Reading KeHE and UNFI movement data in SPINS

Why this matters

KeHE/UNFI movement data is how SPINS sees the long tail of independent natural retailers. After the direct-scan accounts, it's the biggest piece of the natural channel you have to work with. The catch: it reads nothing like direct POS data, and analysts who treat it that way get burned constantly.

Picture a natural-products brand whose KeHE/UNFI movement number in SPINS jumps +18% one week and craters -22% the next. Distribution hasn't moved. Velocity at the direct-scan retailers, Sprouts and Natural Grocers, was flat across both weeks. The KeHE/UNFI cut, on the other hand, is bouncing around like it's on a trampoline.

Most people call that noise and look away. That's a mistake, because the real explanation is more useful than "ignore it." The KeHE/UNFI numbers in SPINS aren't measuring what you probably think they are. They measure distributor shipments to retailers, not retailer sales to shoppers. Shipment data carries structural quirks that direct-scan data simply doesn't have.

So this page walks through those quirks: what they are, when they bite, and how to read KeHE/UNFI movement without getting played by the artifacts.

What KeHE/UNFI movement data actually is

If you want the broader context first, What is SPINS data? and Syndicated vs. panel data both cover it. Here's the part that's specific to KeHE UNFI movement data SPINS users keep tripping over.

SPINS can't see the independent natural retailer channel directly. The regional natural co-ops, the single-store naturals, the specialty grocers that don't license POS on their own. So it leans on distributor-flow data instead. Two distributors do the heavy lifting here:

• KeHE Distributors is a large natural and specialty food distributor with broad reach across natural retailers.
• UNFI (United Natural Foods, Inc.) is the other one. For years it was the primary distributor to Whole Foods Market. That relationship has shifted, and KeHE has picked up share at WFM recently.

What SPINS actually gets from these two is a log of what shipped out of the distributor warehouse to each retailer, by SKU and by week. SPINS turns that into an estimate of retailer sales. But the thing it starts with is a shipment, not a sale.

The whole approach rests on one assumption: over a long enough window, what ships roughly equals what sells, because retailers don't pile up inventory forever or run it to zero. Stretch that assumption down to the week-over-week level, though, and it falls apart faster than most analysts expect.

The four quirks that move the number

1. Retailer inventory cycles

A regional natural chain orders four weeks of stock in a single shipment to catch a distributor promo, then goes quiet for a month. The KeHE/UNFI movement read shows a fat spike in the order week and then near-zero for the next three. Consumer sell-through, meanwhile, was perfectly steady the whole time.

This is the biggest single source of week-to-week noise in distributor-flow data. Smaller retailers order in big infrequent batches, and that buying pattern produces lumpy movement reads at the individual-retailer level even when shopper demand hasn't budged.

2. Distributor pricing promo windows

The distributor runs a deal-of-the-month, and retailers stock up while the price is good. Shipments spike for the month, then sag the next month as those retailers burn through what they bought. Real economic activity, sure. But it's a working-capital decision by the retailer, not a change in what shoppers want.

3. New-retailer ramp and de-ramp

A new independent natural co-op picks up the brand. That first shipment might be two or three weeks of inventory, the initial fill, rather than one week of throughput. So the KeHE/UNFI movement runs hot for the first month or two while retailers fold the brand into their normal ordering cycles. Read those early weeks as "sales velocity" and you'll overstate how hard shoppers are actually pulling the product.

It works the same way in reverse. A retailer delists the brand, places one last order, then nothing. The trend keeps dragging that retailer's old performance for a month after the delist already happened.

4. Backfill from distributor data delivery

KeHE and UNFI don't always hand SPINS data on a clean weekly schedule. When a delivery slips, whether it's a vacation, a system issue, or a holiday week, the affected weeks show zero. Then the data lands, and several weeks of activity backfill all at once. Take that at face value and you see a trough followed by a spike. Nothing unusual actually happened.

The check here is the same one Reading SPINS panel coverage gives for direct-scan suppression: did ACV move when the sales number dropped? If it did, the drop is probably real. If it didn't, you're looking at a delivery artifact.

Reading KeHE/UNFI movement data in SPINS correctly

Five habits that strip out the noise.

1. Use a 4-week rolling average

At the individual-retailer or single-week level, distributor-flow data jumps around too much to read straight. Roll it to four weeks. That window is long enough to smooth the order-cycle lumpiness and short enough to still catch a genuine demand change.

A brand that reports "KeHE/UNFI flat on a 4-week rolling basis" over a 13-week stretch is reading the data right. Look at the same brand week by week and you'll see big swings that aren't really there.

2. Cross-check against direct-scan retailers

The direct-scan retailers in the Natural channel cut, Sprouts, Natural Grocers, and the regional naturals above the SPINS coverage threshold, measure actual POS. Compared with distributor flow, they're effectively noise-free. So if a brand's Sprouts and Natural Grocers reads sit still while the KeHE/UNFI cut whipsaws, the whipsaw is a distributor-flow artifact. It is not a real demand change at the long-tail naturals.

