Retail's Missing Data Solution: Why POS, Panels, and Receipt Apps Fall Short

Most consumer brand leaders will answer “how is retail performing?” with a fast, confident answer. They’ll cite one or more of the following: units moved, dollar share, velocity by banner, distribution gains, promotional lift, or category rank. 

‍

But when you ask consumer brand leaders who is buying their products, they will likely speak in broad generalizations. 

‍

For the entire history of retail, there has been no practical way to build a direct, first-party relationship with retail buyers. 

‍

A shopper picks a product off a shelf, pays a cashier, and leaves. The transaction has always belonged to the retailer, and the shopper belongs to them, if anyone at all. The brand that designed the product, funded the demand, and paid for the shelf space learns only that a unit left the store.

‍

Retail brands have done what capable teams always do with a problem that has no solution: they’ve built workarounds, and gotten pretty good at them. The lack of first-party retail buyer data…hasn’t been considered a problem at all. It’s just been a fact of life that brands build hacks around. 

‍

In my role here at Brij, I have spent an unbelievable amount of hours learning the ins and outs of the history of retail, along with each of these workarounds and hacks. This article is my honest take on what retailer reports, syndicated data, consumer panels, cashback and receipt apps, and generic capture tools do well versus where they fall short. 

‍

I’ll also share how these “solutions” fundamentally fall short in comparison to a brand capturing verified purchase data directly from its own retail buyers. 

‍

Spoiler alert: my conclusion isn’t that these tools are bad. In fact, many of them are excellent at the job they were designed to do. None of them, however, were designed to put the data back into the hands of consumer goods brands - and that has quietly cost brands more than they realize. 

‍

‍

The Four Kinds of Retail Data

‍

Before comparing tools, it’s helpful to discuss the types of retail data that have been accessible to brands for years. They are not interchangeable, and the type of data determines what a brand can legally, technically, and practically do with each downstream.

‍

Aggregated data counts transactions, without people. For example, a POS report will highlight that 14,182 units sold across 611 stores last week. These reports offer units sold as the primary datapoint for analysis. 

‍

Modeled or projected data estimates the total population from a sample. A panel, for example, observes a few hundred thousand households and extrapolates to a national picture. Depending on the provider, the math can be highly sophisticated and yield useful insights. These are probabilistic insights, though; they’re educated estimates backed by little-to-no deterministic signal.

‍

Third-party or anonymized data describes real, individual people whose identity is held by someone else. A cashback app like Fetch knows exactly who bought your product, because they were the recipient (point of capture) of that buyer’s data. These apps rely on you renting the data from them, without ever owning it. The first-party customer data, despite these buyers being purchasers of your brand, is owned by the app and not your team. 

‍

Deterministic first-party data is the most precise retail data you could ever have. Consider this the golden child of retail data, offering a verified record of a specific person who bought your specific product, collected with their consent, held in your systems, under your control. 

‍

Most established consumer goods brands have access to the first three types of data. These merely answer what sold, what might have happened in the past, or what might happen in the future. 

‍

It is only the fourth type of data - first-party data - that directly answers “who purchased”, putting the reins of control back in the hands of the brand. 

‍

‍

Retailer POS Reports and Vendor Portals

‍

Walmart Retail Link and Luminate, Target's supplier portals, Kroger's reporting through 84.51°, Costco IRMA, Amazon Brand Analytics, and Whole Foods' vendor portal are the closest thing to ground truth a brand can get about what has sold.

‍

• Where they win: Retailer POS reports and vendor portals are highly accurate when it comes to transaction reporting, digging down to the store-level and item-level lines of detail. These reports turn around quickly enough for operational decisions relating to inventory management. If you need to know whether your product is actually on the shelf in Dallas Whole Foods stores, this is the source. It is also the data your buyer is looking at, which makes it the shared language of retailer conversations. 

‍

• Where these fall short: Retailer portals are scoped to one retailer, formatted differently at each one, and deliberately stop at the transaction. The shopper is the retailer's relationship and the retailer's asset. Even the premium tiers that add shopper behavior deliver it as segments and indices, not as identified individuals you can retain.

