The benchmark deep-dive

What the deep-dive shows

A benchmark tile on the Benchmarks page gives you the headline: your number and a sense of where it sits. The deep-dive is what opens when you tap that tile, and it is where the single number turns into the full picture. It answers the questions the headline cannot: where exactly do I sit in the pack, how has this moved over time, who is behind the number, and what is the gap worth to me.

Everything in the deep-dive is read from data you already have, joined against an anonymized peer cohort. Verinode does not decide anything for you here. It lays out the distribution, the history, and the sources as plainly as a fractional COO would, and leaves the read to you.

If you have not yet read how to read a single benchmark, start there. This article assumes you know what a median, a percentile, and a cohort are, and walks through every panel the deep-dive stacks on top of them.

Where to find it

Open Benchmarks from the sidebar. Every metric on the page is a tile. Tap a tile and the deep-dive opens as a full-screen glass overlay. It slides in over the page, dims the background, and closes on the close button, the escape key, or a click outside. You can swipe between sibling metrics without leaving the overlay, so a metric and the ones next to it read as one set.

The panel stacks its sections top to bottom. Each one appears only when you have the data to feed it, so a sparse metric shows fewer panels than a mature one. An empty panel is not a broken screen, it means the readings, peers, or history behind it have not landed yet. The rest of this article walks through every section in order.

The hero recap

The top of the panel is the recap: your most recent value, big, in the metric's own unit (a percentage, a dollar figure, a day count, or a plain number). Beside it, up to three chips summarize where that number stands.

• Percentile pill. A label like "72nd Percentile" telling you where you rank in the cohort. It reads green when you are at or above the median (the top half) and red when you are below it. Percentile is direction-aware: for a metric where lower is better, sitting low still reads as a strong percentile.
• Trajectory chip. Improving or Declining, with the size of the recent move, when your value has moved meaningfully over the last window. Green when the move is in the good direction for that metric, red when it is against you. A flat metric shows no chip.
• Dollar gap line. Under the number, a plain-language line puts the gap in money: "Closing the gap to median is worth $X/yr" when you are behind, or "You're beating the median by $X/yr" when you are ahead.

Cohort Distribution

This is the core panel and the reason most operators open the deep-dive. It shows the full shape of the cohort, not just the median, so you can see whether you sit in a tight pack or a wide spread, and exactly where your dot falls.

What you see. A header reads the scope and the cohort size, for example "National cohort · 34 peers". Below it, a horizontal distribution track:

• A shaded band spanning P25 to P75, the middle half of the cohort (the interquartile range). Most operators live inside this band.
• Thin markers at P25, Median, and P75, labeled with their values beneath the track: P25 on the left, Median centered and bold, P75 on the right.
• Your dot, placed at your value. It reads green when you are on the better side of the median and red when you are on the worse side, so a glance tells you which half you are in.

Choosing the scope. A picker in the panel header lets you switch the cohort the distribution is drawn from:

• Group, your franchise or association cohort.
• Region, operators in your state.
• National, all consenting operators.

Widening the scope grows the cohort but loosens the comparison, a national median is broad; a group median is people who run the business the way you do. All three scopes always show, even when one is too thin to draw, so you can see which cohorts have enough peers and decide how tight a lens to use.

What to do. Read your dot against the band first. Sitting inside the P25 to P75 band is normal; the operators worth learning from sit past P75 on the good side, and the gap worth closing is the distance from your dot to the median. If your group cohort is too thin, widen to Region or National for a read, then narrow back as your cohort fills in.

Empty state. When a scope does not have enough peers yet, the panel reads "Not enough peers in the [scope] cohort yet. Try a wider scope above." That is the anonymity floor doing its job, not an error.

Development Over Time

The distribution is a snapshot of today. This panel is the movie: how the metric has moved week over week, both for the cohort and for you.

What you see. A weekly chart covering the trailing window (about six months). The cohort is drawn as a shaded band from P25 to P75 with a median line through it, so you can watch the whole pack drift. Your own value rides on top as dots, one per week, each colored green or red depending on whether you beat that week's median. A caption names the date range and reminds you what the band and dots mean.

What to do. Watch the direction, not just the level. A dot climbing week over week while the median holds flat is real, durable improvement, not a one-month blip. A dot drifting toward and then past the median is an early warning worth acting on before it shows up in your margin.

