Misleading and Misaligned

In the never-ending quest to prove the value of marketing, improve advertising performance, and simply do more with less, many marketers have embraced Multitouch Attribution (MTA)—a seemingly perfect tool to track every clickable moment on the customer journey. MTA is chock-full of benefits: countless dashboards provide granular reporting, real-time insights and instant credit assignment. It gives digital marketers the sense that they could pinpoint exactly which ads, placements, and even message sequences were driving conversions.

But for all the appeal of MTA, there’s a harsh reality: MTA suffers from fatal flaws that relegate it to second-tier measurement status well behind Marketing Mix Modeling (MMM). As with most dreams, you wake up—usually to discover that all the “reporting” was built on cookie crumbs, tenuous user identifiers and probabilistic device matching that vanish at the slightest hint of privacy regulations and adblockers. MTA’s eroding foundation is a systemic challenge, further weakened by short-sighted measurement windows that gaze upon a narrow slice of the marketing ecosystem and the broader market.

Below, we’ll walk through the three core flaws that doom MTA to be Misleading, Marginal, and Misaligned. We’ll then address common questions we have heard in response to this material. By the end, you may wonder if your MTA efforts are doing more harm than good. (Spoiler: we have a multi-touch attribution white paper addressing exactly that.)

MTA makes a splashy promise: to pinpoint which clicks drive which sales. But peel away the veneer and its credibility suffers from measuring the wrong output metric, recency bias and stale, out-of-date models.

1. Focusing on Associated, Not Incremental Sales
   Instead of showing which marketing actions lifted revenue, MTA typically reports on where a customer clicked before buying. This distinction might sound subtle, but it’s critical. “Incrementality measurement provides marketing leaders with the linchpin directly connecting marketing actions to improved business outcomes,”¹ explains one Gartner report on the topic. MTA’s obsession with “associated sales” conveniently skips that incremental step, leaving you with correlation instead of causation. Your CFO wants to see your incremental impact on the business to the unique impact of marketing deduplicated from other growth investments.
2. Short and Arbitrary (Lookback) Windows Can’t See Long-Term Effects
   Brand advertising’s influence and shifting consumer preferences don’t fit into a 30- or even 90-day clickstream window. But guess what? That’s precisely how MTA tools at Google Ads (default 30-day lookback, capped at 90 days), Adobe Analytics (capped at 90 days) and others operate. Consequently, the long-term payoff of brand marketing campaigns—which can transform a business, increasing revenue ROI an average of 90% according to ROI Genome—rarely gets its due in an MTA model. As WARC has recently trumpeted, marketing has a “multiplier effect” when measurement considers brand and performance, but MTA is designed for immediacy, encourages impatience, and can’t observe the multiplying power of brand.
3. Static Models Fail the Accuracy Check in a Dynamic World
   MTA frameworks are notoriously slow to adapt. They’re fueled by data but not data driven. Like a manufacturing line, the processing of raw materials does not change the line’s machinery. Once the attribution rules are locked in, they’re practically stuck, regardless of how much information flows through the system. Consumer behavior can shift overnight (hello, global pandemic), but MTA rulesets don’t. Ironically, MTA’s “real-time” results often rely on unchanging data processing models.

Half of brands should decrease paid search budgets and the same for retail media. Why? In part because MTA consistently makes bottom of the funnel investments appear more attractive than they are.

The End Result? The promise of model accuracy quickly collapses under these constraints, leaving you with incomplete and misleading takeaways that goad flawed marketing decisions. The resulting inaccurate models contribute to poor budget allocation with too much investment going to the bottom funnel. According to ROI Genome^®, 50% of brands should decrease paid search budgets and a similar percentage should cut their retail media network investments². Why? In part because MTA consistently makes bottom of the funnel investments appear more attractive than they actually are. As you might expect, Analytic Partners ROI Genome observes the opposite effect as well: underinvestment where consumers are generally less likely to click. For example, eight of 10 brands should be allocating more to CTV and streaming video.³

By design, MTA delivers an anemic view of your marketing ecosystem. Think of it as a microscope looking at only part of the specimen, leaving much of your marketing strategy unexamined.