‍

• Why this matters: Practically, it’s borderline impossible for marketers to parse through these reports. Every retailer is a separate island with a separate login, a separate schema, and a separate definition of a week. Reconciling them is a person's full-time job, and no combination of portals will ever tell you that the same household bought your product at Target in March and at Amazon in June, because neither retailer can see the other and neither would tell you if they could. POS reports and portals answer how much moved, where, and when. They cannot answer who purchased and if that buyer ever returned to purchase again. 

‍

‍

Syndicated Data

‍

NielsenIQ, Circana, and SPINS built the measurement layer the packaged goods industry runs on. Syndicated data aggregates POS scan data across retailers into a comparable view of a category, then spits out a report for your brand. 

‍

• Where it wins: Syndicated data wins when it comes to competitive context. No other solution will identify your share of the category, whether the category is growing, how your velocity compares to the item next to you on the shelf, or whether a competitor's promotion is stealing your volume. This is the language of category reviews and line reviews. It supports the annual planning cycle, the trade investment case, and the argument you make to a buyer for more space. For those jobs, it is the right tool, and there is no equally viable substitute.

‍

• Where it falls short: Syndicated data is aggregated and lagged by design. Typical reporting latency runs several weeks, which makes it a more long-term strategic tool than a real-time operational one. Coverage varies by channel and provider, and things like DTC sales and Amazon sales are removed from these reports entirely, so a truly omnichannel brand is forced to evaluate their sales channels in silos. 

‍

• Why this matters: Syndicated data cannot identify your brands’ buyers and deliver brand-level buyer behavior. It cannot distinguish a first-time trial from a loyal shopper stocking up, because both look identical at the scanner. It cannot tell you that a unit sold to someone who had never purchased from you before, which is the single most valuable fact a marketer could learn from a retail transaction. Potentially the most important, it can’t tell you where buyers are discovering you and how they transact across your sales channels. 

‍

‍

Consumer Panels

‍

Panel data was created specifically to fill the "who" gap in POS data, and it deserves credit for a fair attempt. Providers including NielsenIQ, Circana, and Numerator recruit opted-in households who report their purchases, typically by scanning receipts through an incentivized app, and project that behavior to a national population.

‍

• Where they win: Panels are the best available tool for demographic and psychographic understanding at the category level. They deliver some broad-stroke insights into buyer profiles, repeat rates, trial rates, cross-purchase, and basket affinity. They can also provide source of volume, identifying whether your growth is coming from the category expanding or from a specific competitor losing share. These are real, defensible insights. 

‍

• Where it falls short: A panel is a sample. Its output is an estimate about a population, with confidence intervals that widen as you narrow the question. That is fine for a broad national category read but highly unreliable for an emerging brand, a regional launch, a new item, or a specific retailer. These are precisely the situations where a scaling brand most needs answers.

‍

• Why this matters: Ultimately, panel households are not fully representative of your customers, and you certainly cannot attain the buyer data to build ongoing relationships. They are the panel provider's respondents, contractually and legally. You receive a portrait of a buyer archetype, but these insights are actionable at the high-strategy level, not the immediate revenue impact level. 

‍

‍

Cashback and Receipt-Scanning Apps

‍

Rebate and cashback platforms sit at an interesting midpoint. They genuinely identify individual shoppers and verify real purchases with real receipts. Brands use them to drive trial, move volume during a promotional window, and support various retailer programs. 

‍

• Where they win: These apps tend to deliver fast, measurable volume. They can verify purchase at the item level, and prove useful for a launch, a seasonal push, or hitting a sales target within a specific window. The offer mechanic is proven and the consumer experiences are heavily tested to reduce friction, so brands typically get the sales movement they’re looking for. 

‍

• Where they fall short: These apps own the shopper relationship. That is their business model, and it is non-negotiable. The shopper engages with the app (such as Fetch or Ibotta), joins the app's audience, and receives offers from the app. This almost definitely includes a number of offers from your competitors in the same category. At the end of your campaigns, brands receive aggregated redemption reporting and a performance summary. Consumer identities, contact records, and the the ability to build ongoing relationships are never to be seen, trapped inside the platform.

