Why this matters
A functional-beverage brand doing roughly $6M in annual SPINS-measured revenue asked me why its "adaptogen" segment share jumped four points year-over-year when its own sales had barely moved. The answer was not a market shift. It was product attribute tagging: SPINS had refined the adaptogen criteria between the two pulls, and a batch of competitor SKUs that used to sit outside the segment now sat inside it. The brand had booked a share change that was pure definitional drift. If you build competitive sets out of attribute tags, and in natural and specialty CPG you almost have to, you need to know how the attribution layer is built and where it moves under you.
How SPINS product attribute tagging works
Product attribute tagging in SPINS is a maintained layer over the UPC, not a field a brand fills in. It works in three moves.
-
Classify the product type. Every UPC gets a product-type attribution, the granular descendant of the broad category code. "Energy Drinks" becomes "sparkling energy," "energy shot," or "RTD coffee with function," depending on the item.
-
Apply the attribute tags. On top of product type, SPINS layers ingredient, certification, and functional-benefit tags: plant-based, non-GMO, USDA Organic, gluten-free, keto, adaptogen, immunity. This is the layer no other syndicator matches on natural and specialty, and it is the part brands most often take for granted.
-
Version and revise. SPINS periodically re-reviews the criteria for a tag and reclassifies the SKUs that qualify. The tag name stays the same; the membership changes. This is the step that quietly breaks year-over-year comparisons.
The reason this is hard to copy, and the reason SPINS charges for it, is step two. Anyone can read a category code off a UPC. Deciding that a specific mushroom-based cold brew is "adaptogen" and "functional" and "plant-based" takes a maintained taxonomy and a human review process, and that is the moat Circana and NielsenIQ have not closed in the natural channel. The SPINS vs. Circana vs. NIQ comparison walks through where each provider leads; attribute tagging is where SPINS does.
The layered tag stack
A single UPC carries descriptors at several levels at once. Reading them as a stack, rather than a flat label, is what makes attribute tagging useful.
| Tag level | Example values | Analyst use |
|---|---|---|
| Category | Energy Drinks, RTD Coffee | The shelf set and the buyer who owns it |
| Product type | Sparkling, Cold Brew, Energy Shot | The form factor, for like-for-like velocity |
| Ingredient / base | Plant-based, Dairy, Mushroom-based | Sourcing and clean-label positioning |
| Certification | Non-GMO, USDA Organic, Gluten-Free | Verified claims that gate some retailers |
| Functional benefit | Adaptogen, Energy, Immunity | The job the shopper hires the product for |
The competitive-set logic lives in the bottom three rows. When you filter to "tagged adaptogen and plant-based, in the RTD coffee product type, natural channel, last 52 weeks," every one of those conditions is an attribute tag, and the resulting segment is a peer group a buyer will accept as fair.
Attribute tags versus the category code
The instinct of most brand teams new to SPINS is to run everything off the category code, because it is the one field that is always populated and never ambiguous. The problem is that a category code answers "what shelf is this on," and almost no interesting competitive question stops there. "Energy Drinks" at Sprouts is a $200M-plus set that lumps a legacy sugar-heavy sparkling brand in with a $9M adaptogen upstart, and comparing the two on category-level share tells you nothing a buyer will act on.
Attribute tags are what let you draw the line where the shopper actually draws it. A clean-label adaptogen brand is not competing with the sugar-heavy incumbent for the same purchase, even though they share a category code, so its true competitive set is the attribute cut, not the category. This is also why two analysts pulling "the energy category" can walk out with different numbers: one filtered on the category code and one filtered on a stack of attribute tags, and neither wrote down which. Any competitive-set methodology that does not name its attribute filters is not reproducible, and a category review built on a set nobody else can rebuild is a set the buyer will discount the moment it is questioned.
The discipline is to treat the category code as the coarse first pass and the attribute tags as the definition that actually travels: write the tag stack into the report so the next person, or the buyer's own analyst, can rebuild the exact same segment.
Worked example: sizing a non-GMO functional beverage segment
The version-drift problem is easiest to see on one segment. A brand wants to track its share of the "non-GMO functional beverage" segment in the natural channel, so it pulls the segment total in two consecutive annual reports.
