Close this search box.

GUEST COMMENT: Big Data and marketing attribution reveal the bigger picture

This is an archived article - we have removed images and other assets but have left the text unchanged for your reference

by Adit Abhyankar

The term ‘Big Data’ has become a key marketing buzzword over the past twelve months. But while there has been much rhetoric about its potential benefits, what does it mean in practice?

Big Data is a term applied to describe data sets whose size exceeds the ability of software tools to capture, manage and process information within a reasonable time frame. The size of Big Data can range from a few dozen terabytes to multiple petabytes of information in a single data set. For marketers, Big Data can deliver huge volumes of information regarding their interactions with prospective customers, yielding insights that help create strategies to optimise marketing performance. As a result, many companies want to be first in the queue to harness Big Data for competitive advantage – and the rise of marketing attribution technology makes the handling of this data fairly routine.

Sampling solutions

Marketing attribution has completely transformed the analytics landscape by enabling marketers to move away from more traditional methods of understanding data segments and patterns. Traditional data analytics processes involved creating a representative group from the total universe of available data, which was an extremely time-consuming and arduous task that could result in multiple iterations of data sets. It also ran the risk of creating and analysing a sample group that may not have been representative of the entire universe of users.

However, there is no longer a compelling reason why a business would take shortcuts in performing attribution or other data analytics that involve Big Data. With advanced marketing attribution technology, the entire universe of available data can be used to analyse and yield the insights that translate into actionable steps to optimise marketing performance.

New approach; compelling results

So how do traditional analytics methodologies differ from marketing attribution that is Big-Data capable? The traditional route involves creating a test group from a subset of data by subjectively applying a number of filtering criteria. For example, if a financial services company wanted to identify a group of potential customers who will churn within the next month, the data set could be limited to include only those accounts that:

  • Are behind on payments

  • Have a last customer touchpoint within 10 days of the account falling behind on payments

  • Have an outstanding amount of £50 or more

  • Have a credit score of less than 500

The issue, however, with narrowing down the with narrowing down the universe of data to a manageable and actionable subset is that subjective decisions are made about the elements included in the data group. This runs the risk that essential elements – of which the business may be unaware – may not have been factored in. As a result, this selected dataset could be systematically biased, and if used to do analysis, could lead to spurious conclusions.

By contrast, the Big Data marketing attribution approach to the same issue is very different. Instead of using a subset of accounts, the entire universe of users is segmented into different ‘buckets’ by a defined range within each of these groups, and an analysis is then undertaken to explore what is discovered about each range and in combination within individual buckets. Examples of some of the criteria for the buckets in this example could include:

  • Account status within the past month

  • Amount that account is behind on payments

  • Last touchpoint frequency

  • Last touchpoint type (e.g. bill payment)

  • Last touchpoint recency

Once multiple groups have been created from all the available data, more advanced questions can be asked of the data set, including:Which of the buckets yields the highest levels of customer churn?

  • What combination of factors yields the highest churn rate, i.e. credit score, amount and length of time that account has been behind on payments?

  • Of those customers that churned, did any have a specific touchpoint that could be highlighted? For example, a complaint with a high frequency (>3) within a certain time period of cancellation (<10 days).

When selecting a technology to perform marketing attribution, it’s essential that it is Big-Data-capable so your business can benefit from the advantages it delivers including the speed, accuracy, objectivity and multi-dimensionality of its analysis. Big Data offers an array of benefits when it comes to ensuring your marketing optimisation decisions can achieve a competitive edge, most important of which is seeing the bigger picture.

Adit Abhyankar is executive director of Visual IQ

Read More

Register for Newsletter

Group 4 Copy 3Created with Sketch.

Receive 3 newsletters per week

Group 3Created with Sketch.

Gain access to all Top500 research

Group 4Created with Sketch.

Personalise your experience on