‍

• Why this matters: These programs are highly efficient at reaching deal-seeking shoppers at volume, because that is who uses them. However, these campaigns are highly ineffective for building brand loyalty. When a promotion ends, the audience you built (regardless of if they’re a good one or not) does not transfer, because you did not build an audience; you rented access to one.

‍

One nuance worth keeping in view: cashback and receipt-scanning apps are typically funded out of trade promotion budget. This budget is typically held separate from the standard marketing budget (sometimes categorized as shopper marketing, sometimes purely called trade marketing).

‍

A brand can run a cashback program and an owned capture program simultaneously without conflict. The question is not which one to fund; it’s whether the trade program is the only thing you have, because if it is, every customer relationship it generates belongs to someone else.

‍

‍

Generic Data Capture Tools

‍

The fifth category is the one brands assemble themselves, and it often leads to an utter mess of data spread amongst disconnected spreadsheets. Typically, I see these as a combination of QR code generators, landing page builders, and embedded forms.

‍

Sometimes, they’re a mix of detached point-solution tools across rebate, sweepstakes, and product warranty offerings. At best, we’re looking at an on-brand landing page with a properly built Zapier connection to Klaviyo. 

‍

• Where they win: These tacked-together tools are typically cheap, fast, and DIY within a couple of weeks. For an isolated campaign with a short-term goal, a QR code pointing at a form can collect email addresses, sure. 

‍

• Where they fall short: These tools deliver fragmented data, typically do not verify purchases (ie, anyone can scan a QR code and enter a sweepstakes), and they almost never deliver results in a format that is actually actionable. 

‍

• Why this matters: Two critical points here:
  
  • Fragmentation is extremely painful for businesses. If a brand is using a sweepstakes platform, a rebate platform, and some other separate email capture form - none of them unify customer profiles across platforms. The same shopper appears three times as three strangers. Building the unified customer record is left as an exercise for a team that does not have time for it, which is why the integration project gets scoped every year and finished in none of them. 
  • Most importantly, these tools do not action on the data in any way. Getting a hashed, consented, verified purchase event into Meta, Google, and TikTok in a format their systems accept, with match quality high enough to be useful, and simultaneously into the CRM with the right attributes attached, is a different technical problem than a form submission. Generic tools stop well before that line.

‍

Deterministic Signal Solutions

‍

The sixth approach is the one that has not existed as a category until recently, which is exactly why only the savviest, most cutting-edge marketers are searching for it: signal. 

‍

• Where it wins. It is deterministic, meaning it records what actually happened to a specific person rather than estimating what probably happened to a population. It is owned, meaning the record sits in your systems and keeps working after any single campaign ends. It is complete at the individual level rather than sampled. It spans retailers, because the shopper identifies to you rather than to a store. And it is directly activatable, because ad platforms and CRMs are built to accept exactly this kind of record.

‍

• Where it falls short: Admittedly, not every shopper will identify themselves – but depending on the category, we see 30 - 50% opt-in rates. Capture rate will always depends on the strength of the incentive you offer.  This approach also does not produce category share, competitive velocity, or source-of-volume analysis, which is the point. It’s not a replacement for syndicated data or panels; it’s the layer underneath of them that has been missing. 

‍

The capture mechanic is simple: 

• Give the retail shopper an incentive to offer up their email or phone number (product registration, warranty claim, rebate, etc)
• Verify the purchase with a receipt upload 
• Build a CRM record: a real person, a real item, a real store, a real date, a real price, collected with consent.

‍

The activation is where it becomes powerful: 

• Send the data directly to your email and SMS platforms, so a retail buyer enters lifecycle programs the same way an online buyer does.
• Hash the transaction information, then forward it to Meta, Google, and TikTok as purchase events. Immediately, this improves reporting to deliver more accurate CAC and ROAS figures, as well as halo effect measurement. Over time, platforms will optimize audience targeting against a more complete lookalike buyer profile, rather than just a DTC-looklike one. Better targeting = more efficient spend. 

