As privacy changes lead to signal loss, digital advertising is diverging from user-level data and specific measurement methods like multi-touch attribution (MTA) and media mix modeling (MMM) in favor of bigger-picture analysis of marketing performance and a multi-methodology approach.
Relying solely on MTA and MMM is not enough. There’s also a third method: incrementality. This measurement approach uncovers how marketing activities drive KPIs relative to what would’ve happened without those activities. And with advances in artificial intelligence and machine learning, incrementality measurement becomes more accessible and precise, leaving its A/B testing days behind.
To explain that distinction and incrementality’s role in the privacy-era marketer’s measurement toolkit, let’s break down how incrementality complements MTA and MMM, what AI-enabled incrementality entails and how it will position marketers to succeed.
The measurement trifecta
MTA, incrementality and MMM are all key to measurement.
Attribution is most adept at providing a short-term view of marketing performance. A mobile marketer can use it to see the app installs a campaign generates. Multi-touch attribution goes a layer deeper than last-touch attribution, showing not just which channel contributed last to conversion but all the touch points along the customer journey. MTA is still useful, but it’s becoming more restrained with privacy changes from Google and Apple.
Then incrementality comes in, giving marketers the confidence to run new campaign types with new channels in new markets.
Let’s say you’re an ecommerce app marketer with a strong presence in the US looking to expand into Brazil. With attribution, you will immediately see installs. But you won’t know whether these installs would’ve happened organically or how your performance stacks up to competitors. With incrementality, you can compare your likely campaign outcomes to similar ecommerce apps and evaluate incremental lift.
Plus, the increasing power and availability of artificial intelligence (AI) and machine learning (ML) are enhancing incrementality even further. Traditionally, incrementality relies on A/B testing where one group is exposed to a campaign and another isn’t. An AI-based approach relieves marketers from the tedious process of creating potentially flawed control groups.
Instead, they can now analyze a vast amount of historical data for campaigns similar to theirs to gauge incremental lift. This makes the process more precise and efficient. A Fortune 500 brand with a large, dedicated data science team can spend weeks comparing campaign results to historical data, assessing patterns in conversions and ad spend. Or AI can do it in minutes.
Finally, there’s MMM, which will be most helpful for marketers looking for strategic budget allocation, in addition to those capable of taking a long-term view of performance. Through MMM, marketers can determine their daily optimal spend allocation across channels based on their budget and campaign goals. MMM also allows marketers to assess the correlation between channel investments and performance over, say, a six-month period to inform recalibration of the marketing mix.
Making the case for multiple measurement approaches
Marketers are being asked to do more with less as fears of a recession persist and layoffs strike left and right, especially in tech. The ones thriving in this privacy-centric ecosystem are embracing changes and adopting innovative measurement approaches.
Today’s growth marketers need to use all three measurement models when making decisions. MTA provides the short-term view. (Where are we immediately driving results?) Incrementality provides the mid-term view. (Are we getting the most bang for our buck relative to how we’d be performing over the same timeframe without marketing?) And MMM informs strategic budget allocation based on a long-term analysis. (How should our media mix be adjusted to drive outcomes while taking into account factors such as seasonality?)
Of course, no single measurement approach will rule the day. But marketers grappling with privacy changes need more than the short-term, user-level-data-prone view that MTA provides. With the long-term analysis provided by MMM and AI-driven incrementality, marketers can round out their measurement approaches and answer the biggest question on their minds: Which of my marketing tactics is really delivering the maximum bang for buck?
“Data-Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media.
Follow Adjust and AdExchanger on LinkedIn.
For more articles featuring Katie Madding, click here.
