What Syndicated Data Does Not Tell You

Syndicated POS is brick-and-mortar scan data. It misses most DTC and much of e-commerce, it cannot see retailers who decline to participate, reporting lands a week or more after the selling week, and it records what happened rather than why. The expensive mistakes happen at these edges, where teams assume coverage that is not there.

The e-commerce hole

Syndicated POS is built from physical register scans, and the channels growing fastest for many emerging brands (your own DTC site, much of Amazon, Shopify storefronts) typically sit outside that surface.

The picture is changing, though. Providers now offer integration of Amazon and other online retailer data, and Circana's MULO+ expansion brought major e-commerce into the multi-outlet universe [1]. What matters is that this volume is captured and projected under a different methodology than traditional POS, so it is not simply more of the same data. Delivery-service volume is its own case again, with its own coverage rules.

The distortion is still real where coverage is thin. A brand doing 40% of its business online can look, in a brick-and-mortar read, like a brand 40% smaller than it is. If your buyer is judging total brand momentum from a MULO chart, the online story is yours to bring.

Panel is the usual way to bring it. Because panel measures the household rather than the register, an online purchase is captured the same as an in-store one, which means panel typically has a read on e-commerce volume that POS does not. It will not give you the item-level precision of a scan feed, but it will tell you the size and direction of the channel and who is buying there.

Retailers who do not participate

Some retailers license their data to no syndicated provider at all. Trader Joe's is the most cited example. Its volume does not appear in POS feeds under any contract, at any price, from any provider.

This is a different situation from a retailer whose data flows to one provider only. Where an exclusive relationship exists, other providers may still carry an estimate; where a retailer is non-participating, there is nothing to estimate from. Panel data can still see purchases at these retailers, because measurement happens at the household rather than at the register, one of the clearest cases where panel coverage reaches past POS coverage.

Club, and partial reads

Club coverage has changed, and the old caution is now only half right. Historically club retailers did not share scan data the way grocery chains did, so club figures inside a projected total were partial or modeled estimates rather than direct scans. Costco now reads through Circana, which brings the largest club retailer into direct measurement for brands on that contract. If your provider does not hold the read, your club volume is still estimated, so confirm what you are actually buying before you quote a club number.

Exclusive retailer relationships create a related nuance. Per Bedrock Analytics' 2024 overview, Whole Foods reads directly through Nielsen [2]. If your contract is with another provider, your Whole Foods volume is estimated inside projected totals rather than scanned, and the finer the cut you ask for, the shakier that estimate gets. The provider-by-provider detail is in chapter 3.

The lag, and the moving history

Syndicated data is periodic, never live. The selling week has to close, the scans have to flow in, and the file has to be processed and released, so reporting typically lands a week or more after the week it describes. By the time you read a promotion's result, the promotion has been over for weeks.

And as Chapter 2 covered, the history behind the new period is not fixed. The operational answer to both is unglamorous discipline: know your delivery calendar, date every extract, and treat "which file is this number from?" as a first-class question.

No "why" in the file

The deepest limit has nothing to do with coverage. Syndicated POS records outcomes: units, dollars, distribution, price, promotional conditions. It carries no perspective on what drove consumer decisioning: no marketing exposure, no shelf placement quality, no competitor activity outside the data, no shopper intent.

When velocity drops 12% in one market, POS can tell you it happened, decompose where and when, and rule candidates in or out. It cannot tell you that a regional distributor missed deliveries or that a competitor reset the shelf.

Panel narrows the gap without closing it. It can tell you whether the households buying you shrank or whether the same households bought less, whether your lapsed buyers went to a specific competitor or left the category, and which retailers absorbed them. What remains outside both datasets is everything that happened away from the purchase: the marketing, the shelf execution, the distributor, the competitor's plan. That story always comes from combining the data with knowledge that lives outside it.

What fills the gaps

The edges explain a growing pattern: syndicated data alongside retailer first-party data and a brand's own internal context [3]. Retailers now sell direct reads of their own stores, and a portal export is a census of one chain (every store, every transaction, no projection) while syndicated is a projected read across many. One is deep and narrow, the other wide and modeled, and mature brand teams carry both, plus the internal knowledge that neither feed contains.

That stitching is the practical reason Sous connects retailer portals and internal files alongside all three syndicated feeds: the question "how are we really doing?" is usually answered across sources, and each source keeps its own math. When each of those feeds updates on its own calendar, keeping the combined picture current is its own job, which is what the drop handles itself is about.

The final chapter turns all of this into a decision: which syndicated data provider do you need?

Common questions

Does syndicated data include Amazon? Increasingly, yes, but not as ordinary POS. Providers offer e-commerce measurement and Circana's MULO+ brought major e-commerce into the multi-outlet universe [1]. The methodology behind those figures differs from register scan data, so treat them as a related dataset rather than more of the same.

Does syndicated data include Costco? Yes, through Circana, which holds the Costco read. On other providers club volume is estimated rather than directly scanned, so confirm coverage before quoting a club number.

What about Trader Joe's? Trader Joe's does not share data with any syndicated provider, so it does not appear in POS feeds at all. Panel data can capture purchases there, because it measures households rather than registers.

How current is syndicated data? Reporting typically lands a week or more after the selling week closes, on your contract's cadence. It describes weeks that have already ended, and even that history can be restated.

Can syndicated data tell me why sales changed? POS can quantify and localize the change and eliminate suspects. Panel can tell you whether your buyer base shrank or your buyers bought less, and where the ones you lost went. Neither observes what happened away from the purchase, so pair both with retailer data and what your field team knows.