Cohorts and scope, who you are compared against

What a cohort is

Every peer number on the Benchmarks page answers the same question in a different way: how does this compare to other operators like me? A cohort is that set of "operators like me." It is the group of anonymized peers whose data goes into the median or rate you are reading. Change the cohort and you change who you are standing next to, so the number moves.

Two independent controls decide the cohort. Scope sets the geographic or membership circle, national, regional, or one of the networks you belong to. Cohort filters narrow that circle to a kind of work or a kind of operator, the same service line, the same job type, the same property type, or operators in similar shape to you. Scope and filters stack: you are always being compared against the peers who match every control you have set.

Verinode is an independent data trust, not a data broker. The peers behind a cohort are always anonymized, aggregated to a minimum operator count before anything is shown, and never sold to carriers. That is what makes it safe to be compared this precisely. For the mechanics of how those aggregates are built, see how benchmarks work; for how to read one once it is on screen, see reading a benchmark.

The scope picker in the header

At the top of the Benchmarks page, next to the metric headline, sits the scope control labeled Compare to. It sets the outer circle of peers before any filter is applied.

The scopes it can offer:

• National, every eligible operator in the network. Always available, and the default.
• Regional, operators in your own state. Available only once your profile carries a state.
• One entry per network you belong to, labeled by name (for example your franchise network, a buying group, or "RIA Network"). One button, or one dropdown row, for each network on your membership. These appear only for the networks you are actually a member of.

There is no generic "Group" option. If you belong to several networks, the picker lists each one by name so you can switch between them; if you belong to none, you simply do not see a network entry. These are the only scopes. Verinode does not invent a circle you are not part of.

How the control renders. With no network memberships, the picker is a small row of pills, Regional and National, with the active one highlighted in copper. Once you belong to at least one network the labels get long, so the control becomes a Compare to dropdown instead, listing your networks, then Regional, then National.

Scopes that are disabled

A scope you do not qualify for is shown but not selectable, rather than hidden, so you can see it exists and what would unlock it.

• Regional is disabled until your profile has a state on file. Hovering it reads: "Add a state on your profile to unlock regional benchmarks." Add your state in your profile and the regional scope lights up.
• A network entry only appears when you belong to that network. There is no disabled placeholder for a network you have not joined.

Under the hood the page computes this availability directly from your account: regional turns on when your profile has a state, and network scopes come from your actual group memberships. Pick a scope you do not qualify for by editing the URL and the page quietly falls back to the default cascade rather than rendering an empty screen.

How the cohort is formed, and the anonymity floor

Whatever scope and filters you choose, a cohort is only ever published once it holds enough distinct operators. The privacy unit is the operator, never the row. One operator with fifty jobs is still a cohort of one, and publishing its "benchmark" would hand a competitor that operator's exact private numbers. So the floor counts distinct operators behind the number, not observations.

The floors match Verinode's signed data-use commitment:

• A minimum number of operators must stand behind a standard benchmark, at any scope, national or regional or network.
• A higher minimum applies to sensitive categories, the financial, labor-cost, and profit-margin metrics, because a P&L figure is more re-identifying and a thin cohort is easier for two colluding contributors to unpick.

When a cohort is below its floor, the peer value comes back empty rather than being shown against a thin sample. On a deep-dive comparison the peer side reads a dash and the line tells you why. This is honest by construction: benchmark access and gating covers the difference between "not enough peers yet" and "peers exist but your access is gated."

Narrowing loosens itself to stay publishable

When you stack filters, tightest cohort first, Verinode tries the exact match, and if that falls below the floor it drops the least important dimension and tries again, widening one step at a time until the cohort clears the floor. The health band gives way first, then geography, with the kind of work, the service line and job type, held onto longest because it matters most to a like-for-like read. The comparison line on a deep-dive says exactly what happened, for example: "Compared against 14 operators matching commercial, fire. State and CAT vs daily were loosened to reach a usable sample." You are never silently compared against a looser cohort than the line admits.

