Picture: 123RF/rawpixel
Picture: 123RF/rawpixel

The effective tracking of digital marketing campaigns is not as easy as initially thought. One question that often pops up is: at what point did consumer conversion take place? During a campaign it’s difficult to attribute this correctly, especially with multiple touchpoints such as display and native advertising, sponsored social posts, video ads on mobile or paid-for search results all running consecutively.

Simply attributing a successful conversion to the consumer’s first or last point of contact on a campaign doesn’t tell the complete story. It’s here that multi-touch attribution (MTA) comes into play.

Linking users with touchpoints

Joel Rubinson, an international expert on the subject of MTA, believes that by using it marketers can find out which of their digital marketing tactics are working and which are not, and that thereby they are enabled to fine-tune campaigns and increase return on investment (ROI).

Simply put, MTA tries to connect user data across multiple campaign touchpoints. It tracks patterns of ad serving and content consumption on a user level, analysing campaign data as it’s unfolding. Doing so provides marketers with a granular view of a campaign’s various hits and misses. “I’ve seen as much as a 30-to-1 difference around how effective different creatives or methods of delivering advertising are. Knowing this [enables you to] do more of what works really well or to deploy money away from what isn’t,” says Rubinson.

Identity resolution the bugbear

Effective MTA is not a one-man job, nor a quick one. For brands, the linking of user data across silos – identity resolution – will be the biggest stumbling block. It involves connecting the dots between data gleaned from browser cookies and tracking pixels, known first-party data and customer relationship management software. It’s done to figure out who was exposed to what (and when did they click) during a digital marketing campaign. As can be expected, identity resolution will require the help of the media agency and ad tech firms who offer relevant services as well as media partners such as Google.

Standing in the way of competent identity resolution are challenges such as the walled gardens created by large publishers. These severely limit the data returned to clients after campaigns and often don’t allow pixel tracking nor the sharing of specialised ad products. Even when usable data is provided it can’t be linked to other walled-garden data sets. 

Online retailers such as Amazon that allow brands to advertise on its properties don’t share sales data at a user level, making it difficult to determine the outcome of advertising. And if this isn’t making identity resolution difficult enough, SA companies are hampered by the unavailability here of LiveRamp, one of the largest identity resolution companies in the US.

Notable workarounds

However, there are workarounds. The Mobile Marketing Association (MMA) believes brands should consider going all in with strategic publishing partners. The emergence of data clean rooms, where aggregated data from publishers is mixed with first-party data from advertisers, can produce more complete media-effectiveness analysis and attribution. Google’s Ads Data Hub is a good example of this, providing advertisers with campaign data such as cookies, mobile device IDs and anonymised Google sign-ins, to be combined with data of their own. 

Understandably, brands that have talked to the MMA are reluctant to put all their advertising eggs in one basket with only one publisher; while sharing their own first-party data with publishers also raises privacy concerns.

Set out on the right MTA path

It might initially be difficult to see the wood for the trees, but the MTA process does get easier. “The data-readiness part of MTA is arduous, but once this [has been seen to] the analysis part does not take long. For companies that have this infrastructure in place, future projects can be turned around much quicker,” believes Rubinson.

The five stages that MTA projects need to move through for it to be successful, he says, are briefly:

  1. Initiating the MTA process, which involves identifying an executive sponsor and establishing a project leader;
  2. Establishing data readiness, which incorporates understanding the digital data ecosystem, identifying conversion data and making a deep review of linkable data;
  3. The setting up of the first MTA project, which requires choosing the campaign test case and preparing data for the first test;
  4. Implementing the project, which involves managing the project plan and quantifying the lift in marketing ROI; and
  5. The full deployment of MTA, which includes a phased roll-out plan and the expansion of available data.

No shortcuts to success

While the full roadmap Rubinson presents is far broader in scope, he says it’s important to move through each stage deliberately without trying to find shortcuts. “It’s good to start with the end state in mind, but you have to proceed through the journey map methodically, marking each step complete as you go.”

Finally, Rubinson points out that, despite the benefits involved with MTA, it should not be seen by all marketers as a full-on replacement of other marketing mix models. “Marketers should look at MTA as a critical part of a portfolio of models, not the silver bullet. It will help them move from okay ROI to superior ROI by focusing on incremental sales improvements.”

Sarah Utermark is the country director of MMA SA. The organisation will, as an ongoing project, be tracking local MTA progression and benchmarking results against global business figures. In order to take measurement forward, local companies are invited to join the MMA’s Marketing Attribution Think Tank (MATT), which will assist marketers to apply MTA with confidence.

The big take-out

MTA allows marketers to find out which of their digital marketing tactics are working and which are not.