Bad science: common pitfalls in attribution and digital planning

November 03, 2015
LRegan
5 min read
Beginner
Stacked electronic devices and a book showing ports and connectors.

The digital marketing industry continues to struggle with its interpretation of ‘attribution’, and how to achieve it. In its most basic sense, attribution means determining the role that any given channel plays in informing and influencing the customer journey. But in an increasingly complex and fragmented marketing landscape where the consumer journey flits between a tangled web of digital, offline, social, paid and earned media channels, quantifying the contribution of individual touch points is increasingly tricky. Too often, marketers exercise poor execution in their attribution strategy, and as a result, they rely on the old adage that “half my marketing works, I just don’t know which half”! Consequently, marketing spend, and the overall sales cycle, is not optimised nearly as much as it could be. There are two huge challenges facing marketers: the extreme difficulty in tracking cross-device behaviour, and understanding the genuine incremental contribution of a channel or campaign. Sophia Evgeniou, Head of Customer Acquisition at House of Fraser, recently told Econsultancy: “Attribution is still a difficult topic for many companies as there isn’t one set way to attribute and this can be an extremely unsettling discussion point for any business.” Let’s examine three of the common pitfalls that marketers face, in the hope that we can shed some light on them and offer up solutions…

  1. Cross-device attribution

study by Facebook revealed that over 60% of online adults in the US use at least two mobile devices every day, with close to one quarter using three devices. Switching devices is also a significant part of the purchase journey. More than 40% of online adults sometimes start an activity on one device only to finish it on another. These trends present new opportunities for cross-device attribution and open up significant challenges in understanding consumer behaviour and usage patterns.

On the face of it, smartphone traffic converts poorly and cross-device attribution remains a pipedream for many European advertisers. Clearly, the savvy marketers out there know this is a misnomer. A recent study by comScore showed that 80% of local searches on mobile phones convert into sales. This makes sense in light of the top searches cited in the report: restaurants (23%), auto service establishments/dealerships (10%) and arts and entertainment (9%). From an ecommerce perspective, those who have the ability to track their logged-in users can understand cross-device behaviour, and hence devise an attribution model that accounts for the impact of mobile. For everybody else, there is external help at hand. Some vendors have devised ways to help marketers with cross-device attribution and Google Adwords’ Attribution reports and metric ‘cross device conversions’ is a good place to start.

  1. Correlation isn’t enough

Too often marketers invest in a channel which appears to convert to revenue, but for which they don’t have any measurable proof. Relying on opinion is risky business. While 91% of marketers say they understand the value of attribution, many struggle to justify spend on attribution technology. But the truth is, you don’t need to collect lots of data and construct the perfect attribution algorithm in order to measure the value of one channel or touch-point. If we think back to how medicine has dealt with the challenge of attribution, it’s used randomised control trials whereby double-blind studies are carried out with some participants receiving a placebo and others a treatment. Neither the doctor nor the patient, know which is which. How does this apply to marketing? One possibility is to use geo-targeting to test one batch of locations versus another for specific channels. If we extend the analogy this means one set of towns receive the ‘treatment’ while another receive the ‘placebo’, i.e. are excluded from the targeting of this marketing channel. Such testing can help to shed light on whether correlation is in fact converting to actual, incremental revenue. Another way to use this in display advertising is to run an A/B test showing half your participants a charity banner and half your brand banner. To illustrate the opportunity in this area, 90% of marketers who are properly measuring attribution say they are seeing significant benefits, while a fifth of companies are planning to increase their attribution technology spending during 2015 (according to Econsultancy’s Marketing Budgets 2015 Report).

  1. Be wary of agencies, publishers or channels that fail in areas of transparency

Every agency will tell you that they are media neutral when it comes to planning, but the reality is often very different. Publishers and media owners offer incentives and commission, and seek to build relationships by media planners and buyers. If an agency is part of your digital planning, be aware that your media plan may be biased towards the incentives on offer (although some agencies do pass this discount back to their clients). If you choose to get third parties involved in attribution, tread carefully with those that fail to offer transparency, or the granular reporting you require. It is, after all, in their best interests to look like they’re providing positive results, especially if they are earning commission. In some scenarios is might become a case of agencies and publishers ‘marking their own homework’, so don’t rely solely on your supplier to analyse the data for you. Ensure that someone in your marketing team understands the data just as well as they do.

Contributing Experts

LRegan

Mentioned in this article

Explore more insights