Measuring brand equity with retail scan data

Why this matters

Measuring brand equity usually means commissioning a survey, and surveys are expensive, slow and answer a question adjacent to the one a commercial team needs. Awareness and favourability are real constructs. Neither tells you whether you can hold a 40-cent premium when a private-label competitor arrives next quarter.

The commercially useful question is narrower: can this brand hold price, hold its shoppers, and earn distribution ahead of what its raw velocity would justify? All three are answerable from scan data you already receive, on a quarterly cadence, at no incremental cost.

None of the four measures below is brand equity itself. Each is a proxy that moves when equity moves, which is what makes them useful as a tracked series rather than as a single number. Read together they triangulate something a survey cannot: whether the brand's strength is currently being converted into money, and where it is leaking.

The methodology

Measure 1: price premium at matched velocity

The sharpest single signal. Find the closest comparable item, ideally private label in the same size and format, and compare price at similar unit velocity.

Premium = (your price - comparator price) / comparator price

The discipline is the matching. Compare like sizes, like formats and like distribution breadth, and check the velocities are genuinely close before reading the premium as equity. A brand priced 40% above a comparator selling three times the units does not have 40% of equity, it has a distribution problem and a price problem at once.

What to watch is the series, not the level. A premium eroding a point a quarter is the earliest reliable warning that equity is being spent, and it shows up long before a share number moves.

Measure 2: promotional dependence

Split baseline from promoted volume, and track baseline as a share of total.

Baseline share = baseline units / total units

A brand doing 85% of its volume at everyday price is being chosen. A brand at 45% is buying its volume, and its reported velocity is partly a trade-spend artefact. The post-promo lift and baseline method covers how to split the two credibly, which is the load-bearing step, because a sloppy baseline makes this measure meaningless.

The direction matters more than the level here too. Baseline share falling while total volume holds is the signature of equity being converted into short-term volume, which is a decision someone made and which is worth surfacing before it becomes structural.

Measure 3: distribution ramp

Compare velocity in stores where the item is newly distributed against stores where it is established, at matched store counts and comparable formats.

Ramp ratio = velocity in new stores / velocity in established stores

A high-equity brand arrives with demand: shoppers recognise it and buy it quickly, so a new store reaches a large fraction of established velocity within a quarter. A brand that takes three or four quarters to converge is being discovered rather than sought, which is a materially different asset.

This measure has a practical use beyond diagnosis. The ramp ratio is what should inform the velocity you promise a buyer in a new-distribution pitch, and it is more defensible than an average that mixes mature and new stores together.

Measure 4: repeat rate

Where basket or loyalty data exists, measure the share of buyers who purchase again within a defined window.

Repeat rate is the closest a transaction file comes to observing preference directly, and it separates trial from habit in a way none of the other three can. Its limitation is availability: most syndicated feeds do not carry it, and in convenience it usually requires the retailer's own loyalty data.

Where you cannot get it, market basket analysis gives a partial substitute. A brand appearing in stable, repeatable basket combinations is being bought habitually, even if you cannot follow the individual shopper.

Assembling the four into a read

Do not average them into an index. They fail in different directions and the pattern is the diagnosis:

The last two rows are the ones that change a plan. Recognition without satisfaction means the marketing is outperforming the product and more distribution will amplify a leak. Under-known but loyal is the opposite and is the best case for spending on distribution rather than on advertising.

Worked example

An illustrative better-for-you snack brand, tracked over four quarters against a private-label comparator.

Read the volume line alone and this is a good year: units rose from 34 to 36, up about 6%. Read the equity measures and it is a bad one. The premium fell ten points, from 40.2% to 30.2%, and baseline share fell sixteen points, from 84% to 68%. The brand bought its 6% volume growth with price and promotion, and the comparator gained units in every quarter regardless.

That is the third-and-fourth-row pattern: premium eroding and baseline falling together, which the table above reads as equity genuinely deteriorating rather than as ordinary competitive pressure.

