In this article, we wanted to shed light on what are the differences between the three methods of measurement: Last Touch (or Last Click) attribution, Media Mix Modeling (or Marketing Mix Modeling, or MMM), Incrementality measurement
The three different methodologies are not “better” or “worse” than one another. It is not about you as a marketer making a choice “which attribution methodology should be used?”
All three attribution models may be relevant to your marketing activities.
We wanted to provide this short, but comprehensive guide to marketing attribution.
Starting with a comparison of the methodologies:
Last Touch Attribution
Last Touch attribution works only for Digital marketing where the Advertiser is able to assign a dedicated URL for consumers to use to reach the conversion point.
I.e. rather than going to www.incrmntal.com - the user will click a URL with various parameters allowing the Advertiser to attribute the conversions to the relevant vendor or other more granular parameters such as campaign, creative and so on.
For example: https://www.incrmntal.com/?utm_source=vendor&utm_medium=mobileapp&utm_campaign=businesscampaign
As we cannot (and should not) expect consumers to type in a long URL with various parameters presented on a TV campaign or billboard - Last touch attribution is limited to Digital Media.
Some Advertisers try and bridge this gap by using “promo codes”, but as there is no way to tell if a user saw an offline ad and chose to use or not use the promo code - this approach is not very accurate.
Last touch attribution's biggest benefit is that it allows a digital marketer to associate the last ad the user engaged with before the conversion point - in real time.
There’s no need for any historical data , as the attribution happens based on a match between the click and the conversion.
For any advertiser spending a large sum of money for digital advertising - real time conversion data can provide a critical feedback loop for optimization. Last touch attribution can act as a great proxy to the performance of creatives or audience segments.
With recent changes in mobile operating systems and browsers - user level identification is becoming impossible and moving to a more aggregate cohort level.
While this move makes conversion level matching impossible - the mechanics of this method remain the same where attribution happens based on “last touch” , while the reporting of attributed conversions would be reported in aggregate form rather than a single conversion.
Over attribution is extremely common with Last Touch attribution. As a product becomes more known and popular, consumers search and interact with it more and more, leading to a point where consumers may unknowingly and worse off - involuntarily engage with an ad on their way to the conversion point, causing the attribution platform to credit a vendor for the conversion, while the user intent was to convert organically.
Last Touch attribution is susceptible to “attribution gaming” as fraudulent publishers discovered that if they trigger an ad engagement, with or without the user knowledge before the user reaches the conversion point - they can claim that the conversion was a result of their activity, thus, gaining budget or higher prices from the Advertiser.
Media Mix Modeling
Media Mix Modeling or Marketing Mix Modeling or in short: MMM is a statistical method to estimate the impact of various marketing tactics on sales in order to better forecast and come up with a better marketing strategy.
The method was developed in econometrics for the consumer packaged goods industry and has become common with brand cross platform Advertisers in the last years.
Media Mix Models require historical data to have any helpful outputs. Often, the data needs to include external influencing factors such as competitors activity, product launches, financial events, weather and any major event that may have influenced the performance of a product (i.e. during an Olympics year, more people buy sport goods).
Due to these requirements - Media Mix Models work best for refining a strategy, expecting influencing factors such as changes in the media mix and/or external factors to help understand what would be the best media mix to market a product over time.
Media Mix does not work well for new product launches, as without historical data - there are simply too many unknown variables.
B2B2C brands (i.e. consumer goods, retail, consumer electronics) often must rely on Media Mix Modeling to analyze the effectiveness and impact of their paid marketing activities.
Attempts to use a simple version of Media Mix Modeling are tested regularly by Advertising being turned off completely , allowing Advertisers to analyze sales activities with no Advertising in place - however, switching off all Advertising is extremely hurtful to most Advertisers as while doing so - competition may consume market share, thus, hurting long term brand equity for an Advertiser experimenting with turning off the lights across all marketing activities.
Incrementality measurement a method to understand the true value of Advertising spend. It is more operational and tactical than Media Mix modeling , but is not as granular nor real time as Last Touch attribution.
Incrementality measurement does not assume to replace attribution. Incrementality measurement relies on last-touch attribution data to indicate if paid and/or attributable conversions had any value to the Advertiser or if the attributable conversions are cannibalizing the organic new user base or the user base arriving from other paid media sources.
Incrementality attribution can provide insights in granular levels over campaigns, demographics, vendors, geo location, contextual features. These insights can be used in an operational - tactical level.
As this form of measurement operates simultaneously to attribution, Advertisers can forward or stream conversions and ad spend data from any platform used to understand the true value of their spend, including non digital mediums such as TV , Radio and obviously - Digital.
Incrementality measurement requires a heavy investment into technology, utilizing causal inference algorithms, difference in different techniques, concepts from game theory, market seasonality data, disrupted time series and machine learnings to come up with the right hyperparameters to provide actionable insights for the Advertiser to consider.
A common misunderstanding of incrementality is the utilization of “incrementality testing” only for reattribution (or retargeting) activity, where media vendors use a randomized control group and serve those with ghost ads showing that the segment being served normal ads does perform better. This type of testing still utilizes Last-Touch attribution , thus making the result of the test skewed.
- (no comments)