US advertisers spent $60.32 billion on retail media in 2025, which makes the question of how a network recognizes its shoppers a budget question, not a technical footnote. AI powered identity resolution for retail media networks is where that promise lives or dies. Here is how to vet it before you sign.
Why identity resolution decides your retail media ROI
Every metric a retail media network (RMN) hands you depends on a match. Reach, frequency, new-to-brand rates and closed-loop sales all assume the network linked an ad exposure to the right person, then linked that person to a purchase.
When the link is wrong, the report still looks tidy. That is the problem.
Budgets are also concentrating. EMARKETER projects spend will reach $71.09 billion in 2026, and the same source says Amazon and Walmart will capture 89% of incremental spending. So you will increasingly test smaller RMNs, grocers and delivery apps whose identity stacks vary widely in quality. A vetting process is no longer optional.
An RMN’s identity graph is the measurement instrument. If you cannot inspect how it matches people, you cannot trust any number it produces.
What “AI powered” should actually mean
Vendors use the phrase loosely. Strip it down to the mechanics and you find three layers.
- Deterministic matching: logins, loyalty IDs, hashed emails and payment tokens tied to a known customer.
- Probabilistic matching: models that infer links between devices or sessions from signals like shared networks and behavior patterns.
- Hybrid methods: deterministic anchors, with models filling the gaps.
The IAB/MRC Retail Media Measurement Guidelines are blunt about the tradeoff. Probabilistic approaches are generally considered less accurate than deterministic ones for pairing devices, because they rest on inferred or modeled data, though they can scale further. The same document says the sources of assignment data and the integration methods should be disclosed to measurement customers.
That last point is your leverage. Ask for the disclosure. A network that cannot describe which share of its matches is deterministic and which is modeled is telling you something.
The AI part usually sits in the probabilistic layer: entity resolution models, graph clustering, propensity scoring. Useful, yes. But AI does not turn a thin loyalty file into a rich one. It estimates. Your job is to find out how much of your reported audience is estimate.
Person, household or device: which one are you buying?
This is the question that quietly wrecks frequency caps.
The IAB Tech Lab Identity Solutions Guidance describes identity resolving to different granularities: a device or screen, a person (multiple signals linked into one ID for an individual), or a household (signals tied together across all devices and individuals in one home). Each answers a different business question.
- A household ID suits grocery and CPG, where one basket serves a family. It overstates individual reach.
- A person ID suits beauty, apparel and anything personal. It is harder to build and rarer than vendors imply.
- A device ID is the weakest basis for cross-channel claims, since one shopper may appear as several people.
Ask which granularity the network reports at, and whether it changes between onsite, offsite and in-store. A retailer might report onsite reach at person level, then offsite at household level, and hand you a blended frequency number that means neither.
The vetting checklist
Run these questions in order. Most weak vendors fail in the first three.
- What is the deterministic share? Request the percentage of matched IDs anchored to a login, loyalty account or payment record, split by channel.
- How is the match rate defined? Match rate against what denominator? Your CRM file, the network’s file, or the ad-exposed audience? The three produce very different numbers.
- How fresh is the graph? Email addresses change, households split, loyalty members lapse. Ask how often links are refreshed and expired.
- What are the consent mechanics? Identify how opt-ins are captured, how opt-outs propagate to downstream partners and how long that takes.
- Is the method auditable? Look for third-party accreditation or willingness to let an independent party inspect the process.
- Can you test it blind? Offer a holdout and see whether the network agrees to your design.
That last item separates partners from presenters. If you are already stress-testing outside reporting vendors, the logic in our piece on a method audit first transfers directly: inspect the method before you believe the output.
Clean rooms: helpful, not a substitute for scrutiny
Most RMNs now pitch clean rooms as the answer to privacy and matching at once. They help with the first. They do not guarantee the second.
A clean room lets you and the retailer join data without exchanging raw records. But the join is only as good as the keys on each side. If your CRM carries stale emails and the retailer’s graph is mostly modeled, the room will faithfully produce a poor match, privately.
Pilot in small pieces. We covered that approach in our look at how clean room pilots work in modules. Start with one use case, such as overlap analysis, and compare the result with what the retailer’s own dashboard claimed. A gap tells you what the dashboard was hiding.
Consent and compliance are your risk, too
Brands often assume the retailer carries the legal exposure. Not so. If you upload customer lists, activate audiences or accept modeled segments, you share responsibility for how that data was gathered and used.
Ask three things in writing. Where did the consent for each identifier originate? What happens to a suppression request, and how fast does it reach offsite partners? Which sub-processors touch the data? Our guidance on consent checks before connecting ad platforms applies here with the same force.
Agentic tools add another wrinkle. As AI agents begin to buy and optimize media on your behalf, they will need to authenticate against the same identity layer. That is why agent and customer identity alignment belongs on your roadmap now, not after launch.