Under the 2024 attribute set, the segment is $42.0M. Under the 2025 attribute set, the same segment reads $47.3M, a 12.6% apparent gain. Sales did not grow: SPINS moved a handful of borderline SKUs, a few sparkling-energy items that newly cleared the non-GMO criteria, into the segment. If the brand's own dollars held flat at $5.0M, its computed share fell from 11.9% to 10.6% without a single lost sale. That is a phantom 1.3-point share loss, and it walks straight into a category review deck unless someone catches the version change.
| Measure | 2024 attribute set | 2025 attribute set |
|---|---|---|
| Segment total | $42.0M | $47.3M |
| Brand dollars | $5.0M | $5.0M |
| Computed share | 11.9% | 10.6% |
| Actual sales change | baseline | 0% |
The only number in that table that changed for a real reason is the segment total, and it changed because the tag definition changed. Read the share row in isolation and you fire off a "we are losing the functional-beverage segment" email. Read it next to the segment-total row and you see the truth: the denominator grew under you. This is the most common way product attribution quietly corrupts a share trend, and it stays invisible unless you carry the version alongside the number.
The fix is procedural, not clever: pin the attribute version. Ask your SPINS rep which attribute vintage each extract was built on, rebuild the prior-year segment under the current version before you compare, and note the version on the slide. It is the same discipline as holding a category definition stable so the historical comparison stays honest.
Where attribute tagging breaks
Three failure modes account for most of the bad numbers I have seen come out of attribute tags.
Version drift, from the worked example: the segment total moves because the tag membership moved, not because the market did. This is the most common and the most expensive, because it looks exactly like real growth or decline. It is worst in the fast-moving functional segments a natural-channel brand most wants to track, since those are exactly the tags SPINS re-reviews most often, so the categories where attribute tagging is most valuable are also the ones where the drift risk runs highest.
Overlap you did not account for. Attribute tags are not mutually exclusive. A SKU can be tagged both "granola" and "snack bar," so summing two product-type segments double-counts it: add a $12M granola cut to an $8M snack-bar cut and you may report $20M for a real universe of $17M, because $3M of items sit in both. Before you trust a total, know whether your cuts overlap, and dedupe on UPC rather than on segment subtotals.
Off-taxonomy items falling out. A genuinely novel product, the first mushroom-based energy shot in a set, may sit unclassified until SPINS adds a tag for it. During that window it is missing from every attribute cut, and a zero in an attribute segment is easy to misread as "no sales" when it really means "not yet tagged." The glossary entry on attribute tagging covers the definitional side of this in brief.
Auditing how your own SKUs are tagged
Before you build a single competitive set, audit your own attribution. Most brands underuse product attribute tagging in their first SPINS contract year for the same reason: they never checked how their SKUs were classified, so every downstream cut inherited a mistake they could not see. Four questions are worth raising with your SPINS rep early.
-
Which product types carry each of my UPCs? A SKU with two product-type attributions, both "granola" and "snack bar," lands in two competitive sets at once. You cannot define a defensible peer group until you know which assignments are in play, and whether a competitor's item you expected to see is tagged the way you assumed.
-
Which attribute tags did you apply, and which did you decline? If your cold brew markets itself as adaptogen but SPINS did not tag it that way, you are absent from the exact segment you are trying to own. That is a fixable mis-tag, but only if you find it.
-
What attribute version is each extract built on? Ask for the vintage on every pull, and write it on the report. This is the one habit that prevents the phantom-share-loss table above.
-
How often does this category get re-reviewed? Fast-moving segments like functional beverage get re-tagged more often than shelf-stable staples, so the drift risk is not uniform. Knowing your category's cadence tells you how suspicious to be of a year-over-year segment move.
None of this requires special data access. It is an hour with your rep and your own UPC list, and it is the cheapest insurance a category analyst can buy against a bad number in a buyer meeting.
Doing this in Scout
The grind here is that attribute tags arrive inside a weekly SPINS extract and then get flattened the moment the file lands in Excel, so every competitive set has to be rebuilt by hand from the raw tag columns. Scout keeps the SPINS attribute layer intact and queryable, so a segment like "adaptogen and plant-based RTD coffee, natural channel, trailing 52 weeks" is a saved view your whole team can open, not a pivot table stranded on one analyst's laptop. When the attribute version changes, you re-point the same view rather than rebuilding the filter, which is what makes the version discipline above actually sustainable. Scout surfaces the product attribution SPINS already assigned; it does not re-tag SKUs or overrule the buyer's category definition.
Summary + further reading
- Product attribute tagging is a maintained SPINS layer over the UPC, product type plus ingredient, certification, and functional-benefit tags, and it is the natural-channel moat other syndicators have not matched.
- The layer's biggest risk is version drift: reclassification can move a segment apparent size 5 to 15% with no change in sales, so pin and note the attribute version before any year-over-year comparison.
- Attribute tags overlap and lag genuinely new products, so confirm how your own SKUs are assigned before trusting any segment total.
Related: What is SPINS data? · Category definition