How to read a benchmark

A single benchmark is not a scoreboard number. It is a picture of the whole market for one metric, with your own position marked on it. The point is never "here is your labor ratio." The point is "…

11 min read·Updated July 11, 2026
On this page

What a benchmark tile shows

A single benchmark is not a scoreboard number. It is a picture of the whole market for one metric, with your own position marked on it. The point is never "here is your labor ratio." The point is "here is what every comparable operator's labor ratio looks like, and here is where you sit inside that spread." Reading a benchmark well means reading the distribution first and your dot second.

Every benchmark on the Benchmarks tab, whether it renders as a scan row in the main list or as a tap-to-open tile, carries the same five things:

  • The metric, named plainly, with the cohort it is drawn against right underneath it (National, Region, Group, or Your Offices) and how many operators stand behind it, for example "National · 40 operators." Verinode writes the plain count, never the statistician's "n=40."
  • The peer distribution, drawn as a horizontal strip: a middle-half band, a median tick, and the axis it all sits on. This is the market.
  • The median caption above the strip, reading "Peer median" (or "Industry median" on the network view) with the number beside it.
  • Your value, shown two ways at once: as a colored dot dropped onto the distribution strip, and as a number in the right-hand column.
  • The direction color, green or red, that tells you at a glance whether your number is the good side or the bad side of the market for that metric.

Verinode does not grade you or tell you what to do about the number. It reads the data you already have, positions it honestly against anonymized peers, and leaves the read to you. Those peer numbers are trustworthy for one reason: Verinode is an independent data trust, and operator data is never sold to carriers.

The distribution strip, piece by piece

The strip is the heart of the benchmark. Left to right it is a number line for that metric. Three marks sit on it.

  • The middle-half band is the shaded rectangle. Its left edge is the 25th percentile (P25) and its right edge is the 75th percentile (P75). Half of all comparable operators fall inside this band. It is the "normal range," the fat middle of the market. The gap between its edges is the inter-quartile range, or IQR.
  • The median tick is the short vertical line, usually near the center of the band. This is the 50th percentile (P50), the exact middle operator. Half the market is worse than this line, half is better.
  • Your dot is the round marker. Where it lands relative to the band and the median is the entire read. Inside the band, you are in the normal middle. Past either edge, you are in the top or bottom quarter of the market.

Note

Some metrics have no spread to draw. A share-of-revenue style metric where every operator sits near the same value collapses to a single median line with no band, because a middle-half range would be misleading. That is expected, not a rendering bug: the median is still the honest reference.

What P25, P50, and P75 actually mean

Percentiles sound technical but the plain-language version is short.

  • P50, the median, is the middle of the pack. Line every comparable operator up worst to best and P50 is the one standing in the exact middle. It is the fairest single "typical" number, more honest than an average because one giant outlier cannot drag it around.
  • P25 is the operator one-quarter of the way up from the bottom. A quarter of the market is below it.
  • P75 is the operator three-quarters of the way up. A quarter of the market is above it.

So the middle-half band running from P25 to P75 is where the ordinary half of operators live. If your dot sits inside it, you are unremarkable in the honest, reassuring sense: normal for the market. If your dot is past P75 on a metric where higher is better, you are in the top quarter. If it is below P25 on that same metric, you are in the bottom quarter and this is a number worth a second look.

Verinode derives these display quartiles from the cohort's own mean and spread, so the P25 and P75 edges move as the real distribution moves. For counts that cannot go negative, like days-to-pay or dollar figures, the low edge is floored at zero rather than drawn into an impossible negative.

Ahead or behind, in one glance: the color

You should never have to do arithmetic to know which side of the market you are on. The color does it for you.

  • Green (the Expand signal) means you are at or better than the median for that metric.
  • Red (the Analyse signal) means you are worse than the median.
  • Copper means Verinode has your number but not yet a peer median to judge it against, so no good-or-bad claim is made.

The critical thing the color handles for you is direction. For some metrics higher is better (net margin, collection rate). For others lower is better (days to pay, labor as a share of revenue, cost ratios). The tile knows which is which. So a dot sitting to the left of the median can be green (good) on a "lower is better" metric and red (bad) on a "higher is better" one. Never read the raw left-or-right position as good or bad. Read the color, and let it account for the metric's direction.

Heads up

Below the median is not automatically a problem, and above it is not automatically a win. The color tells you the direction-correct read, but context still matters: paying below the market labor rate reads green as a cost, yet can be a retention risk. The color is the fast read. The deep-dive is where you check whether it is headroom or a warning. See the benchmark deep-dive.

The value and dollar columns

To the right of the strip, the row spells out the same read in numbers.

  • Your value appears in the value column, colored the same green or red as your dot, with "You" beneath it (or "Network" on the HQ office view). This is your own figure, the one being positioned.
  • The dollar impact, when it applies, is the last column: a signed figure like "+$18k vs Peer / Yr" in green, or the same in red with a minus sign. This is a rough translation of the gap between you and the peer median into annual dollars, and it only appears for revenue-scalable percentage metrics such as margin, collection rate, and cost ratios. It is deliberately never shown for a compliance or safety percentage, where turning percentage points into dollars would be nonsense. Below a small threshold the column stays quiet rather than crying wolf over pocket change.

Peer read vs industry read

A benchmark can be built from two different reference layers, and the label always tells you which one you are looking at.