Prove it with a holdout
Documents and demos only go so far. Structure a test that the network cannot game.
- Pick one product line with clean sales data on your side.
- Split a seed audience of known customers into exposed and holdout groups, set before the campaign.
- Ask the network to report matched reach, frequency and sales lift for each group.
- Compare against your own first-party records, such as ecommerce orders or a retailer sell-through feed.
The IAB/MRC guidelines include a full chapter on incrementality, covering randomized controlled trials and synthetic controls. Use that as your shared vocabulary so the network cannot redefine “lift” mid-conversation.
When results arrive, treat them with discipline. Our five checks for AI-generated insights are a good filter before any finding moves budget.
Do not ask a network whether its identity resolution works. Ask it to survive a test you designed.
FAQs
What is identity resolution in retail media?
It is the process of linking identifiers such as emails, loyalty IDs, device signals and payment data to a single shopper, household or device. RMNs use it to target audiences, cap frequency and connect ad exposure to purchases.
How is AI used in identity resolution?
Mostly in the probabilistic layer. Models infer likely links between devices and sessions, cluster records that appear to belong to the same person and fill gaps where deterministic data is missing. Ask what share of matches is modeled rather than anchored to a login.
What is a good match rate?
There is no universal benchmark, and the number is meaningless without its definition. Ask what the denominator is, how much of the match is deterministic and how freshness is maintained. Then validate it against your own data in a pilot.
Do clean rooms solve identity problems?
They solve a privacy problem, letting parties join data without sharing raw records. They do not repair weak or stale keys. A poor graph inside a clean room still produces a poor match.
Who is responsible for consent in retail media?
Both sides. The retailer collects consent, but a brand that uploads lists or activates audiences shares exposure. Get consent origin, suppression timing and sub-processor details in writing.
Which standards should I reference when vetting an RMN?
Start with the IAB/MRC Retail Media Measurement Guidelines for measurement and incrementality, and the IAB Tech Lab Identity Solutions Guidance for identity concepts and granularity. Both give you neutral language for contract and RFP questions.
Next step: before your next RMN commitment, send the six checklist questions in writing and run one blind holdout on a single product line. Decline to scale until the network’s matched numbers agree with your own sales records.
Top Influencer Marketing Agencies
The leading agencies shaping influencer marketing in 2026
Agencies ranked by campaign performance, client diversity, platform expertise, proven ROI, industry recognition, and client satisfaction. Assessed through verified case studies, reviews, and industry consultations.
Moburst
-
2
The Shelf
Boutique Beauty & Lifestyle Influencer AgencyA data-driven boutique agency specializing exclusively in beauty, wellness, and lifestyle influencer campaigns on Instagram and TikTok. Best for brands already focused on the beauty/personal care space that need curated, aesthetic-driven content.Clients: Pepsi, The Honest Company, Hims, Elf Cosmetics, Pure LeafVisit The Shelf → -
3
Audiencly
Niche Gaming & Esports Influencer AgencyA specialized agency focused exclusively on gaming and esports creators on YouTube, Twitch, and TikTok. Ideal if your campaign is 100% gaming-focused — from game launches to hardware and esports events.Clients: Epic Games, NordVPN, Ubisoft, Wargaming, Tencent GamesVisit Audiencly → -
4
Viral Nation
Global Influencer Marketing & Talent AgencyA dual talent management and marketing agency with proprietary brand safety tools and a global creator network spanning nano-influencers to celebrities across all major platforms.Clients: Meta, Activision Blizzard, Energizer, Aston Martin, WalmartVisit Viral Nation → -
5
The Influencer Marketing Factory
TikTok, Instagram & YouTube CampaignsA full-service agency with strong TikTok expertise, offering end-to-end campaign management from influencer discovery through performance reporting with a focus on platform-native content.Clients: Google, Snapchat, Universal Music, Bumble, YelpVisit TIMF → -
6
NeoReach
Enterprise Analytics & Influencer CampaignsAn enterprise-focused agency combining managed campaigns with a powerful self-service data platform for influencer search, audience analytics, and attribution modeling.Clients: Amazon, Airbnb, Netflix, Honda, The New York TimesVisit NeoReach → -
7
Ubiquitous
Creator-First Marketing PlatformA tech-driven platform combining self-service tools with managed campaign options, emphasizing speed and scalability for brands managing multiple influencer relationships.Clients: Lyft, Disney, Target, American Eagle, NetflixVisit Ubiquitous → -
8
Obviously
Scalable Enterprise Influencer CampaignsA tech-enabled agency built for high-volume campaigns, coordinating hundreds of creators simultaneously with end-to-end logistics, content rights management, and product seeding.Clients: Google, Ulta Beauty, Converse, AmazonVisit Obviously →