‍

‍

The TLDR: What Each Can and Can't Do

‍

Retailer POS reports and vendor portals

- Can: precise sales, inventory, and store-level performance at one retailer.

- Can't: identify a buyer, connect behavior across retailers, or feed an ad platform or CRM.

‍

Syndicated data

- Can: category share, competitive benchmarking, velocity, distribution, the retailer conversation.

- Can't: identify a buyer, distinguish trial from repeat, cover DTC and most ecommerce, or move fast enough for campaign decisions.

‍

Consumer panels

- Can: buyer demographics, repeat and trial rates, basket affinity, source of volume at category scale.

- Can't: give you a person you can contact, hold up at small sample sizes, or produce anything activatable.

‍

Cashback and receipt apps

- Can: drive verified trial volume quickly during a promotional window.

- Can't: transfer the customer relationship to you, prevent competitor offers reaching the same shopper, or leave you with an owned asset when the program ends.

‍

Generic capture tools

- Can: collect contact information cheaply for a single campaign.

- Can't: verify purchase, unify identity across mechanics, or deliver a usable conversion event to ad platforms and CRM.

‍

Owned deterministic signal

- Can: identify the individual retail buyer, verify the purchase, follow them across retailers, and activate in paid media and lifecycle.

- Can't: measure category share or competitive position, or capture every shopper.

‍

‍

The Five-Question Retail Data Stack Quiz

‍

Take each data source you pay for and ask yourself:

‍

1. Does it contain a person, or a count? If you cannot open a record and see an individual, everything downstream is inference.

‍

2. Do you own it, or are you licensed to view it? Check whether the contract lets you retain and market to what you receive after the term ends. Most do not.

‍

3. Can an ad platform ingest it as a conversion event? Aggregated reports and projected estimates cannot be sent through a Conversions API. Verified individual purchases can.

‍

4. Can your email or SMS platform act on it tomorrow morning? If the answer requires an export, a manual match, and a data engineer, the answer is effectively no.

‍

5. Does it survive the end of a campaign? A promotional program that leaves you with a performance deck has not built an asset. A capture program that leaves you with fifty thousand identified buyers has.

‍

Most brands run this exercise and find that every source they pay for fails at least three of the five. 

‍

‍

MMM, MTA, and Incrementality Tools

‍

If you’re currently thinking “she forgot about my MMM tool!”, fear not. I haven’t. 

‍

Multi-touch attribution, marketing mix modeling, and incrementality testing are all legitimate disciplines, but none of them substitutes for what is described here, because they operate one layer above it.

‍

Multi-touch attribution models the online path to conversion. These tools are pixel based, and without direct data injection, they will always be blind to retail and marketplace purchases. 

‍

Marketing mix modeling estimates channel contribution from aggregate spend and outcome data. When retail sales are not tied to identified buyers, the model infers retail performance rather than observing it, and the estimate is only as good as the inputs.

‍

Incrementality testing measures whether a given activity caused lift. Signal makes incrementality models smarter by identifying offline lift and associating it back to those given activities. 

‍

The key takeaway here is that all three of these tools types interpret signal, rather than creating it. Deterministic signal makes each more accurate, and running them without a solid retail data layer means modeling a partial picture for your brand.

‍

‍

The Impact of Identifying Your Retail Buyers

‍

With deterministic first-party retail data, three major shifts occur with compounding gains:

‍

Paid Media Gets Better Inputs

‍

Sending verified offline and marketplace purchases into Meta, Google, and TikTok has both immediate and long-term effects. Immediately, retail sales that were always happening finally get attributed, and reported return on ad spend rises to reflect reality. Over time, the algorithms learn from a fuller picture of who actually buys, and optimization improves.

‍

There is a second effect that gets less attention: suppression. When the platforms know who already bought, they can stop serving new customer acquisition ads to them. Brands routinely pay to reacquire customers they already won at retail, because the algorithm has no way of knowing those people exist. Every impression saved there is budget returned to genuine prospecting.