  • The peer read is the real thing: an anonymized median drawn from operators like you who have consented to benchmarking, once that cohort clears Verinode's minimum sample and anonymity threshold. It is captioned "Peer median" and footed with the operator count and scope. This is the moat, the number you cannot get anywhere else.
  • The industry read is the honest fallback for when a peer cohort has not formed yet. Rather than show you a dead "cohort building" placeholder, Verinode drops in a published industry reference for an operator your size, captioned "Industry typical" or "Industry median." It is grounded, source-attributed reference data, safe to show from day one, and it is clearly a different thing from a peer number so you are never misled into thinking a survey figure is your peers.

The distinction is not cosmetic. A peer median is your actual market. An industry typical is a stand-in until enough peers contribute. As the cohort fills, the industry reference gives way to the real peer distribution. Which cohort the peer read is drawn against, National, Region, or your Group, is set by the scope picker over the list. See cohorts and scope for how that resolves.

"Show my position," or hide it to read the market clean

At the top-right of the Benchmarks tab is a Show my position toggle. It is the signature control of the whole data room.

  • On (the default) overlays you onto every distribution: your dot on each strip, your value in the value column, the dollar delta in the last column. This is the "where do I stand" read.
  • Off hides all of that. Your dots disappear, and the value column swaps to show the market's own median (or the industry number) instead of your figure. Now you are reading the clean market distribution with no ego in the frame, which is exactly what you want when you are studying the shape of the market itself rather than scoring yourself against it.

That is what the tab subtitle means by "Hide it to read the market clean." On the HQ network view there is no single "you" to overlay (the rows are the network's own offices distributed), so the toggle is simply absent and the strips read as the network's own spread.

The same benchmark shows up in two shapes depending on where you are.

  • On the main Benchmarks tab, metrics render as dense scan rows grouped into categories (Profitability, Cash Velocity, Cost Structure, and so on), so you can run your eye down a whole category in seconds. A Cards / Table switch lets you flip the same data between the visual row layout and a plain comparison table with You, Peer, Delta, Cohort, and count columns.
  • On surfaces like the Margin page's Benchmarks slice, the same metrics render as tap-to-open hero tiles in a grid. Each tile leads with a large count-up hero number (your value, colored by direction), a percentile pill such as "72nd · Top Quartile," "Above Median," "Below Median," or "Bottom Quartile," a comparison bar with the middle-half band and the median line, and a plain-language dollar-gap line ("Closing the gap is worth $X/yr," or "You're beating the median by $X/yr"). A trust footer names the reference and the sample.

In the tile grid, the first, most important metric of each category, its net margin, its collection rate, its labor ratio, is promoted to a featured tile at double width. That is a deliberate hierarchy cue: the wide tile is the headline number of that category, the one to read first before scanning its siblings.

Whichever shape you are looking at, tapping a benchmark opens the full deep-dive: the complete distribution, methodology, history, and sources. That is covered in the benchmark deep-dive.

Every empty and gated state

Benchmarks are honest about what they do not yet know, so you will meet several partial states. None of them is a broken screen.

  • You have no value yet, but a reference exists. The tile shows the industry (or peer) typical number as its hero, tags it "Industry typical" or "Peer typical," and draws your own marker as a dashed ghost at the start of the bar with a one-tap "You, add data." The board reads as a populated market with your line as the single empty slot, and your real number lands there the moment data flows in.
  • You have no value and there is no reference. The tile reads "Add data to see where you stand."
  • You have a value but no peer cohort yet. The tile reads "Comparison loads when the cohort fills." In the row layout the strip reads "Cohort building. Peer data appears once it clears the consent and sample threshold."
  • A curated sample, not your full access. Some peer numbers carry a "Sample, unlock the full set" pill. That is a single headline stat shown to everyone as a teaser. You unlock the full family by contributing your own data or becoming a Member. See access and gating.
  • A category with nothing to say hides itself. Rows that can show no peer distribution, no industry reference, and no value of your own are dropped, and a category left entirely empty disappears rather than filling the page with dead ghosts. A shorter board is the honest one.

Tip

If a whole category is missing that you expected to see, it is almost always a data-coverage gap, not a defect. The metric needs a source you have not fed yet (a P&L, timesheets, lead data). Forward the document and the category lights up, positioned against peers from the first upload.

Reading one benchmark end to end

Say you open Cost Structure and land on the labor ratio row. The strip shows a middle-half band with the peer median tick near its center, captioned "Peer median 31.0%," footed "National · 40 operators." Your dot sits to the right of the band, past P75, and it is red, because for labor-as-a-share-of-revenue lower is better, so being in the top quarter of the market is the wrong quarter to be in. The value column reads your figure in red under "You," and the last column reads a red dollar figure "vs Peer / Yr," the rough annual cost of that gap. In two seconds you know: your labor share is heavier than three-quarters of comparable operators, and roughly what it is worth to close. Flip Show my position off and the same strip reads as the clean market, your dot gone, so you can study how tightly the market clusters before deciding whether your gap is a crisis or a rounding error. Then tap the row to open the deep-dive and work out why.

Data sources

Data sources

  1. 1.Your own operating metrics, from the data you have fed in. Your business.
  2. 2.Anonymized peer distributions across consenting operators. Verinode network.
  3. 3.Published industry reference figures, source-attributed. Research.
Was this helpful?