3. Separate KeHE and UNFI when possible

The two distributors cover overlapping but distinct retailer sets. KeHE has historically leaned specialty and gourmet, UNFI core natural, and the balance has shifted over time. Reporting them apart catches problems that hit one set and not the other, like a UNFI delivery delay that dings half the long tail and leaves the rest alone.

In practice the standard SPINS Natural channel cut blends the two. The separate distributor views live in custom cuts.

4. Treat the absolute dollar level cautiously

Shipments to retailers don't line up perfectly with sales to shoppers. SPINS' projection methodology bridges that gap, but the bridge isn't exact. When the absolute dollar figure actually matters, lean on distributor-flow movement for the trend and not the level.

5. Flag promo periods explicitly

Any promo in your analysis window, whether the brand ran it or the distributor did, drags inventory effects across the surrounding four to six weeks of KeHE/UNFI movement. Mark those weeks in the report and handle them with care.

Worked example: the apparent volatility

A wellness brand's weekly KeHE/UNFI movement reads across 8 weeks:

Read the raw KeHE/UNFI column week by week and it's chaos. The numbers run from $28K to $68K, a 2.4x swing.

Read the 4-week rolling column instead and it settles down: a $42K–$53K range, roughly flat, no real demand change.

The direct-scan column tells the same story. Steady across all eight weeks, $179K to $184K. Consumer demand isn't moving.

Put it together and the KeHE/UNFI week-to-week variance is just order-cycle and delivery-cadence artifact. The brand's actual demand in the independent natural channel never wavered. A team staring at the raw KeHE/UNFI weekly numbers would have burned the quarter chasing movement that was never there. A team reading the 4-week rolling next to direct-scan would have called it correctly: the brand held steady.

Where KeHE / UNFI movement data is the only signal

Here's the part that's easy to miss after several hundred words about noise: distributor-flow data is the only way SPINS, or any syndicator, sees the long-tail independent naturals at all. The single-store naturals, the regional co-ops, the specialty grocers below the SPINS direct-scan threshold, none of them license POS on their own. Strip out KeHE/UNFI flow data and the natural channel cut collapses to Sprouts plus Natural Grocers plus a handful of regional chains. Half the channel just disappears.

Four questions where distributor-flow beats direct-scan outright:

• Distribution breadth across independent naturals. Direct-scan is blind to the long tail. Whether the brand ships to 200 independent natural retailers or 1,200 is a number only distributor-flow can give you.
• Regional distribution patterns down at the long-tail level. KeHE and UNFI flow data picks up regional shipment patterns to independents that nothing else captures.
• Distribution onboarding speed for a new SKU. Launch a new SKU into the natural channel and distributor-flow data is your first read of which independents stocked it. Direct-scan won't catch it until the product reaches the bigger chains.
• Catching distribution loss before it shows up at the chain. Independents sometimes drop a brand ahead of the chain buyer. A trend break in KeHE/UNFI flow can warn you a month or two before direct-scan would.

For these questions, the noise this page spends most of its time teaching you to filter out is the actual information. What reads as noise for weekly demand turns into signal for distribution trends.

Anti-patterns

• Reading single-week KeHE/UNFI movement as a sales signal. It's too noisy. Roll it to four weeks, or make peace with the fact that you're chasing artifacts.
• Comparing distributor-flow ACV to direct-scan ACV head to head. ACV built from shipment data has different timing than ACV built from POS. A brand that looks "down in ACV at the long-tail naturals" may just be looking at a shipment lag, not a delisting.
• Adding KeHE/UNFI dollars to direct-scan dollars without saying so. The blended Natural channel total is fine for most reporting. But when accuracy carries weight, like a board report or an investor read, split the two and flag what share is distributor-flow.
• Treating distributor-flow promo lift as the same thing as direct-scan lift. A promo at a KeHE/UNFI-distributed retailer shows up in shipment data as stock-up-then-sell-through, with the lift smeared across four to six weeks. The same promo at Sprouts lands as a sharp spike in the promo week itself. Compare those two lift profiles without normalizing and you'll draw the wrong conclusion.
• Assuming the KeHE and UNFI relationships never change. The WFM dynamics have moved in recent years, and other major retailer-distributor pairings shift too. Check the current structure before running a longitudinal comparison that crosses one of those changes.

Doing this in Scout

Scout puts KeHE/UNFI movement in its own column, separate from the direct-scan retailers in the Natural channel cut, and the report view defaults to a 4-week rolling option. The KeHE-vs-UNFI split shows up wherever the SPINS extract carries it. Because anomalies in the distributor-flow column sit right next to the direct-scan column, the "real demand change or shipment artifact?" question turns into a glance instead of its own investigation.

Summary + further reading

• KeHE and UNFI movement data is distributor shipment data, not retailer POS. SPINS projects shipments into estimated sales, but the shipment number it starts from carries inventory-cycle quirks.
• Read it at the 4-week rolling level, not single-week, and cross-check direct-scan retailers to tell real demand changes from distributor-flow artifacts.
• Promo periods contaminate four to six weeks of distributor-flow data around them. Flag those windows.

Related: What is SPINS data? · Syndicated vs. panel data · Reading SPINS panel coverage