Attribution-model-emarketerAttribution modeling: it’s one of the biggest challenges marketers face in assessing the value of a marketing strategy and connecting the dots across channels. In ad tech, advertisers can leverage a variety of attribution models, but the two most common are last-touch and multi-touch. Which approach is best? What are the pros and cons of each?

The deal with last-touch:
A last-touch attribution model gives 100% of the credit for conversion to the very last thing the customer did before the purchase.

Last-touch attribution: pros

  • Proximity. Advertisers know exactly which clicks and views directly preceded a conversion.
  • Low data requirements. Since you’re only tracking one touch point, you don’t need an elaborate technical setup.
  • Easier mobile attribution. Several mobile networks record only clicks, not impressions. In this case, last-touch may be the simplest model to get an apples-to-apples comparison across channels and campaigns.

Last-touch attribution: cons

  • Bias against upper-funnel tactics. When you only track the last click or view, you don’t know how the customer found you or how many times they interacted with your brand before converting. This could inadvertently move budget away from channels that help drive customers to discover and engage with your brand.
  • Poor cross-channel tracking. No tactic acts alone; a recent Stanford study discovered that people who saw a display ad campaign made 5–25% more campaign-related searches than people who didn’t see the campaign. Last-touch attribution models may keep you from correctly assessing the value of channel and campaign overlap.

The deal with multi-touch
“Multi-touch” attribution models look at multiple touch points along the customer journey. These can include linear, time decay, and custom models.

Multi-touch attribution: pros

  • More data. You get more information on the customer, from discovery to conversion – which means greater insight on what is most relevant users at each stage of the buy cycle.
  • More effective budget investments. A fuller picture of how your marketing tactics are performing will better inform how to allocate spend across upper and lower-funnel efforts.
  • Cross-channel lift. Display ads running in combination with search ads have demonstrated a 48% lower CPA (cost per action) compared to search ads alone. Your various marketing channels impact each other, and multi-touch attribution can capture that data for more accurate performance tracking.

Multi-touch attribution: cons

  • Complexity. Multi-touch attribution gives you a lot of information, but it can be difficult to separate the signal from the noise. Make sure you have resources for data analysis and a game plan for implementing what you learn.
  • Risk of overvaluing non-conversion actions. Touch points that don’t lead directly to conversions (for example, branding campaigns) have value, but how much? It can be challenging to assign appropriate value. Tactics like incrementality testing will help you optimize your multi-touch attribution model.

Google Analytics pioneered the last-click attribution model, which has been the default attribution setting for marketers for some time. Recently, however, there’s been an industry-wide shift in the way marketers view attribution. As the customer journey continues to become increasingly complex and multifaceted, marketers are starting to experiment with new ways of approaching multi-touch.

Want to learn more? Download our research report Attribution for the Data-Driven Marketer.