Tactics > customer recruitment
Understanding the customer journey
By Tim Pemberton
The case for attribution modelling
T
hese days, when I talk to clients, there are two main issues that
crop up time and again: “How come your sales results aren’t closer to ours?” and “Why should we spend any money on online display
advertising when it doesn’t give us the ROI we need.” Getting to the bottom of both these queries is relatively easy but requires some work from both client and agency. The key to these two problems comes in the analytics and understanding how the data is “counted”, or often, “double counted”.
How do we get around the differing results issues? Clients tend to forget that there are numerous different digital and offline channels all promoting brands and products at the same time. Online retailers can have press advertising, catalogues, inserts, email, paid search, display advertising and SEO campaigns running simultaneously. Certain tools are good at tracking some data. For agencies, the default ad server, DoubleClick, can be used to net out sales from bought media, such as display, search and affiliates, but it becomes a blunt instrument when email and social media are added into the mix. Both email and social media are tracked and reported using different tools. The best solution we have found is to use one tool that tracks all the different digital elements and can net out the sale across all the channels. It sounds easy, and it is easy, but it requires additional coding onto the site and yet more reporting in the short-run— approximately three to four months—whilst all the internal and external stakeholders become comfortable with the results.
The second question that we always get
asked by our clients is around investing in online display advertising when it doesn’t seem
24
to show the required return on investment. The real question here is that display doesn’t work with the “last click wins” model, even with a 30-day cookie window and post-impression sales being tracked. This is often true in a linear world where you are saying that the consumer decision-making process is really simple—one
attributing model in place and be comfortable with the results.”
click and a sale is recorded. We know this not to be true. There is a library of research in the offline world about the decision-making process for consumers and we have lots of simple models to help describe this process, such as AIDA (Attention, Interest, Desire, Action), for example. We know that it can take up to six months to purchase a new car and a consumer goes through a whole raft of information while researching his purchase; he asks for opinions and reads up on the subject before creating a wish list for his chosen vehicle. Some decision-making is simpler and may take 24 hours—replacing a broken washing machine, for instance. Some purchases take longer—up to three weeks for clothing and gadgets. There are now a growing number of tools in the online world that are allowing us to track the consumer decision-making process and in doing so allowing us to reattribute the sale not just to the “last click” channel—usually a brand search term—but to all channels and advertising exposure that a consumer sees prior to making the sale. Take a simple example of this: a consumer might see a display advert, not click on the ad, see a generic keyword, click on it and go to the site but leave without making a purchase. A few days later he sees an affiliate advert, clicks on the ad, but again does not make a decision. He receives an email, opens it but does not click through; uses Google and enters the site through organic search; does
Direct Commerce Catalogue e-business
www.catalog-biz.com
“The challenge is to get the
some more research; reads some comments on a blog and posts some questions; sees a display ad, clicks and revisits the website and finally decides to make a purchase by using a branded paid-search keyword. Now that’s quite a customer journey and it’s
only the online element of the journey. The usual “last click” model would attribute the sale and the revenue to the branded paid search. The new approach is to weight the effect of each channel and the overall communication— bought, earned and owned—for the sale, or a collection of sales, to a combination of channels. This creates some really interesting
marketing communication challenges. The reallocation of part of the sales to different channels throws up all sorts of results. The sales revenue of paid search can go down, while display goes up—and becomes a profitable channel against the ROI targets. Affiliates need to be managed as “paying for a sale” that they might have claimed may not be totally true; they contributed to part of the sale. Most interestingly, it also throws up real challenges internally where the SEO and website’s claim on sales can sometimes change dramatically. The effect of using the attribution, or
exposure to conversion, model is that it shows a more holistic effect of online on a sale, measures both bought (paid search and display), earned (social media) and owned (website) media on your sales. The challenge is being able to get the attribution model in place and being comfortable with what the results mean to the online communications budget. The benefit is to be able to net out sales and start to reallocate the sales against the different digital channels and the financial investments made in them.
Tim Pemberton is a director at The Digital House, a media agency that manages paid search, SEO, display advertising, affiliates, cold email, list buying and analytics. The Digital House is part of the Specialist Works group of companies—a CatEx Direct Commerce Association gold sponsor.
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52