‍

Brands Drive LTV of Retail Buyers

‍

A retail buyer who identifies themselves enters the CRM with a verified purchase attached: item, store, date, price. That record can trigger a post-purchase flow, a replenishment reminder timed to the product's actual consumption cycle, a cross-sell, a review request, or a subscription offer.

‍

These are the same programs that already work for ecommerce buyers, applied to a group that has been structurally excluded from them. Consider the LTV of your average DTC buyer, if you have one. Now, multiply that by the number of assumed retail buyers you have each year. That is a lot of revenue you could be generating, but simply cannot without contact info. 

‍

Retail Insight Becomes Deeply Specific

‍

Not "our category buyer skews toward this demographic index," but "these are the people who bought this item at this retailer in this month, here is what motivated the purchase, and here is what they bought next." That level of specificity changes assortment conversations, launch planning, and the case you bring to a buyer.

‍

‍

How Brij Fits In

‍

Brij makes your buyer data from retail and marketplaces visible, turning those purchases into owned, deterministic signal that lowers CAC and grows LTV.

‍

Brij captures verified identity alongside receipt-level purchase context and post-purchase survey responses. Identifiers are hashed server-side, the event is typed to what the campaign optimizes on, and it moves in two directions at once: into Meta, Google, and TikTok as a deterministic purchase event, and into the brand's CRM and email or SMS platform as an identified customer with purchase context attached. Both sides of the LTV-to-CAC equation move together.

‍

What makes Brij different than the old-school solutions:

• It is deterministic at the brand's own customer level
• It is connected to the major ad platforms, automatically optimizing your paid media outcomes
• It is connected to the brand's own CRM and lifecycle tools, automatically driving LTV with trigger-based segmentation. 
• Brij delivers data directly to the brand, so you own your own customer data. No more renting it. 

‍

Brands including Chobani, Heineken, Skullcandy, Feastables, Quip, Black + Decker, Health-Ade, Caraway, TUSHY, and Bobbie use Brij. 

‍

If you’ve made it this far and you’re wondering about security, Brij is SOC 2 Type 1 compliant, GDPR and CCPA aligned, with SHA-256 hashing and customer-owned records. 

‍

Learn more about Brij here → 

‍

‍

Frequently Asked Questions

‍

Is syndicated data still worth paying for?

Yes, for category share, competitive benchmarking, and the retailer conversation. It is the wrong tool for identifying buyers or feeding ad platforms, and it was never designed to be.

‍

Isn't panel data first-party data?

No. Panel households belong to the panel provider. You license insights derived from them. First-party data means a person who gave their information directly to you, under your terms, held in your systems.

‍

Doesn't our attribution platform already handle this?

Attribution platforms model and interpret the events they receive. Pixel-based tools receive nothing from a shelf. Deterministic retail purchase events sit underneath every attribution model and make all of them more accurate, which is why measurement vendors generally treat this as a complement rather than a competitor.

‍

Can't we just use the retailer's clean room?

Clean rooms allow matched analysis against retailer data under strict privacy controls, which is genuinely useful for measurement. They do not release identified shoppers to you. You leave with an answer, not a customer record.

‍

What percentage of retail buyers actually identify themselves? Higher than most teams expect. Brij experiences convert at roughly 30 percent in packaged goods and 50 to 60 percent in durable goods, which is around ten times a typical brand email opt-in rate. The variable that moves the number most is the strength of the reason you give the shopper, followed by the visibility of the prompt on pack or in store and the friction of the flow. Durable goods run higher because a warranty is a reason people already act on.

‍

Do we still need the pixel and standard Conversions API setup?

Yes. Those cover web events. Retail and marketplace purchases are a separate stream that requires a source of verified offline events. The two are complementary halves of the same picture.

‍

How is this different from a rebate or cashback program?

The mechanic can look similar to the shopper. The ownership is opposite. In a cashback program the platform keeps the customer relationship. In an owned capture program the brand does, permanently.