Currently, Slevomat is leading the way by using a data-driven attribution approach to marketing channels optimization. For a company always looking to edge the competition and spend money wisely Slevomat challenged ROIVENUE™ with this goal and the platform delivered incredible return on investment of over 19% within key marketing channels.

 

         The Process

 

By analyzing the current performance of Slevomat’s digital marketing mix, the ROIVENUE™ platform was able to quickly:

 

1. Break down all incoming traffic into 45 meaningful manageable channels​

 

2. Analyze ROI performance of said channels based on data-driven attribution models

 

3. Identify performance and propose budget reallocation to channels with higher ROI

 

This allowed the ROIVENUE™ team to analyze the campaign’s limit saturation, and find opportunities for growth using ROIVENUE’s Budget Optimizer feature. The optimization process was to reallocate budget from low performing channels to those identified by the ROIVENUE™ attribution suite as high performing channels.

 

         Results

 

Using ROIVENUE™ Budget Optimizer the Client Success Team recommended a two wave course of action.

 

During the first wave, the data collected suggested that Slevomat increase their spend on Facebook and Adwords by only 0.20%. Basically the overall budget wasn’t changed, but reshuffling budgets between channels resulted in revenue increase by +19.67% and ROI increase by +19.41%.

 

During the second wave, the ROIVENUE™ Budget Optimizer aggregated the marketing attribution data to suggest a focus of marketing efforts on the Criteo platform.The ROIVENUE™ Client Success team suggested to split Criteo ROI measurement from one overall channel to three distinct channels within then traditional funnel: Upper, Middle, and Lower level.​

 

This allowed Slevomat to use ROIVENUE’s multi-touch attribution models to optimize budget spending precisely in the section of the funnel that would give them the greatest marketing ROI.

 

By redistributing budget from more saturated parts of the funnel to the lesser saturated ones, (the ones with higher ROI according to data driven attribution), overall ROI of the platform went up.This time, budget was slightly decreased, and even with that and proper budget split the overall performance resulted in revenue increase by 5.92% and marketing ROI boost by 12.64%.​

 

“Optimizing only selected channels with ROIVENUE™ during the initial phase assured us that data driven approach is the right way. Moreover, the level of ROIVENUE’s support was just phenomenal.”

 

David Matoušek, ​CMO Slevomat

Challenge: Optimizing Acquisition Costs Per Predicted Loan-Seekers’ Success

We expected that a certain number of loan-seekers, which were targeted with ads, would start filling the questionnaire. Out of those who finish the form, 70% of the loan requests will be refused outright, and 25% of loans granted will default. If we were to succeed in moving those percentages (conversion rates), it would translate into a significant lowering of the marketing costs associated with getting the initial requests and reduce the burden posed by defaults.

 

Our Solution: Optimization Of Ad Investment Relative To The Probability Of Loan Repayment

We used the Roivenue Performance Monitor to analyze the cost of every conversion step and measure the performance of the different conversion stages. We then created prediction models to measure the effectiveness of marketing channels from the overall view.

 

Roivenue can evaluate cost of marketing investment relative to Customer Lifetime Value (CLV), and differentiate between new customer acquisition and retention of existing customers. This simplifies the optimization of campaigns to a very basic level: invest more in effective channels and lower investment in problematic ones. The next step was to analyze the actual form-filling process – in real time.

 

We used the data-mining algorithm in R to calculate the probability of a given user’s progress on filling out the loan – uncovering several mistakes in form code and possibilities for improvement of the UX. Most importantly, we were able to classify different types of  applicants, predict the level of loss in relation to segments, and discover problematic (potential) clients. This included certain patterns that were in fact attempts at fraud.
​

Result: Stable Growth And Extremely Low Level Of Loan Defaults

The continuous improvement of performance parameters lead to stable growth and an extremely low level of defaulting clients – just 15%. Thanks to several revisions of the form based on Roivenue data, we succeeded in raising the conversion rate for those who filled the form to 36%, and lowering the cost for new client acquisition to 30% CLV.