4. Reliance on Shrinking Digital Exhaust
   MTA relies on masses of granular consumer tracking data. But third-party cookies are crumbling, tracking legislation is on the rise and consumers are savvier about privacy. The large quantities of “granular tracking data” MTA requires no longer exists: 73% of consumers browse at least some of the time in no-tracking (or incognito) mode, and 67% have a junk email account for spam⁴. What began as a leaky bucket has morphed into a colander, letting crucial data slip away at an accelerating rate as digital platforms fortify their walls, consumer behaviors shift, and privacy regulations tighten.
5. Isolated to the Trackable, Digital Subset of Marketing
   Let’s say you sell through both e-commerce and brick-and-mortar. MTA will try to track digital leads, but what about in-store displays? Sponsorship of the local sports teams? Out-of-home ads? Many real-world tactics are excluded from the MTA conversation, robbing you of a full view of your marketing universe. Since media channels work together and complement one another – each additional channel, on average, drives an additional 11% ROI improvement – this partial view is catastrophic for those trying to optimize ROI across channels.
6. Ignores Marketing Fundamentals
   Marketers know terms like “adstock,” “wear-out,” and “diminishing returns” for a reason: Advertising impact accumulates and depreciates over time, influenced by factors like consumer sentiment, seasonality, and the broader economy. MTA rarely accounts for these variables, creating a simplified (and flawed) perspective on how marketing truly works. This oversimplification weakens both the accuracy of measuring yesterday’s spend and your ability to leverage today’s “insights” for forward-looking decisions.

Translation? Don’t use MTA to allocate your channel or program budgets – it’s based on incomplete, short-term metrics that ignores real-world interplay of your brand efforts, offline channels, and marketplace dynamics.

MTA will not help you do more with less, but by chasing the wrong metric – ROAS – you will do less with less over time. This is the consequence of a measurement approach based predominantly on data availability rather than causality and econometrics.

At its core, effective marketing analytics should address business critical questions: competitive threats, market expansion, and even product lifecycle queries. Here’s where MTA often falls off a cliff. (Maybe that is a good thing given all the above?)

7. Disregards Commercial Questions
   Your C-suite wants to know how your marketing programs respond to competitors’ moves or the trade-offs between increasing brand investment and opening new customer-facing locations. MTA shrugs — “not my job.” On the other hand, MMM can be built with a commercial view front and center by including factors like competitive density, share of voice and relative price. This equips CMOs with realistic and accurate scenarios that expands internal partnerships with Finance, Operations, Product and/or Sales. MTA was never designed to answer a broader set of C-suite questions.
8. Neglects Your Customers’ Conditions
   Your customers live in a world of shrinking discretionary budgets and rising interest rates—a place where luxuries and long-term loans don’t exactly sell themselves. While macroeconomic factors vary across industries and lie beyond your control, you still need them woven into the analysis. MMMs do that, allowing you to plan against multiple possible scenarios or re-plan when unexpected economic shifts occur. Unfortunately, MTA often ignores these market forces entirely, as if a default 30-day lookback were sufficient to grasp evolving economic headwinds.
9. Fails to Integrate Essential Business Factors
   Marketing success isn’t just about media performance—it’s shaped by distribution networks, pricing shifts, product launches, and customer experiences. Yet, MTA overlooks these critical factors. By focusing solely on click data, it fails to capture how business operations and customer satisfaction interact with marketing efforts.Consider pricing strategy: When a company plans a price increase, strong advertising can offset potential revenue loss. Analytic Partners’ ROI Genome found that for a billion-dollar brand implementing a 10% price increase, strategic advertising support unlocked an additional $33 million in opportunity. With MMM you have a brand-specific simulator that allows you to test various pricing and marketing scenarios—helping you find the optimal balance for growth.

“Integration of MTA and MMM” was the top problem statement in 2024 according to the MMA Global. MTA will never reconcile with MMM, often not even directionally, due to the nine factors mentioned in this blog.

Why Does Misalignment Hurt? With MTA, marketing decisions become siloed from the rest of the enterprise—fine for browsing interesting (though misleading and marginal) dashboards, but terrible for guiding holistic strategies that truly move the needle. It’s no wonder the industry is still grappling with bridging MTA and MMM. According to the MMA Global, “Integration of MTA and MMM” was the top problem statement in 2024.⁵ MTA will never reconcile with MMM, often not even directionally due to the nine factors mentioned above.

