Benchmark by Segment: Are You Above or Below the Market?

Every store owner does benchmarking, even those who never used the word: they open the Instagram of the nearest competitor, compare the posting pace and the coupon of the week, and conclude they are doing well (or badly). The problem with that comparison is statistical, not about effort: it is a sample of 1. The e-commerce benchmark by segment swaps the mirror for a ruler: instead of measuring yourself against one store, you measure yourself against the median of a group in the same sector. This article explains how to compare yourself with the market the right way: which metrics make sense to aggregate, how to read the median and percentiles without fooling yourself, and what to do when the number says you are below.

Why comparing yourself with a single competitor misleads you

A sample of 1 has no distribution. You know your neighbor’s figure, but you do not know whether it is typical: it could be the most aggressive store in the segment or the most idle. And there is a silent selection bias: the competitor you chose to watch is usually precisely the noisiest one, because it was the one that showed up in your feed. Whoever posts more gets seen more, and becomes a reference more often than it should.

A concrete example: a supplements store that compares itself with the niche leader, which runs five campaigns a week with a team of twelve people, concludes it needs to quintuple its cadence. Against the median of the fitness segment, the same store could be sitting comfortably in the top half. The opposite mistake also happens: whoever compares themselves with a sleeping competitor feels safe until the day they discover the entire sector has sped up.

The numbers on Brazilian e-commerce help to size the game, but the national average mixes fashion with auto parts and supermarkets with SaaS. To decide promotion cadence, the useful cut is your own segment; and, within it, a group large enough to dilute outliers.

What makes sense to aggregate in a benchmark by segment

Not every metric gets better when aggregated. The practical rule: aggregate behavior, not scale. Behavior (how many campaigns per week, which type of campaign, on which network) is comparable between a 3-person store and a 300-person one. Scale (revenue, follower count, media budget) is not: aggregating it would only produce a number nobody can use.

A one-off survey, like the discount benchmarks by category, answers “what is the typical depth of my market right now”. That is the photo. The continuous benchmark answers a different question: “am I above or below my sector this week, and is the gap widening or closing?”. That is the film: the same indicator, calculated the same way, every week, against the same group. The two complement each other, but do not replace each other.

Median, P75 and P90: how to read them without fooling yourself

Why median and not average? Because the average breaks with a single outlier, and every segment has one. Imagine five stores with 2, 3, 3, 4 and 30 campaigns in the week. The average is 8.4; the median is 3. If you ran 4 campaigns, you are above more than half the group, but the average would say you are “below the market”. In a sector benchmark, the average punishes everyone for the behavior of one.

In practice, each number serves a different decision:

• Median (P50): where the typical participant sits. It is your ruler of normality.
• P75: the floor of the most active quarter. A realistic target for anyone who wants to gain share of attention without becoming a different company.
• P90: the territory of the most aggressive 10%. Read it with suspicion: a very high cadence can be margin burning, not excellence.

Being at the P90 is no trophy. If your segment goes into a coupon war, the P90 rises, and chasing it means joining the war too. A high percentile describes behavior; it recommends nothing on its own.

How a continuous, anonymous, opt-in benchmark works

The mechanism matters as much as the number, because it is what guarantees nobody sees your individual data. In Batedor, the sector benchmark works like this, on the Sector comparison page of the dashboard:

1. You declare your sector. There are 18 options (fashion, beauty, home and decor, electronics, pet shop, SaaS, among others) and an explicit opt-in authorizing the anonymized use of your metrics. You can turn it off at any time.
2. The aggregation runs every week. On Monday, the system closes the previous week and calculates, for each sector, the median, the P75 and the P90 of campaigns detected per monitored competitor, plus the mix by type and by platform.
3. Nothing is published with fewer than 3 participants. If the sector does not reach the minimum, that week’s benchmark simply does not exist. And only aggregated statistics appear: median and quartiles, never the figure of an individual account.
4. The dashboard positions you. A “You” block compares your median of campaigns per competitor with the sector’s and delivers the reading ready-made: “18% above the sector median”, or “in line with the sector” when the difference is smaller than 10%.

If you do not use the dashboard yet, the 14-day trial with no card already includes the benchmark page: choose your sector, enable the opt-in and the comparison appears as soon as your segment reaches the minimum number of participants. A note of honesty: in a small sector, that can take a few weeks.

You are below the median: now what?

The first reaction (crank everything up) is usually the wrong one. Below the median is a diagnosis with three possible causes, and each one calls for a different answer.

1. Separate pace from mechanic. A below-par cadence with a mix similar to the sector’s is a pace problem, and the answer is operational: calendar, automation, a campaign plan. A cadence in line but a mix concentrated where the sector is not (you only run direct discounts, the sector lives on free shipping and launches) is a mechanic problem.
2. Check the channel. If the sector’s mix by platform concentrates activity on Instagram and your bet is somewhere else, the gap may be about channel, not effort. Changing the stage is cheaper than doubling production.
3. Decide whether you want to close the gap. Being below the median with a healthy margin and high repurchase can be positioning: a premium brand promotes less by choice. The benchmark tells you the distance; whether it is a problem is decided by your strategy.

And what if you are well above? Above the P90, the question flips: how much of that cadence is buying share of attention, and how much is just burning margin the sector does not require?