According to an eMarketer survey, more than six in ten executives globally said that achieving a more complete view of the customer helped them more accurately predict their needs. So it’s not surprising that our RampUp Virtual Summit session titled Multi-Touch Attribution across Facebook, Digital, and Beyond, was packed (virtually). The webinar revealed how this advanced form of measurement is helping to create a unified omnichannel view of consumers that 81% of marketers find challenging to implement, as Experian found in its study of 1,100 global marketers.
During the webinar, we received many great questions about attribution technology and methodologies for presenter Wayne St. Amand, CMO at Visual IQ, a provider of marketing intelligence.
Here are his responses to some of the questions he wasn’t able to get to:
Can you walk us through some multi-touch attribution methodologies?
While there are many different types of multi-touch attribution (MTA) models, the common denominator is that each tracks the consumer journey and de-duplicates it across channels and tactics to assign credit for a KPI event (e.g., conversion, lead, etc.) to one or more touchpoints. However, the level of sophistication of each model can differ dramatically:
- A subjective rules-based methodology relies on humans to define the rules of how credit is allocated to one or more points in the consumer journey. These models can be used to better understand the consumer journey. Examples of rules-based approaches include last touch, first touch, even weighting, position-based, and time decay.
- An algorithmic methodology is based in objective, statistical modeling and machine-learning techniques. The output of an algorithmic model can be used to predict outcomes to help marketers plan or optimize future marketing efforts.
But even algorithmic models vary in terms of their modeling sophistication and granularity of predictions. The most sophisticated algorithmic MTA approaches can evaluate every dimension of every touchpoint before assigning fractional credit. This means marketers can understand optimal budget allocation at the channel and subchannel levels and also at the most granular levels like keyword, placement, and creative.
As marketers are always running test campaigns on multiple channels, how do marketing attribution technologies take into account consumers’ differing exposures on these channels to uncover the combination of marketing tactics delivering the greatest results?
Algorithmic multi-touch attribution approaches use the full dataset of both converters and non-converters to perform billions of A/B comparisons between those exposed and those not exposed to different dimensions of marketing and media.
The result is a breakdown of every single touchpoint at the most granular level (e.g. ad size, placement, publisher, creative, offer, etc.) This in-depth understanding of how each touchpoint and dimension changes the propensity of a consumer to take a desired action enables marketers to make more specific and impactful optimization decisions.
When running cross-platform campaigns, we can’t properly keep a true control audience, as the reality is the control group may be exposed on Facebook in-app (where brand partners aren’t given line of sight). Are there any new developments coming there to put brands at ease as they distribute cross-channel campaigns?
The fundamental idea of a controlled experiment is to compare data groups that are virtually identical. In the controlled mode, one variable is changed (test group) while the other (control group) is left undisturbed. The results obtained from the test group are then compared to the control group to measure the impact of that change.
The challenge is that most marketers aren’t looking at a single data point like creative message, but rather at data from multiple sources. The sheer volume of data makes it virtually impossible to get a holistic view of marketing performance without time consuming manual data collection and manipulation.
That’s where multi-touch attribution come in. It does the hard work for you. By integrating dozens of channels and multiple terabytes of data, multi-touch attribution enables marketers to compare the performance of every channel and tactic across their entire marketing mix.
Moreover, more sophisticated attribution technology will use test and control groups to verify the accuracy of the measurement and optimization recommendations it delivers by comparing the predicted results with the actual results. If this validation process identifies an unacceptable margin of error, the algorithms will be refined and updated to ensure predictions of future marketing performance are more precise.
The recent offline purchasers segment is very interesting to us as a retail brand. Is there a way to understand/get information from the customers as to why they are purchasing offline rather than online?
With consumers exposed to hundreds, if not thousands of messages a day, experience itself is proving to be the primary driver of competitive advantage. Marketing relevance is key for attracting new customers and building more valuable relationships with existing customers.
While MTA doesn’t necessarily enable marketers to predict consumer behavior, it gives them much more clarity into the marketing messages and tactics that influence and are relevant to a specific audience. Marketers can then leverage this granular insight to test, iterate, and refine their techniques for each audience to drive the online and offline KPIs they care about most.
In case you missed it, read Wayne’s first RampUp post, Choose Your Own Marketing Attribution Adventure, or listen to his RampUp Virtual Summit session below: