Coverage labels and the anonymity floor

What this page is protecting

Every number on the Benchmarks tab is doing two jobs at once: telling you something true, and refusing to tell you something it shouldn't. Coverage is the first job. It is Verinode's honesty check on your OWN network: has enough of your own franchise data come in to trust a distribution, or is a number still too thin to lean on? The anonymity floor is the second job. It is the wall between your network and everyone else's: how many operators outside your network have to be behind a reference number before Verinode will show it at all, so that number can never be traced back to one operator's real business.

Both live on the same page for a reason. A benchmark that is confidently wrong (drawn from one or two offices) is as useless as a benchmark that leaks (drawn from one outside operator dressed up as a "median"). Verinode would rather show you nothing than show you either.

Where to find it

Open Benchmarks from the HQ sidebar, under Intelligence, or go directly to hq.verinode.ai/benchmarks. It is the same four-tab hub operators see on IQ, Benchmarks, Ratings, Library, Industry News, with the Benchmarks tab reshaped for your seat: instead of "you vs. your peers," it is "your offices vs. each other, with the wider industry as one reference line." That reframing, and the full tour of the tab's categories and controls, is covered in Benchmarks: how HQ reads the network. This article is narrower: it is about the two gates that decide what you actually see.

Coverage: is there enough of your own network to trust this number

Every metric card on the Benchmarks tab, Gross Margin, Days to Pay, Cycle Time, and every other row across Profitability, Cash Velocity, Cost Structure, Service Speed, and the rest, is really a distribution across your own offices. A distribution needs a handful of data points before it means anything. One office reporting a number is a fact about that office, not a "network median." Verinode moves a metric through a confidence ladder as more of your own offices report it:

• No data. Nobody in your network has reported this metric yet. The card simply does not appear, there is no placeholder tile sitting there labeled "no data."
• Early signal. A couple of offices have a value. This is too thin to call a shape, so Verinode still holds the card back rather than show you a "distribution" of two points.
• Indicative. Enough offices are reporting that a real shape starts to appear, a genuine P25-median-P75 spread. This is the point where the card starts showing up on the Benchmarks tab. Read it as a direction, not a verdict.
• Observed. A solid slice of your network is reporting. The spread is dependable enough to make decisions from.
• Verified. Most or all of your network is reporting. This is as solid as a network read gets.

You will see this exact ladder, in these words, in two other places in HQ: the Margin & Cash tiles label a metric Awaiting Data, Early Signal, Trending, or Observed/Verified depending on how much of your network is behind it, and the Custom Cohort Builder (below) prints No data, Early signal, Indicative, Observed, or Verified directly on its result tile. It is the same underlying ladder every time, Verinode does not invent a new confidence scale per screen.

On the Benchmarks tab itself, the ladder does not show up as a printed badge on the tile, it shows up as a gate. A metric that has not cleared Indicative simply is not on the page yet. As your offices flow more data in, cards you have never seen before start appearing on their own, and you do not need to do anything to unlock them.

What you see once a card is open

Tap a metric tile and its deep-dive opens with an Across Your Offices panel. At the top, a count reads how many of your offices have reported this metric ("6 offices"), plus, when it is available, a single Industry figure alongside it, the broader-market reference described below. Below that sits a horizontal bar: an IQR-shaded band running from your network's P25 to P75, with Median centered and labeled underneath in the actual unit (percent, dollars, days). Below the bar is the ranked list itself, every reporting office by name, best to worst for that metric, each row showing its rank, its name, and its value. A value colored green sits in your network's strong quartile for that metric; red sits in the weak quartile; anything in between reads in plain foreground color. This is how you see, at a glance, which office to coach first on a given metric and which one to learn from.

Empty state. If no office has reported the metric yet (which should not normally happen, since the card would not have appeared at all), the panel reads: "No offices have reported this metric yet."

Underneath the office panel, a How It's Calculated note explains the read in plain language, something like: median across your reporting offices, with each office ranked; the band is your network's own P25 to P75 spread; the industry line, when shown, is the median of the operators contributing to it outside your network, a reference, never a competitor. A Data Sources panel beside it either names the outside peer cohort behind the Industry line (when one exists) or reads: "Sources will appear once cohort sample size and research coverage land for this metric."

The anonymity floor: the wall around everyone outside your network

Your own offices are named because they are yours, Verinode already has a direct relationship with every one of them, and ranking them against each other is the entire point of this page. The Industry reference line is a different animal. It is the median of operators who are not in your network at all, independent shops, other franchise systems, competitors you will never see the name of. Showing that number at all is a promise Verinode has made in its data-use terms: a benchmark never gets published from a peer group too small to hide inside. Below a certain number of contributing outside operators, publishing a "median" would really be publishing one operator's, or a couple of operators', real number with a label on it. Verinode will not do that, so instead of a thin number, you simply see nothing.

That threshold is not the same for every metric. Financial and cost-structure numbers, margin, labor ratio, material ratio, loaded labor rate, and similar figures pulled from an operator's books, are treated as more sensitive than an operational metric like cycle time or days to pay, because a real P&L figure is more identifying than a process timing. Sensitive metrics need a wider outside cohort behind them before the Industry line will show at all. You do not need to know the exact numbers behind either bar to use the page: what matters is that the floor exists, it is higher for financial figures, and it never bends for a network that "really needs to see it this week."

What this looks like on screen. When the floor is cleared, the Industry figure appears next to the office count at the top of the deep-dive, and the Data Sources panel names the peer cohort behind it, worded as a count of operators "contributing under data dividend." When it is not cleared, there is no Industry figure, no partial number, no "based on 1 operator" asterisk. The reference line is simply absent until enough outside operators are behind it.

Real networks never see demo numbers. If your network is a live, real franchise system, its Industry reference is built only from other real operators. Verinode keeps a fully separate demo/showcase universe for onboarding and product tours, and a real network's benchmark will never quietly borrow from that showcase data to fill a gap. Early on, before enough real operators outside a given network have contributed a metric, that means the Industry line reads as simply not there yet, "warming up," rather than showing a number pulled from the wrong world. This is deliberate: a number that is present but wrong is worse than a number that is honestly absent.

Why peers outside your network are never named

This is the second half of the same promise. Even once the Industry line clears the floor and appears, it is always a single number, a median. It is never a named operator, never a company, never a city-level breakdown small enough to guess at. There is no drill-down that turns "Industry $X" into a list of contributing businesses, the way your own office ranking is a list of contributing offices. That asymmetry is intentional and it is the whole point of the privacy boundary: your own network is a relationship Verinode already sits inside of, benchmarking your own offices by name is oversight you already have. Everyone outside it is a stranger's business, and Verinode's entire value as an independent data trust rests on never letting one franchise system reach through the platform to see another operator's real numbers. The floor above is the mechanism; this is the reason it exists.

Building your own slice: the Custom Cohort Builder

If you want to test a specific hypothesis, "how do my Southwest offices compare to the rest of the network," or "how does this subset of my group stack up against everyone else," Verinode has a second tool built on the same math: the Custom Cohort Builder, at hq.verinode.ai/benchmarks/cohort. Pick any operators from your directory on the left, choose a metric and a comparison universe (Rest of network or Rest of this group) on the right, and press Run benchmark. Nothing here is saved, the cohort you build exists only for that query.

The result shows three tiles: your selected cohort's median with its P25/P75 spread and how many operators are behind it (labeled "Cohort · n="), the comparison universe's median the same way (labeled "Rest of network · n=" or "Rest of group · n="), and a percentile tile showing where your cohort's median sits inside the comparison universe, colored green when it is ahead and red when it is behind. That third tile also carries the same coverage label as the rest of HQ, No data, Early signal, Indicative, Observed, or Verified, reflecting how solid a read this particular cohort-vs-comparison slice is. The percentile itself goes blank, reading "Not enough rest-side data," if either side of the comparison is too thin to publish safely, the identical anonymity floor described above applies here too: your cohort is your own data and always shows, but the comparison side is still subject to the same protection against a thin outside group.

Empty state. If your directory has no active members with an operator profile yet, the tool reads: "No active operators yet. Cohort math needs at least one active member with an operator profile."

How to use both together

1. 1Open a metric card on the Benchmarks tab. If it is visible at all, your own network has cleared the bar to show a real distribution, read the office ranking first: it is entirely your own data, always visible, and it tells you exactly which office needs coaching this month.
2. 2Check for an Industry figure next to the office count. If it is there, use it as a single outside gut-check, "how does my whole network compare to the wider market," not as a target for any one office. If it is absent, that simply means not enough outside operators have contributed that metric yet, keep making decisions from your own office ranking in the meantime.
3. 3For a sharper question than the fixed categories answer, build a custom cohort at hq.verinode.ai/benchmarks/cohort and compare exactly the slice you care about against the rest of the network or the rest of your group.
4. 4Come back as your network grows. Coverage only ever moves in one direction, more of your offices reporting, more outside operators contributing, so cards that are hidden or missing an Industry line today will fill in on their own over time.

Related reading

• Benchmarks: how HQ reads the network, the full tour of the tab's categories, the Ratings/Library/Industry News tabs, and the controls this article assumes.
• Network health, for the office-standing leaderboard and regional roll-up, a different lens on the same office data, entirely internal to your network.
• Standards and Programs, for how HQ turns a benchmark gap into a network-wide action.
• Broadcasting to your network, for sharing a benchmark finding with the offices it concerns.