Demand problem or distribution problem? How to read your scanner data

CPG Consulting · 8 min read

The call usually comes on a Friday. Sales are down 30 or 40 percent versus last year, the board deck is due, and everyone has a theory: the packaging is stale, the price is wrong, the category is dying, a competitor is eating us alive.

Sometimes one of those theories is right. But in our experience tearing down emerging brands' Circana and Nielsen data, the single most common finding is simpler and stranger: the demand never went anywhere — the distribution did. The brand lost doors, facings, or entire retailers, and the topline collapse everyone attributed to shoppers was actually a shelf-space story.

This matters because the two problems have opposite fixes. A demand problem needs marketing, pricing, or product work. A distribution problem needs sales work — winning back doors and defending the ones you have. Spend a year fixing the wrong one and the brand may not get another year.

The one division that makes the diagnosis

Take your total dollar sales and divide by your total distribution points (TDPs) for the same period. That gives you dollars per point of distribution — your velocity. Trend it weekly or monthly alongside raw sales, and one of three pictures emerges.

Picture one: sales down, velocity flat or up. Distribution problem. Your product sells fine wherever it's still on shelf; there's simply less shelf. Find the lost doors — a delisting, a distributor change, a reset that cut your facings — and the fix is a sales conversation, often a winnable one, because your velocity is your best argument.

Picture two: sales down, velocity down, distribution flat. Genuine demand problem. Same shelf presence, fewer shoppers choosing you. Now the packaging, pricing, and competitive theories deserve the meeting.

Picture three: sales up, velocity down. The sneaky one. You're expanding into doors faster than demand supports — growth that reads great in a press release and sets up a delisting wave 12 months out. This is the picture to catch early, because it looks like winning.

Where brands go wrong with the data

They compare months instead of indexing against the category. An October drop means nothing by itself — if the category dropped further, you gained share. Always ask the data "versus the category" before "versus last month."

They read national when the fight is local. A national trend is an average of retailer-level battles. Your velocity at your most important chain is a more actionable number than any total-US figure.

They let promo pollute the read. If half your volume is sold on deal, your "velocity" is partly a record of your trade calendar. Split base and incremental volume before drawing conclusions about underlying demand.

What to do with the diagnosis

If it's distribution: build the door-loss map (which retailers, which items, when), then take your velocity numbers back to those buyers. "Our $/TDP was above category average when we were cut" is a genuinely strong re-entry pitch.

If it's demand: resist the urge to fix everything. Rank the suspects — price gap versus the category anchor, promo dependence, review scores, share of voice — and test one at a time where you can measure the result.

Either way, the diagnosis costs an afternoon with data you're probably already paying for. The wrong guess costs a year.