There are numerous reasons why Multi-Touch Attribution (MTA) is a reckless choice for today’s marketers. One of the biggest and riskiest is MTA’s dependence on unreliable and incomplete user-level data.
At a time of stringent privacy regulations, the phasing out of third-party cookies, and the proliferation of Walled Garden media platforms, MTA’s challenges with data collection far outweigh its utility – and without user-level tracking, MTA is basically dead.
Long considered one of the original (OG) attribution methods for marketers to gain (what used to be) an accurate and comprehensive view across various marketing touchpoints, MTA’s strength was its ability to address digital media at a granular level and reveal insights about particular users across their digital journeys.
This granular approach involved assigning fractional credit to each interaction a user had with a brand's marketing efforts, from initial ad impressions and social media engagements to website visits and email opens. To achieve this, MTA models employ sophisticated algorithms that analyze user-level data, including cookies and other tracking identifiers, that connect disparate touchpoints and construct a complete customer journey.
But in our consent-driven era, MTA has proven to be an inadequate methodology for today’s marketers as they navigate the complex, privacy-first, cross-channel media measurement and optimization world.
Here are six more of MTA’s dangerous drawbacks:
1. The Correlation Trap: MTA, like its cousin Media Mix Modeling (MMM), is entirely reliant on correlation-based data. That means it observes patterns in user behavior and infers relationships between marketing touchpoints and conversions. However, this reliance on correlation rather than causation is its greatest weakness. Just because a user interacts with a series of ads and then makes a purchase doesn't mean those ads caused the purchase. There could be numerous factors at play: organic search, word-of-mouth recommendations, pre-existing brand affinity, or even unrelated seasonal trends. MTA models, by their nature, struggle to disentangle these intertwined influences.
2. Addressability Gaps: MTA thrives on the granular, user-level data generated by addressable media, where individual interactions can be tracked and linked. This means it excels at collecting and analyzing individual, user-level conversion events to determine the impact each has on the customer journey. This focus on addressability creates a significant blind spot, and MTA struggles to account for non-addressable media such as traditional television, radio, print, and out-of-home advertising. These channels often play a vital role in brand building and influencing consumer behavior, but they lack the individual-level tracking capabilities necessary for MTA to succeed. In the privacy era, this limitation is particularly damaging for cross-channel media measurement and optimization.
3. Lower Funnel Death Spiral: MTA has always been incapable of optimizing beyond low funnel performance marketing, also known as the Lower Funnel Death Spiral. Marketers who rely on Google Analytics (GA) or MTA to inform their cross-channel investment decisions are particularly susceptible given that GA or MTA give them extremely limited visibility into upper-funnel media performance, which can result in misallocation of budget, a media mix low in incrementality that over-indexes on high-intent consumers, and a disproportionate focus on lower-funnel activities. MTA is a reckless attribution method for marketers who need a reliable and efficient way to maximize spend, measure, and optimize across the upper, mid, and lower funnel.
4. Fragmentation Fiasco: MTA requires a unified view of customer interactions across numerous channels and devices. However, data is often scattered across many different platforms and systems such as CRM databases, website analytics, social media platforms, advertising networks, email marketing tools, and more. With each platform using its own unique data formats, tracking methodologies, and identifiers, this creates inconsistencies and incompatibilities, and integrating and cleaning this data is a major hurdle if not an impossibility.
5. Temporal Lags: MTA strives to connect marketing touchpoints with eventual conversions, but it often operates within a framework that struggles to account for the inherent delays in the customer journey. This is especially true in industries involving high-value purchases, complex decision-making, or extended consideration periods. This time lag between initial exposure and final conversion can stretch over weeks, months, or even years and creates a significant attribution challenge for marketers. Additionally, rapidly changing market conditions can render attribution models obsolete, requiring frequent updates and adjustments.
6. Steep Costs and Complexity: MTA is an extremely difficult exercise to land. With each new channel added to the digital marketing mix comes another level of added complexity. MTA can take months to implement and it is expensive and complicated. MTA involves complex algorithms and models, and choosing the right model (e.g., linear, time decay, U-shaped) can be challenging. Each model has its own assumptions and limitations, and marketers may struggle to determine which one best reflects their desired customer journeys. Also, setting up and managing these models requires specialized expertise.
Where MTA Fails, Incrementality Wins
In the final analysis, MTA is an ill-advised attribution method for marketers. Incrementality measurement, however, solves the attribution problem, but in a very different way from MTA. Incrementality testing helps marketers understand which media spend is actually driving conversions, versus what would have happened without media. Instead of relying on flawed methodologies — incrementality testing isolates the true impact of an ad.
For most channels or platforms, MTA cannot measure views or impressions, so you’re essentially just measuring clicks. Incrementality measurement accounts for the impressions and clicks within each of the platforms and therefore gives marketers a more accurate view of the true contribution of their media across their entire portfolio.
By combining advanced Media Mix Modeling with incrementality testing and platform data for a more accurate, actionable, and scalable measurement framework, marketers have a much more credible way to ensure that their brand is capturing the full impact of their entire media mix based on causal ground truth experimentation.
With this modernized approach, marketing teams can finally achieve a single, reliable source of truth for measurement and optimization.
For a master class on how to integrate incrementality testing and a modernized approach to MMM into your measurement strategy, download our latest guide: The Future of Media Mix Modeling: How Incrementality Testing is Redefining MMM for a New Generation of Marketers.