What market share is
Market share is one brand's sales stated as a percentage of total sales in a defined category, over a defined period, across a defined set of stores. The arithmetic is a division. The three definitions in front of it are where every argument about market share actually happens.
I learned that in a category review where our brand walked in at 24.5% and the buyer had us at 21.0%. Nobody had made an error. We were quoting dollar share and he was quoting unit share, and our shelf price sat 50 cents above the category average, which is enough to open a three and a half point gap on the same 52 weeks of the same data.
How to calculate market share
Brand sales divided by category sales, in the same currency, for the same weeks, over the same stores:
Market share = brand sales / total category sales
Every hard part is in the denominator. Change which stores feed it and the number moves. Change which items count as the category and it moves again. Change from dollars to units and it moves a third time without a single transaction changing.
Dollar share and unit share are different numbers
Here is a refrigerated-kombucha category read, latest 52 weeks, one regional grocery banner. Wildroot and Copper Hill are made up; the arithmetic is not.
| Brand | Dollars | Units | Average price | Dollar share | Unit share |
|---|---|---|---|---|---|
| Copper Hill | $5,712,000 | 1,904,000 | $3.00 | 47.6% | 47.6% |
| Wildroot | $2,940,000 | 840,000 | $3.50 | 24.5% | 21.0% |
| Banner own-brand | $2,100,000 | 875,000 | $2.40 | 17.5% | 21.9% |
| All other | $1,248,000 | 381,000 | $3.28 | 10.4% | 9.5% |
| Category | $12,000,000 | 4,000,000 | $3.00 | 100.0% | 100.0% |
Run it: 2,940,000 / 12,000,000 = 24.5% dollar share, and 840,000 / 4,000,000 = 21.0% unit share. Copper Hill prices exactly at the $3.00 category average, so its two shares are identical at 47.6%. The banner's own brand does the reverse of Wildroot: 21.9% of the units at 17.5% of the dollars, because it rings 60 cents under the category average.
The read to take from that table is that dollar share is a revenue claim and unit share is a demand claim, and price sits between them. A brand whose dollar share grows while unit share falls has taken price, not shoppers. That is a perfectly good quarter and a bad twelve months, and reporting only the dollar line hides which one you are having.
Which number to lead with depends on who is asking. A buyer allocating shelf space is buying units, because units consume facings. A finance team modelling revenue is buying dollars. Bring both, and say out loud when they disagree.
The category boundary sets the answer
Before the division there is a decision: what counts as the category. Move kombucha's boundary out to include refrigerated functional beverage and Wildroot's 24.5% falls by more than half, because the denominator triples while the numerator does not move at all. Pull it in to glass-bottle kombucha only and the share jumps.
Neither boundary is wrong. Both are answers to different questions, which is why how you define a category has to be settled and written down before anyone quotes a share number. The rule I use: the category is the set of items a shopper would consider substitutes for each other, and if two people on the call disagree about a SKU, the share number is not ready to present.
This is also where fair share enters. Fair share compares your share of category sales against your share of category items or facings, so it inherits every boundary decision above it. The mechanics of that comparison sit with category management.
Why the same brand shows a different market share in two panels
The most common market share dispute is not a calculation error. It is two reads with different store universes underneath them.
SPINS defines its MULO (Multi-Outlet) channel as a combination of conventional grocery, drug, mass merchandise, dollar and military stores, and puts it at over 110,000 retail locations. A natural-channel read covers a few thousand stores that skew heavily toward the shopper an emerging brand actually sells to. Same brand, same weeks, two denominators that are not remotely the same size or the same shape.
| Read | Brand dollars | Category dollars | Market share |
|---|---|---|---|
| Regional banner, natural set | $2,940,000 | $12,000,000 | 24.5% |
| Same brand, broad multi-outlet read | $3,360,000 | $196,000,000 | 1.7% |
Wildroot's dollars barely move between the two, up 14% because the wider read picks up some conventional distribution. The share collapses from 24.5% to 1.7%, because 3,360,000 / 196,000,000 = 1.7%. Both numbers are correct. Only one of them answers "am I winning where I compete", and the other answers "how much of the total category do I own".
Three things to check before you accept that two panels disagree:
- Store universe. Which banners report, and which are projected rather than measured. SPINS panel coverage walks the projection and suppression mechanics.
- Category definition. Each provider maintains its own hierarchy, and a brand can sit in different parent categories in each one.
- Period alignment. A 52-week read ending on a Saturday and one ending on a Sunday are not the same 52 weeks, and neither is a calendar year against a retail one.
Market share, share of shelf, and share of wallet
Three share metrics get used interchangeably in the same meeting, and they measure three different things:
| Metric | Numerator | Denominator | What it tells you |
|---|---|---|---|
| Market share | Your category sales | Total category sales | Position against competitors |
| Share of shelf | Your facings | Total category facings | Physical presence at shelf |
| Share of wallet | Your sales to a buyer group | That group's total category spend | Depth of a customer relationship |
A brand at 24.5% dollar share and 14% share of shelf is over-performing its space and has a real expansion case. The reverse is a SKU-rationalization conversation you would rather start than receive. Share of wallet is the one that needs a buyer-level denominator, which scan data alone does not carry.
Where Scout fits
Market share is arithmetic. The work is holding the denominator still: same category definition, same store set, same weeks, every time somebody asks. Scout connects your retailer POS or syndicated feeds and computes dollar and unit share against a category boundary you set once, so two people pulling the same number in the same week get the same answer. It reports the share and shows the denominator behind it. It does not tell you which boundary is the right one for your business, because that is a category decision rather than a data one.
The short version
- Market share is brand sales divided by category sales, over a stated period and a stated store set. The three definitions matter more than the division.
- Dollar share and unit share diverge by exactly as much as your price differs from the category average: 24.5% versus 21.0% for a brand priced 50 cents above a $3.00 average.
- The category boundary is a decision, not a fact. Widen it and your share falls without a single unit changing hands.
- Two panels can both be right about the same brand. A 24.5% read on a natural store set and a 1.7% read across a 110,000-store multi-outlet universe are answers to different questions.
Sources: SPINS, "MULO (Multi-Outlet) Channel".