Pricing analytics a category owner can act on

Four questions, four numbers. Where do we sit against the market, and by how much? On how many SKUs are we lowest, matched or above? Is that spread tight, or two different populations averaged together? And what does closing a gap cost in margin? Price index answers the first, position mix the second, gap distribution the third, margin impact the fourth. All four are computed from the timestamped observations collected by price monitoring, so any figure traces back to the page it came from.

What pricing analytics has to measure

A pricing dashboard fails in one of two ways: a single average that hides everything interesting, or forty charts and no way to tell which one to act on. These five measures each change a different decision, and each carries a trap worth knowing before you present it.

Price index, with the arithmetic shown

price index = (your price / market price) * 100

market price = mean or median of matched, in-stock competitor listings

Worked example. Your price is $189.00. Four matched competitors are in stock at $179.00, $199.00, $205.00 and $184.00. Their total is $767.00, so the mean is $191.75. Your index is 189.00 / 191.75 = 0.9857, times 100, so 98.6. You are 1.4 percent below market on that SKU.

Why one listing can rewrite the answer

Add a fifth listing: a marketplace seller clearing stock at $129.00. The total becomes $896.00 across five listings, so the mean is $179.20 and your index moves to 189.00 / 179.20 = 1.0547, which is 105.5. The same shelf price now reads as 5.5 percent expensive instead of 1.4 percent cheap, and nothing about your business changed.

So the market price is configurable. You choose mean or median, whether out-of-stock listings count, whether marketplace sellers count alongside authorized retailers, and which competitors define the market at all. The median of those same five listings is $184.00, giving 189.00 / 184.00 = 102.7. Neither number is wrong. Pick one rule and hold it, so week-on-week movement means something.

Rolling up to a category without lying

A category index is a weighted average of SKU indexes, and the weight matters. Take two SKUs: one indexes at 96 on $40,000 of revenue, the other at 108 on $10,000. Unweighted, the average is 102, which says the category is expensive. Weighted by revenue it is (96 x 40,000 + 108 x 10,000) / 50,000 = 98.4, which says the opposite. Revenue weighting is the default here, because an unweighted number lets a long tail of low-volume SKUs outvote the products that pay the bills.

Position mix and gap distribution

One number for a whole catalogue is good for a board slide and useless as a work queue. Position mix is the count version of the same question.

Example: across 500 monitored SKUs you are lowest on 120, matched within your tolerance on 95, and above on 285. That is 24 percent lowest, 19 percent matched, 57 percent above. A catalogue can index at 99 and still sit above market on more than half its SKUs, because a few deeply discounted products pull the average down.

The shape behind the average

Gap distribution buckets every SKU by its gap to the cheapest in-stock competitor: more than 10 percent below, 5 to 10 below, 0 to 5 below, at parity, 0 to 5 above, 5 to 10 above, more than 10 above. Two categories can both average 3 percent above market. In the first, every SKU sits between 2 and 4 percent above, a deliberate premium. In the second, half are at parity and half are 7 percent above, a backlog nobody worked through.

The far right bucket is the useful one, and it exports as a list rather than a chart. SKUs more than 10 percent above the cheapest in-stock competitor are where a rule in repricing earns its keep. The price position analyzer runs the same bucketing on a CSV if you want the shape before connecting a store.

Category volatility sets your cadence

Volatility is competitor price changes per SKU per week, reported alongside the median size of a change. If 200 SKUs in a category recorded 640 competitor price changes in a week, that is 3.2 changes per SKU. A category recording 0.4 is a different world and should not be watched the same way.

• It sets refresh rate honestly. Four-times-daily collection on a category that moves twice a month buys noise. Paying for hourly refresh where prices move several times a day does not.
• It sets alert thresholds. In a volatile category a 1 percent threshold produces an unreadable feed. Start at 3 percent for general merchandise and tighten only where the data earns it.
• It separates one reseller from a market. Volatility broken out by seller shows whether a category is genuinely moving or whether one automated reseller is oscillating around your price. Those need opposite responses.

Margin impact: the arithmetic that stops a bad cut

Every other measure here is about position. This one is about whether you can afford it. Load unit costs by CSV or from your store, and each proposed change is reported as a change in unit contribution, not only in percent margin.

Worked example. A product costs $118.00 and sells at $189.00, so unit contribution is $71.00 and gross margin is 37.6 percent. Matching a competitor at $179.00 leaves $61.00 of contribution and 34.1 percent margin. Percent margin fell 3.5 points, which sounds tolerable. Contribution fell 14.1 percent, which does not. At 100 units a week, weekly contribution goes from $7,100 to $6,100, and holding the old total needs 7,100 / 61 = 117 units, a 17 percent volume lift just to stand still.

The weekly digest, and why it is short

Live alerts go to people who act inside a day. Most stakeholders do not need that, so a weekly digest goes to email or Slack with five things and nothing else: how your index moved by category, SKUs that crossed from below market to above or the reverse, the largest competitor moves, MAP violations opened and closed, and any source whose data degraded. That last line comes straight from data quality, because a pricing summary that cannot say whether collection was healthy is a summary of unknown data.