Empty states. History accrues one weekly reading at a time, so this chart is quiet at first:

• Before enough readings have landed, it reads "Your trend history is building. This chart fills in as new weekly readings land."
• When your own readings exist but the peer band is not yet unlocked for you, it reads "Peer band unlocks with benchmark contribution. Your readings shown." and plots your dots alone. Contributing your data to the benchmark family unlocks the peer band, see peer benchmarks and contribution for how that exchange works.

Compare by Cohort

On job-grain metrics, a second axis lets you slice the same metric by the kind of work, so you can compare like with like. Pick a cohort (by peril or property type, for example) and the panel adds a comparison row showing that slice's average and how many jobs stand behind it. This is a lighter cut than the full distribution above, meant for a quick "how does this look on water jobs versus the whole book" read. It appears only on metrics that carry a job-level cohort.

Materials: your suppliers and carrier reconciliation

On material unit-cost metrics, the deep-dive adds two panels built from your own invoices and supplement outcomes:

• Your Suppliers. Your own suppliers for that material, each with the unit price you actually pay and, where a cheaper supplier of yours exists, roughly what switching to it would save per year. Your cheapest is flagged "Your best." These are your real vendor names; the peer side stays anonymized. If a row is not really a supplier for that material, remove it with the small control on the row and Verinode stops attributing it.
• Carrier Reconciliation. How your own carriers reconcile the billed unit price of that material line, worst cutter first: what you bill, what they approve, and how often they cut. A carrier that pays as billed reads green; one that consistently approves less than you bill reads red.

How it's calculated, and the data sources

The foot of the panel is where Verinode shows its work, in two columns.

How It's Calculated. A plain-language note on how the metric is built from your data and compared against the cohort matching your scope. When a metric has no bespoke note, it falls back to the standard explanation, including the reminder that every comparison clears a minimum cohort size before it shows, so individual operators can never be re-identified from a median.

Data Sources. The lineage behind the number, as a short list:

• The peer cohort, with the count of operators contributing to it under the data dividend.
• Research, when a research-layer benchmark exists for the same metric. It is shown as "Verinode Research" or the generic "Industry research" with its value, never as an external company's name.

Empty state. When a metric has neither enough peers nor research coverage yet, the sources list reads "Sources will appear once cohort sample size and research coverage land for this metric."

The HQ view: across your offices

When you open a benchmark inside Verinode HQ, the distribution panel is replaced by an Across Your Offices view. Instead of an anonymized peer cohort, it ranks the network's own offices on the metric, best to worst, against a single broader-industry reference line. Offices on the strong end of the distribution read green (learn from them), offices on the weak end read red (coach them), and the rest sit in between. An HQ leader benchmarks offices against each other, so the scope picker and the peer cohort do not appear in this frame. When no office has reported the metric yet, it reads "No offices have reported this metric yet."

Best-practice example

Say you open your collection-rate tile. The hero reads 68%, with a red 31st Percentile pill and a line: "Closing the gap to median is worth $54k/yr." The Cohort Distribution shows the median at 79% and your dot sitting left of P25, well behind the pack. You widen from Group to National to confirm it is not just a thin cohort, and the story holds. Development Over Time shows your dots drifting down over the last two months while the median line stays flat, so this is a real slide, not a seasonal dip. The dollar line tells you what it is worth, the distribution tells you how far you have to move, and the history tells you it is getting worse, not better. That is enough to make collections your priority this month. Verinode laid out the read; the decision to act is yours.

Getting the most from it

1. 1Open the tile and read the hero: your number, your percentile, and the dollar gap if the metric carries one.
2. 2In Cohort Distribution, find your dot against the P25 to P75 band, then widen or narrow the scope to get the tightest comparison that still has enough peers.
3. 3In Development Over Time, check the direction of your dots against the median line, a slide matters more than a single week.
4. 4Read How It's Calculated and Data Sources so you know exactly what stands behind the number.
5. 5If the metric matters to you, watch it so Verinode tells you when it moves.

For where the deep-dive sits in the wider Benchmarks surface, see the Benchmarks overview. For how the composite Verinode Score rolls several of these metrics into one read, see its article.