The cohort filters

Below the headline, next to Compare to, sit the filters that narrow the scope's circle to a kind of work or a kind of operator. They are independent and combinable, and each one is anonymized against the same floor as everything else.

• Service line, the book of business, for example Exterior & Roofing or Reconstruction. This is how a specialized operator compares like with like: a roofer picks Exterior & Roofing to benchmark their exterior work against other operators' exterior work, not against the whole restoration field. Defaults to All.
• Job type, the peril, for example Water, Fire, or Mold. Defaults to All.
• Property, Commercial or Residential. Defaults to All.
• Operators like you, the "similar shape" health-band toggle, described next.

Setting a service line, job type, or property narrows the metrics to the same kind of work and re-reads them against peers doing that same work, showing your own value for that slice, the peer cohort mean, and the operator count behind it. Whole-business metrics like margin and supplements are not sliced this way; when you set a job-grain filter, the page tells you to clear the filter to see them.

"Operators like you", the similar-shape cohort

The Operators like you toggle narrows the cohort by the shape of the business, not the shape of the work. It compares you against operators whose overall health is close to yours.

It works off your Operator Health Score. That score, once it is real, lands in a band:

• At Risk
• Needs Work
• Fair
• Strong
• Excellent

Turning the toggle on narrows every peer number to operators in your own band plus one band on either side, so a Strong operator is compared against Fair, Strong, and Excellent peers rather than the entire field. Operators whose score is still forming are always excluded from this cohort, on purpose: comparing against a not-yet-settled sample would mislead. When it is on, the tooltip confirms who you are standing next to, for example "Comparing against operators in similar shape to yours (Strong)."

When it is disabled. The toggle greys out until your own score is set. Hovering it then reads: "Available once your score is set." Your score has to be real and out of its cold-start window, and you have to have peer access, before the similar-shape cohort can resolve. On a deep-dive, the same idea reappears as chips grouped by Job type and Property; tapping the active chip again toggles the cohort back off.

Changing scope or cohort re-reads the whole page

Scope and the cohort filters do not just re-label what is on screen. Each one resolves a fresh set of peer numbers, because every metric carries its own scope-specific and cohort-specific peer data. There is no single cached "peer column" that gets re-sorted.

The two controls do this differently, on purpose:

1. 1Scope (the Compare to header control) writes a scope value into the page URL and re-renders the whole page on the server. Every metric is re-resolved against the pinned circle, national, regional, or the chosen network, in one pass. This is a full, server-side re-read, so it is the heavier of the two and is why scope lives in the URL: the view is shareable and reloadable at that scope.
2. 2Cohort filters (service line, job type, property, and the health toggle) update in place without a page reload. Picking a job type fetches just the job-grain comparison for the new cohort and swaps the results panel; the similar-shape toggle alone re-scopes the already-loaded list to your health band with no round-trip. Only the comparison updates, so it feels instant.

Either way, the number you end up looking at is always the peer read for the exact scope and cohort shown in the controls, resolved against a cohort that cleared the anonymity floor, or shown empty if it could not.

Getting the most from it

1. 1Add your state to your profile so the Regional scope unlocks, then compare national against regional to see where you sit locally versus nationally.
2. 2If you belong to a franchise, buying group, or association, use its Compare to entry to benchmark inside your own network.
3. 3Set the Service line filter to your book of business so a specialized operation is read against peers doing the same work, not the whole field.
4. 4Once your Operator Health Score is set, turn on Operators like you to compare against businesses in similar shape, then read the comparison line to confirm the cohort.
5. 5When a peer number reads empty, widen the circle: clear a filter or move from regional to national. The cohort was below the anonymity floor, and a wider one will usually clear it.

Verinode chooses your cohort the way a fractional COO would, the closest honest comparison it can publish without ever exposing a single operator. The peers stay anonymous, the floor holds, and the data is never sold to carriers. That is the whole reason the comparison is worth trusting.