higher comparing to any previous project led by SARE. [Mic drop]

Along with this successful campaign, SARE not only increased their business value, but also understood their data better and identified other business areas for strategic and operational optimisation in the future.

Interested in optimizing your campaigns? Let us know and we'll give you a tour!

Subscribe and stay in the loop with the latest on Druid, Flink, and more!

Thank you for joining our newsletter!
Oops! Something went wrong while submitting the form.
Deep.BI needs the contact information you provide to contact you. You may unsubscribe at any time. For information on how to unsubscribe and more, please review our Privacy Policy.

You Might Also Like

No items found.