Nearly nine in 10 retailers are measuring personalization with attribution models that stop short of the full path to purchase, according to a survey of retail and ecommerce leaders from AdRoll and eTail Insights. Some rely on last-click, which credits whatever a customer touched right before converting and skips everything that came before. Others rely on platform-specific reporting that shows performance within one channel but never connects the dots across channels.
The upside is that this is a well-understood problem with a proven fix. Multi-touch attribution modeling already exists, and retailers who use it for retargeting have shown it works. Here's what that model looks like, why it outperforms last-click for personalization, and how retailers are putting it to work.
Most retailers can see personalization results in a single channel, but very few can trace a customer's full path across channels to identify which touchpoints actually contributed to the sale.
Last-click attribution credits whichever touchpoint immediately preceded the conversion, whether that's an email click, a retargeting ad, or a direct visit. Last-click is used by 46% of retailers as their primary approach to measuring personalization's contribution to conversions.
Platform-specific reporting shows how a channel performed on its own terms but doesn't connect that performance to what happened in other channels. This approach is used by another 41% of retailers.
Combined, that's nearly nine in 10 retailers working from a partial view of what's actually driving conversions. A channel can look active and engaged in its own report while still being invisible in the bigger picture of the customer's path to purchase.
The data isn't the problem. Most retailers already have customer interaction data sitting in multiple platforms, sometimes more of it than they know what to do with. What they don't have is a way to connect it. Last-click and platform-specific reporting were each built to answer a narrower question: what happened in this one channel, or what happened right before conversion.
Personalization strategies span email, social, on-site, retargeting, and increasingly CTV. Those single-touchpoint methods were never built to show how those channels work together to move a customer toward a sale.
Everything that happened earlier in a customer's journey gets zero credit under last-click attribution, even when personalization was doing real work at every step along the way.
Consider a single customer's path to purchase:
A dynamic email reminds a shopper of an item they viewed
A retargeting ad follows up on that same intent
An on-site recommendation surfaces a related product
A final ad click happens right before checkout
Under last-click attribution, that last ad gets 100% of the credit. The email, the retargeting touch, and the on-site recommendation all get zero, despite each one moving the customer closer to converting.
You spend weeks building a personalized email flow that re-engages a shopper who abandoned her cart. This flow nudges her back with the right product at the right moment. She comes back, browses, and a few days later converts through a retargeting ad.
The email worked — it's the reason she came back at all. Yet last-click attribution hands 100% of the credit to the ad that ran days later. The email that did the actual work of winning her back shows up nowhere in the report.
Multi-touch attribution modeling asks a different question than last-click does. Instead of "which channel closed the sale," it asks "which channels earned a piece of it, and how much." Go back to the shopper from the last section. Her email flow, her retargeting ad, and the recommendation she saw on-site all contributed to getting her to convert. Multi-touch attribution gives credit to all three.
Here’s what that looks like:
Her cart-recovery email built early interest
The retargeting ad kept her engaged
The on-site recommendation nudged her toward a specific product
All before the final click ever happened.
The core difference is scope:
Last-click attribution looks at whatever happened right before conversion.
Multi-touch attribution looks at the entire path, weighing every touchpoint a customer engaged with along the way.
An email or social ad that introduces a product weeks before purchase will always look ineffective under last-click measurement, since it's rarely the final touch. Under multi-touch attribution, that same touchpoint gets recognized for the role it actually played in moving the customer forward.
With that fuller picture, teams stop guessing at which channels deserve more budget and start knowing:
Campaign optimization can target the touchpoints doing the most work at each stage of that path
Personalization strategy gets evaluated on its actual contribution, since most personalized touchpoints happen earlier in the journey where last-click can't see them
Retailers are putting real budget behind fixing this:
66% of retailers plan a moderate increase in cross-channel personalization investment over the next 12 months.
7% plan a significant increase.
24% name cross-channel attribution as their single biggest investment priority for the next 12 months, ranking it above every other capability.
Where that spending is headed is the more useful signal. When asked which improvements would most accelerate their ability to deliver personalized experiences, 72% of respondents pointed to improved measurement and attribution, ranking it above AI, data unification, and organizational alignment.
From the Personalized Channel Activation in Retail & Ecommerce Report from eTail and AdRoll.
A Senior Vice President at a global retailer summed up why that discipline is important: "It's really about having a very clear framework about what decisions you are making and what data you need for that decision. That needs to stay constant, even as the MarTech space is changing so rapidly."
Omnichannel and multichannel both describe a retailer's presence across several channels, but the scope is different. Multichannel typically refers to digital channels like email, social, CTV, and retargeting. Omnichannel extends that same connected view to in-person experiences too, including store visits and DOOH.
The difference is connection. Multichannel means those channels exist side by side. Omnichannel means every touchpoint, whether it's a digital ad or a visit to a physical store, gets read as part of one connected customer journey. That's what multi-touch attribution is built to measure: crediting each touchpoint for the role it played, digital or in-person.
Omnichannel attribution improves ROI by showing retailers which touchpoints are actually earning results across the full path to purchase. That visibility supports a few concrete decisions:
Reallocating spend toward the channels and touchpoints proven to build momentum earlier in the journey
Justifying personalization budget with data tied to its actual contribution across the customer's path
Prioritizing measurement investment ahead of adding new channels, which tracks with this report's finding that channel expansion ranked last among retailer priorities
Personalization measurement has a proven model to work from. Retargeting has already solved this problem: according to a separate AdRoll and eTail Insights survey on retargeting campaign effectiveness, 82% of retailers use multi-touch attribution modeling to evaluate their retargeting campaigns.
That adoption rate puts retargeting well ahead of where personalization measurement stands today. Retargeting teams already have the infrastructure, the reporting habits, and the organizational buy-in to track a customer across multiple touchpoints.
Personalization teams don't need to invent a new measurement approach. They need to apply the one retargeting has already put to work.
Last updated on September 21st, 2026.