Absolutely. If you’re using any rule-based attribution system – whether it’s time decay, inverse J, uniform distribution, or even your own carefully crafted variant – these nine challenges apply. You must avoid using MTA for decisions across channels, different stages of the customer journey, and, in many cases, even across geographies⁶. If you’re using a single-touch model – and let’s face it, you’re probably using last-touch as it’s the most common — it’s even worse. You’re essentially deciding a novel’s entire plot by skipping to the last page. So yes, all the above still applies for all rule-based attribution solutions.

Sure, there are a few rare, almost extreme exceptions—organizations that own their entire digital ecosystem and don’t rely on third-party platforms to acquire new customers or traffic. In those cases, the marketing function is closer to a customer experience exercise, often with in-app sales as the goal. Some of our watchouts naturally drop off in such a scenario. But unless you’re running a hermetically sealed digital fortress, you’re still going to feel the sting of MTA’s blind spots.

This Doesn’t Apply To My Expensive Gold Standard MTA Model, Right?

If you’re investing over a million dollars a year on a ’world-class’ attribution platform with a “’best-in-class’” identity graph and ’proprietary algorithm’ tailored to your customers, you’ll likely see better accuracy than with rule-based MTA (you’ll have less issues with challenges one and three). However, you’re still doing MTA, meaning many fundamental flaws remain, especially around its marginal and misaligned nature. You may dodge a few pitfalls, but most of the challenges persist—especially if you’re business focuses on new-customer acquisition across multiple channels. Most large organizations, frankly, get hit with the full slate of limitations.

In short, even the most sophisticated, million-dollar solutions can’t escape MTAs underlying structural problems. If your goal is to capture genuine incremental impact to enhance business planning, MTA will always leave significant gaps.

If you are looking for a simple declaration on using MTA, it is this: Only use MTA for high frequency tactical optimization within a single channel. If you’re trying to figure out which set of paid search keywords is associated with conversions this week or which social media creative is outperforming the rest in daily click-through rates, MTA can give you that near-real-time view. It’s handy for quick-turn adjustments—like pausing an underperforming Facebook campaign or reallocating ad spend between Google keyword groups. In these scenarios, frequent (daily or weekly) checks provide immediate feedback to micro-optimize tactics. Yes, this may feel constraining—but that’s exactly the point, given the nine cautions we’ve already discussed. MTA is a tool for making fine-tuned adjustments. You can’t use it to shape your macro-level strategy, just as you wouldn’t rely on a five-minute weather forecast to plan an entire family vacation.

At Analytic Partners, we’ve seen customers abandon their cross-platform MTA programs entirely. Instead, they’ve embraced a robust Commercial Analytics solution to handle overarching measurement and forward planning—an approach that fully meets their strategic needs. For fast, within channel shifts, they rely on built-in platform tools to micro-adjust tactics in near-real time. Essentially, they’ve opted out of ’promise of MTA’ — and found they’re better off without it.

As marketing teams everywhere are asked to do more with less, it’s natural to turn to analytics for insights. The problem? MTA won’t deliver genuine insights—it just delivers reports full of ROAS metrics that seem insightful but are misleading. The truth is that the method and the metrics are misleading and misaligned. If you truly want to understand what’s driving meaningful results, shift your focus to MMM. After all, industry veterans consistently champion it for a reason: it takes a holistic view, accounts for all marketing activations, includes all conversions, and measures incrementality in a way that speaks to business outcomes. If you’re ready for a modern approach, Commercial Analytics is ready to deliver.

Want the Full Story?

To dig deeper into why MTA might be doing more harm than good—and how to course-correct—download our white paper: “Attribution: Is It Doing More Harm Than Good?” We dissect MTA’s structural flaws, showcase real-world gaps in media performance across measurement methods, and outline actionable steps for adopting more robust, business-aligned, data-informed decisioning engine.

References:

1. Gartner. Incrementality Measurement: A Marketer’s Best Friend. Published 30 November 2022.
2. “ROI Genome Highlights” Presented at NorthStar Connect, September 25, 2024.
3. “ROI Genome Highlights” Presented at NorthStar Connect, September 25, 2024.
4. Gartner. 6 Keys to Deliver Best-in-Class Personalization in Financial Services. Published 2 January 2025.
5. MMA Global. State of Attribution Annual Marketer Survey Confidential. June 2024.
6. This happens when data privacy rules vary across geographies. For example a user’s path in the U.S. is more likely to be tracked, while the same journey in Germany might be truncated due to stricter privacy laws.