Categories
Digital Analytics

Algorithmic attribution of digital touchpoints in conversion

The telco client was running a digital customer acquisition campaign. They were advertising heavily across multiple platforms – social media, popular news websites and through agency adtech platform. They wanted to study the impact their ads were having across different digital channels, and how each touchpoint was contributing towards final conversion of the customer. They were also interested in the RoI from each channel and the optimal touchpoint frequency.

Identifying where credit is due

The client wanted to optimize their digital marketing budget using data-driven insights and using them to generate maximum return on investments (RoI)

More accurate distribution of weightage across all touchpoints

Identified cost per acquisition (CPA) and recommendations to reduce CPA

Optimal number of touchpoints for best conversion propensity

The study helped the telco identify leakage in digital campaigns, and optimize their future campaign strategy to minimize cost per acquisition (CPA) and optimize campaign spend.

Too much data. Scarcity of expertise.

The telco was using the rule based attribution models – First-Touch, Last- Touch and Time-Decay based models. All these models were giving more than 80% weightage to Google touchpoints. The client was aware that other touchpoints were contributing more. They also wanted a visibility on the return on investments and the optimal number of touchpoints to maximize conversions.
The biggest roadblock to implementing a more data-driven approach to attribution was the massive volume of data being generated. They lacked the expertise to unlock the insights in conversion data. The other challenge was dealing with a highly unbalanced dataset with less than 0.2% conversion rate. Our team provided the capability to combine our data handling and modeling expertise to unlock the strategic insights.

Beyond Last-touch, First-touch & other rule based attributions

Algorithmic attribution helped our client unlock the key marketing insights that were hidden in their data.

With algorithmic attribution (or multi touch attribution or MTA), the client was able to unlock key insights, including:

a. A more accurate estimation of credits across all touchpoints vis-a-vis rule based attribution models

b. Cost per acquisition (CPA) estimates and better understanding of the touchpoints that work best

c. Optimal number of touchpoints to maximize conversions, and potential campaign cost save

Multi Touch Attribution (MTA) de-duplicates user IDs across all channels, providing a more accurate, customer level view of the conversion journeys.

Leave a Reply

Your email address will not be published. Required fields are marked *