Mapping value across the customer journey

More and more brands are using attribution modelling to find out the true value of every stage of a customer’s purchase journey. But as the picture gets clearer, new challenges are emerging – such as how to split the money fairly.


The online ad industry is well into its teens, and while search accounts for two-thirds of all online ad spend, many people believe that it’s time that other elements of the online purchasing funnel – display, social media, mobile – get the credit they deserve.

The dominance of Google, Yahoo!, Bing and the specialist search comparison sites such as is partly down to the industry’s continued reliance on the ’last click wins’ model.

This set-up rewards the place that was the final port of call for a consumer before a transaction or other action was made with a particular brand. So, a retailer would credit a search engine or price comparison site if it was the direct referrer of a paying customer.

Coupled with the fact that so much traffic begins at and goes through search and comparison engines, it’s no wonder most adspend is distributed in their direction. Increasingly, though, the industry is looking away from the last click model to try to ascertain the true value of each stage in the consumer’s journey.

Sky senior online marketing manager Helen Southgate explains: “At present, the last click wins model is still the most commonly used method to attribute a sale to a channel, and when paying commission on a sale this is still the most suitable model.”

“However, it is important for an advertiser to understand that the channel that was awarded the sale under the last click model is unlikely to be wholly responsible for that sale,” says Southgate. “The customer is likely to have interacted with several online channels, and offline channels.”


Virgin Atlantic has been looking to attribute value across all its online marketing channels. The brand has been working with digital tagging company Tagman to track all its online marketing activity from one piece of code. This is enabling it to identify the sites that are providing the most value.

Allison Wightman, Virgin Atlantic head of marketing systems, says this approach might not necessarily please everyone – especially affiliates, which would often get the lion’s share of credit. But she said the majority have welcomed a more balanced approach.

“A couple of years ago, last click wins was unpalatable for affiliates particularly, but our networks have been supportive in us implementing a clear branding and commission policy,” says Wightman.

“Price comparison engines dominate in last click wins, as the user usually checks where is cheapest for the flight they want as their final check. This is really unfair for content partners who invest significant effort in delivering a richer offering for the user who is in research mode.”

Stripping it down even further is Alex Tait, head of digital marketing at the Post Office and chair of ISBA’s Digital Action Group. He says that the reliance on a “last click” perhaps jeopardises other effective forms of online marketing.
“Imagine a marketing manager looking at a media plan: they’re looking at different media alongside each other, like affiliate, search, behavioural targeting and display,” says Tait. “They might look down that plan and see one as being more expensive – such as display – and so remove it in favour of what they see as driving that conversion without really understanding how it all fits together.

“But if you remove the display, you’re actually going to be impacting the overall performance, and the overall cost-per-sale. You’re not understanding how media works,” he says.


In truth, the drive to look beyond the last click towards true attribution modelling is nothing new. Industry trade bodies including the Internet Advertising Bureau (IAB) have long championed online as a platform that goes beyond just direct response.

The main online players have also made their own attempts to help brands look beyond the last click. In early 2008, both Google and Microsoft launched separate strategies to show advertisers the true value of each element of every stage of the customer journey.

Microsoft’s Engagement Mapping was developed by its technology arm Atlas to provide a scientifically based standard showing how different media affect an eventual conversion.

At launch, Ciaran McConaghy, Atlas analytics manager for EMEA, says: “The act of giving a CPA (cost per acquisition) credit is actually based on about 10% of user activity. If an advertiser is spending a reasonable sum of money, there should be a better standard to measure the exact contribution of a media channel.”

Google’s research has included working with companies including Nielsen and search specialists such as tdSearch to look at the duration and type of journeys consumers make within certain verticals (see box above). The results show that consumers who spend more time researching across multiple sites and search multiple times are of more value to brands.
Once a clearer picture of the full journey emerges, so do new problems – such as how to weight credit. Is it 50% to each stage on a two-step process, 33% on a three-stage, and so on, or does the brand give more weight to some areas, such as display, than to others?

Virgin Atlantic’s Wightman says the company is in the process of developing an attribution model for its global marketing activity.

“We think it will be groundbreaking within the industry. ’Best click’ would be based on revenue value – we hope we get to a point where we can split the revenue by the channels which deliver the most in terms of real value,” she says. “So it could be that we end up with something like: 10% for the first click and 60% for the last click, then the 30% in the middle gets split evenly across all the participants in the path.

“But we need to develop a model alongside the capabilities of the affiliate networks and publishers, as even though the stepped model works best for Virgin Atlantic, it may be impossible to make it work in real life for certain publishers like cashback and voucher sites,” she says.

Virgin Atlantic already has technology to allow it to do this, but the Post Office’s Tait says other companies are not as quick to act.

“The industry’s answer is attribution modelling – but more people talk about it than actually use it,” he says. “Part of the reason is implementation. Upgrading your tagging can take up to six months for some businesses.”

Help is also coming from industry initiatives. Last year saw the launch of UKOM, a common planning currency for online, developed by the IAB and Nielsen to help more brands invest in the web, especially for branding. Likewise, the tail-end of 2010 saw the Association of Online Publishers club together to provide advertisers with a metric for online engagement.
These will help brands understand the full value of each aspect of their online marketing, allowing them to attribute commission or increase investment in those areas that perhaps were performing well without the recognition.
“The key is understanding the role of each channel at each stage of the customer journey and being able to look past the last click to the full path to sale,” says Sky’s Southgate. “The real marketing then comes with tweaking each channel’s advertising strategy to the customer as they interact with several channels throughout their purchasing decision.”



Jonathan Beeston
Client services director (Europe)
Efficient Frontier


Every advertiser wants to understand how their advertising is working, and how different media working together can be greater than the sum of their parts, so that 1+1=3.

So advertisers want to see the whole customer journey, and technology is now at the point where we can start to see all the touchpoints that lead up to a conversion. The fall-out from that is that you can start to make decisions about where to spend your money most effectively.

The first part of the solution is tracking the customer across all online channels, to understand the path they take before purchase. The second part of the solution is doing something meaningful with that information – changing what you are doing on the basis of that analysis.

Monthly analysis isn’t enough to let you respond quickly to user behaviour and optimise a campaign. To be effective, optimisation needs to be done constantly; the only way to achieve this is to automate it. As more and more media are bought by auction, including display and social as well as search, the ability to optimise in real time becomes more important.

A great example is search retargeting. You can follow up a search campaign with a retargeting programme to re-engage those people who didn’t convert as a result of the first campaign. However, it is possible that some of those subsequent conversions would have converted anyway, so it is important to know how many you have genuinely re-engaged. It would be impossible to manually research the incremental uplift for each tactical change you make.

The only solution is an algorithmic approach that can do the calculations for you, rather than you sprinkling the credit along the breadcrumb trail, as some companies are doing at the moment. That is not the wrong thing to do; it is just an early staging post on the way to a more robust way of assigning credit.

We are now at a turning point for attribution modelling. The technology is there. Our universal tracking solution and optimisation algorithms allow us to help our clients move budget instantly between marketing tactics. At every stage of the sales funnel, the optimum mix of media can get campaigns performing more effectively.



    Leave a comment