The majority of marketing attribution approaches are in need of a revamp. Tiffany Carpenter, Head of Customer Intelligence, SAS UK & Ireland addresses the different approaches to marketing attribution and the need for breaking away from these traditional approaches through implementing predictive and prescriptive analytics.
Back in the early 1900s, Philadelphia retailer John Wanamaker supposedly said, “Half the money I spend on advertising is wasted; the trouble is I don’t know which half.”
Wanamaker’s dilemma was that although he knew that traditional advertising of the time – newspapers, billboards, posters and the like – resulted in increased sales, it was never clear which media and the times they were posted worked best and which were not contributing value.
Fast forward 100 years and with the explosion of digital channels and the increasingly complex nature of customer journeys, marketers are still faced with the same dilemma when considering how best to reach consumers and how to best invest and optimise precious marketing budgets. In our research earlier this year, 25 per cent of marketers said they had no ability at all to measure the success of their marketing channels and campaigns.
What is Marketing Attribution and why it matters?
If Wannamaker were alive today, he would most likely have employed marketing attribution analysis to help solve the challenge. Marketing attribution refers to how marketers assess the value or ROI from the various channels that connect them to potential customers. In other words, it’s how the customer came to know and buy a product or service and assigning credit to the relevant marketing touchpoints.
The goal of marketing attribution is to produce more revenue by optimising your spending on the right channels. Knowing which channels drove the conversions also improves your understanding of the customer journey, and enables you to enhance and adjust your marketing strategy accordingly.
There are many different approaches to marketing attribution, but they broadly fall into two main categories – rules-based models and algorithmic models.
Historically the most common rules-based method of attribution was based on ‘last click’ or ‘first click’, which means giving credit to the last or first channel that the user clicked on before “converting” (or making a purchase). Today’s customer journeys span multiple channels, devices and time – and for this reason single-source attribution, where only one channel is credited, has limited value.
Multi-touch attribution gives “credit” to all the channels that are involved in a customer journey. How much credit is dependent on the model you choose. A linear model for example, shares all the credit equally among each channel, a position-based model puts more weight on the first and the last click, whereas a time decay model arbitrarily weights the channel based on the recency of the channel touches across the customer journey.
It’s time to break the rules
The common challenge with all the above approaches is a lack of objectivity. Deciding between different rules-based approaches often comes down to opinion and compromise. Given the many channel-related silos in most organisations, each team tends to want to choose the approach that gives more credit to the channel they own (for fear they will otherwise lose budget).
Marketers are increasingly under pressure to justify their budgets and ad spend but these subjective rules-based approaches to attribution lack the data-driven evidence they need to show business impact. Despite this, our research shows that more than half (53 per cent) of marketers who use attribution models are still reliant on a rules-based approach.
There is a real need to get much smarter about the way in which we measure attribution. We live in a world of big data that allows us to track marketing performance better than ever before, and approachable, scalable analytic capabilities are rapidly becoming part of every marketer’s toolkit.
Attribution is an ideal area to leverage advanced statistics and machine learning to more objectively determine the impact of each touchpoint along each customer journey.
Algorithmic attribution systematically analyses all the available data to determine the true impact of a given touchpoint on conversions, and analytical algorithms assign data-driven credit to each touchpoint.
Alongside a truer picture of channel value, this deeper level of understanding provides a starting point in progressing away from descriptive analytics towards the realm of predictive and prescriptive analytics.
Take dm-Drogerie Markt, a leading chain of retail stores across Europe, which is using SAS Customer Intelligence 360 to integrate all of their digital channels, combine online and offline data to get a complete view of customer activity and algorithmically assign credit to each touchpoint. Using this solution, they can make data-driven decisions about where best to invest their marketing budgets and how best to engage customers in the channels in which they choose to interact.
In today’s multichannel landscape, managing and allocating limited marketing budgets is an ongoing challenge for all organisations. Savvy marketers that can apply analytics and data science to the attribution challenge will find themselves at a significant advantage over those that continue to use a crude rules-based approach.
For more about the value of algorithmic attribution and how SAS can help read our white paper on Marketing attribution: Giving credit where credit is due.
By Tiffany Carpenter
Head of Customer Intelligence