An outlier is a finding. A difference is not
Your review count is lower than theirs. So what: they are older, or they ask harder, or they sell more units. One comparison cannot separate those.
Now put fourteen brands in the same category next to each other. If twelve sit in a band and you sit well outside it, that is a finding with a direction attached, and it is worth an afternoon. If you are inside the band, the original difference was noise dressed up as a gap, and the meeting it would have caused can be skipped.
The value of a pack is exactly this: it converts differences, which are everywhere, into outliers, which are rare and actionable.
Read the position, not the average
The mean of a category is a number that belongs to no company in it, and chasing it produces a slightly worse version of everybody.
Read where you sit instead: comfortably inside, at an edge, or outside entirely. An edge is a decision somebody made, yours or theirs. Outside is either your best asset or your most expensive problem, and it is worth knowing which before somebody else points it out.
The other thing a pack gives you is what the band itself looks like. A category where every brand's entry price sits within a few percent is a market competing on something other than price, and that tells you where to fight.
Three ways this goes wrong
Scales that are not comparable. Some indices normalise each brand against its own peak, so a hundred means "their busiest week" and not "bigger than the others". Put two of those in the same column and the chart is confident nonsense. We treat that as a blocking check rather than a footnote, because it looks perfectly reasonable on the page.
Mixed instruments. Measured facts and modelled estimates in the same ranking produce a ranking of methods. Label them and keep estimates out of anything that gets scored.
A pack that is not a category. Fourteen brands that sell to different buyers are not a comparison set, they are a list. Group by who the buyer is and what they substitute, then compare inside the group.
And the quiet one: a pack of four is not a pack. With a small set, one unusual brand moves the whole band, and a position inside it means very little.
Which measures are worth ranking at all
The ones where being unusual implies an action. Entry price and the width of the range. Review volume against age. Publishing and ad cadence. Catalogue depth. Share of creative carrying a discount. Presence in AI answers.
The ones where being unusual implies nothing: follower counts, badge walls, press hits. They rank beautifully and correlate with very little, which is exactly why they end up on so many category dashboards.
The same filter applies here as anywhere else. If a rank would not change what you do in either direction, it is occupying room a useful measure could have had.
How to read it here
The Benchmark shows every tracked brand rather than only the pair you named, with a category grouping so a comparison stays inside a like set, estimates labelled where the figure is modelled, and each number clickable through to the page it came from. Details holds the evidence per brand.
Read it quarterly, not weekly. Benchmarked measures are lagging by construction, and reading them on a Monday teaches nothing except how to overreact, for reasons that apply to every number on the page.
Pick one competitor. See what we find.
Your first benchmark is free. One domain, measured against you, no card and no call.