It’s much easier to gather data than to interpret it
Of the multiple digital touch points that can influence the purchase of your brand, hand on heart, do you really know which five to buy? And do you know how the five you buy will affect the ones that you don’t?
Most marketers are up to their necks in data. Paid search, SEO and social, display, mobile, video and email, each of these channels produces streams and streams of potentially valuable data. Though the potential value of the data can only be realised through accurate interpretation. An accurate read can transform how to invest your digital budget – you’ll know which five touch points work best for your brand. Needless to say, misinterpretation can be expensive.
The more fragmented the digital landscape, the more marketers need to take charge of the ROI. With social, mobile, display and search all competing for budget, being able to demonstrate ROI across all campaigns is fundamentally important. Yet too many marketers lack a comprehensive understanding of how all these different channels impact their customers.
Holistic view
You see it’s not just the last click that needs to be counted in your ROI. It’s about taking a more holistic view of digital marketing spend, and giving credit to all the touch points along the way that helped to deliver the final purchase.
Let me explain. Imagine you’re a uni student and you’re browsing news websites. You see a compelling display ad on the SMH or The Age website: “Buy a Mac for uni and get $100 for apps, music and more”. While you don’t click through straightaway, you do think that as you need a new computer, it might be a good time to explore further. But you’re not ready to make a purchase there and then and you don’t even click through on the display ads. You think about this for a day or two and browse for other deals.
You click through and explore the other deals available and one or two take really your interest. But you don’t make a purchase there and then, you’re going to think about it and see what your friends say. After a day or so and after chatting with friends you decide to purchase the Mac with the $100 for apps deal. You fire up your browser and type “MacBook pro deals” into Google, click on the paid brand search ad and make the purchase on site.
False read
While the paid brand search ad is simply the endpoint before checkout, the ad receives 100% of the credit for the purchase. The display ad that generated your first level of interest does not receive even a modicum of credit. So you can see, as the Marketing Director of Apple, you may easily be led to think that your display ads aren’t working and you need to channel more budget into paid search. It’s a false read.
Internet ad tracking systems that wrongly give 100% credit for a transaction to the last clicked Internet ad misleads advertisers as to where to allocate Internet ad dollars.
The last click problem
The same is true in the B2B world where many different factors affect the buyer journey. So just a prospect clicks on your paid search campaign and then turns into a paying customer, does not mean you can attribute that success to your paid search campaign when your prospect has probably engaged with multiple marketing touch points before making the decision.
It’s known as the “last click” problem. You see this analysis doesn’t take into account any other paid media that drove the sale. With display ads, you’re not giving any credit to the impressions gained if you’re just measuring clicks. So rather than focussing on the last click, you need to fully understand and give credit to the value of the many digital touch points your customers experience on their path to purchase.
All touch points given a fraction of the credit
Attribution modelling helps marketers to measure channel return more effectively. Like a winning relay running team whose members are all given a gold medal rather than just the one who actually sprints over the line, attribution modelling captures all online media sources from the point where sales are first originated all the way to the final transaction. With attribution modelling, you aim to understand the relationship between different customer touch points, and allocate value accordingly.
You also need to think about incremental attribution, and assign weights to different messages. For example, you’d give a display ad less weight than a personalised email. The time the message was received relative to the eventual purchase should also go into the equation. Typically, you would assign a higher weight to a later message than an earlier one.
Messages are classified by the stage they support, and buyers are tracked as they move through the stages of a purchase funnel. Marketers use this structure to compare the effectiveness of different messages in moving buyers from one stage to the next. This lets them estimate the incremental impact on cost and revenue of spending on different messages/media, allowing a meaningful ROI calculation.
When all touch points are given a fraction of the credit, you can then determine a return on investment based on attributed revenue to spend ratio.
Although attribution modelling has been a hot issue for some time, I’m still seeing marketers struggle with the data and rely too heavily on last click analysis. The complex math can be a challenge but with increasing digital channel fragmentation, attribution modelling provides marketers with the only crisp measure that accounts for many digital touch points including social, mobile, display and paid search. It’s all you need to understand the most effective digital touch points for your brand.
This article was first published in B&T on 5 March 2012