The ramp check

Same brand, velocity in newly distributed stores as a share of established stores:

Reaching 88% of established velocity by the third quarter is a genuinely strong ramp. Shoppers in new stores find and buy this brand quickly, which says the name is working even while the price is not.

Putting the two halves together gives a specific diagnosis rather than a mood. The brand has real recognition, demonstrated by the ramp, and is spending it on price it did not need to spend. The 6% volume growth cost ten points of premium and sixteen points of baseline, and the ramp data suggests distribution would have bought that growth more cheaply than discounting did.

What measuring brand equity this way cannot do

Four limits, each of which has produced a wrong conclusion often enough to be worth naming.

It cannot separate equity from availability. A brand that holds its premium because no competitor is stocked next to it looks identical to one that holds it against full competition. Always check the comparator's distribution before reading a premium as strength, because the cheapest explanation for an unchallenged price is an unchallenged shelf.

It cannot see the brand outside the channel you measure. A brand strong in natural and weak in conventional has two different equity positions, and a blended read describes neither. Run the four measures per channel, and expect the answers to disagree.

It lags. Every measure here is a consequence of equity rather than equity itself, so all four move after the underlying change. A brand damaged by a recall or a supply failure will show intact premium and baseline for a quarter or more while shoppers work through habit. Treat a clean read after a bad event as insufficiently aged rather than as reassurance.

It says nothing about why. The measures detect that a premium is eroding. They cannot tell you whether the cause is a competitor's launch, a packaging change, a quality problem or a price increase that crossed a threshold. That question needs the qualitative work these measures were supposed to replace, which is the honest limit of the whole approach: scan data tells you when to go ask, and cannot answer for you.

The cadence that works

Quarterly, not monthly. All four measures are noisy at short horizons and the underlying construct genuinely does not move fast enough to justify a monthly read. A monthly series produces enough false signals that people stop believing the real one.

The exception is the premium series during a competitive event. When a new comparator lands, weekly for the first eight weeks captures the shape of the response, and that shape, whether the premium holds, dips and recovers, or dips and stays, is the single most informative thing you will learn about the brand that year.

Where to start if you have never run this

One brand, one category, one comparator, four quarters of history you already hold. Compute the premium series and the baseline share series only, and skip the ramp and repeat measures until the first two have told you something.

That is an afternoon of work and it answers the question most commercial teams are actually asking, which is whether the price they are holding is being held by the brand or by the absence of a challenger. Adding the other two measures is worth doing once the first two are running and someone is reading them.

A note on comparator selection

The comparator is the single most consequential choice in this method, and it should be written down and frozen at the start rather than picked fresh each quarter. A comparator chosen quarterly drifts toward whichever item makes the current number look reasonable, which is not a conscious decision and is nonetheless what happens. Freeze it, note why you chose it, and change it only with a written reason attached to the series.

Doing this in Scout

All four measures are the same computation repeated every period on stable definitions, which is where they usually break. A premium series computed against a comparator that was redefined in Q3, or a baseline split whose method changed, produces a trend line that reflects the analyst rather than the brand.

Scout connects your retailer and syndicated data and holds these as saved views that recompute as new data lands, so the definitions stay fixed and the series stays comparable. Repeat rate is the exception and depends on whether the underlying feed carries basket or loyalty data at all; the other three run on standard movement data.

Summary and further reading

• Four proxies compute from scan data: price premium at matched velocity, promotional dependence, distribution ramp and repeat rate.
• Track them as series rather than levels. Direction is the signal; the absolute number mostly reflects the category.
• Do not combine them into an index. The pattern across the four is the diagnosis, and averaging destroys it.
• Premium eroding together with baseline share falling is deterioration; either one alone is usually something less serious.
• A strong distribution ramp with a weakening premium means the brand is spending recognition it could have converted through distribution instead.

Further reading: price elasticity in CPG for the pricing half, post-promo lift and baseline for the baseline split, and the brand equity glossary entry for the definitions.