Every operator knows their own numbers. Almost none can tell you whether those numbers are good.
Across 15 dimensions, read from the systems you already run rather than from a survey you fill in once a year. Filtered to your size, service mix, and region, so the comparison is to businesses actually like yours.
Read, Not Surveyed
A survey asks what you think your margin is. Verinode reads the P&L, the job costing, and the invoices underneath it, which is why the figures reconcile instead of approximating.
Restoration Only
Every contributor is a working restoration operator sharing their own real data. Never scraped, never bought, never a survey panel, and never an adjacent trade averaged in to make the sample look bigger.
Withheld When Thin
No figure publishes until enough distinct operators sit behind it. You will see fewer benchmarks early. That is the trade that makes the ones you do see worth acting on.
Day One
Every benchmark network starts empty, and most fill the gap with invented demo numbers. We would rather start from something public and say what it is. Franchise systems file a disclosure document with state regulators each year, and those filings report what locations bill and how many opened and closed in every state. Verinode reads them as a starting prior, so day one is not a set of blank cells.
What that data is not: it is topline revenue with no margin in it, self-reported and unaudited, and reported by franchise locations that pay royalties, so it does not stand in for an independent operation. Publishing it is optional, so the systems that do are a self-selected group, and the largest one publishes nothing at all. Coverage is a handful of brands rather than a market view, because filings whose tables do not reconcile against their own printed totals are refused rather than guessed at. It is a floor to stand on while your own data arrives, and it is superseded the moment operator contributions are better. The full methodology names every source.
It is never mixed into the peer benchmarks built from operator contributions. Those are a different population under a different legal regime, and the separation is structural rather than a promise.
Common Questions
113 metrics across 15 dimensions: financial health, operations and cash cycle, fleet and facilities, carrier performance, team performance, safety and compliance, vendor cost and concentration, growth and marketing return, subcontractor performance, recruiting and retention, equipment and utilization, exterior and roofing capture, reputation, certifications, labor cost. Each is read from the systems an operator already runs rather than from a survey, and compared against operators filtered to similar size, service mix, and region.
A survey asks operators what they think their numbers are, once a year, and most of them are guessing because the figure is not sitting anywhere they can read it. Verinode reads the actual records: the P&L, the job costing, the invoices, the timestamps. It also updates as the data arrives instead of aging for twelve months, and every contributor is a working restoration operator rather than a mixed-trade panel.
You contribute your own operating data and get the peer view back, which is the trade that makes an independent benchmark possible at all. The protections are structural: your raw data stays in your own encrypted environment, only anonymized aggregates leave it, no operator ever sees another operator's raw data, and nothing is ever sold or given to carriers. The commitments are published in full in the Data Use Policy and reviewed by an Operator Advisory Council.
Because the cohort behind them has not cleared its minimum yet. A peer figure built on a handful of businesses is not a benchmark, and a thin cohort cannot protect anyone's anonymity, so the number is withheld rather than shown thin. Verinode is early and says so: you will see fewer benchmarks at the start, and the ones you see are ones the sample can actually support.
Yes, and the safeguards that keep it lawful are the ones Verinode is built on. A neutral third party collects the data instead of competitors trading figures directly, the output is aggregated and anonymized above a participation floor, competitively sensitive figures carry a 90-day historical lag rather than publishing live, and Verinode never recommends, defaults to, or scripts a price or a wage. The full reasoning and the seven design principles are published at verinode.ai/fair-benchmarking.
Why Operators Join
The other half is where they sit against operators your size running the same work. That is the part you cannot pull out of your own accounting, and the part that tells you what to do next.