Understanding your Operator Health score

What the Operator Health score is

The Operator Health score is the single "how am I doing" number for your business. It rolls the signals Verinode already reads from your data into one figure on a familiar scale, 580 to 820, the same range a credit score uses, so a higher number is plainly better and small moves are easy to feel.

It sits at the top of the Margin page as a gauge, on the left, paired with your Cash Runway gauge on the right. Open it from the sidebar at iq.verinode.ai/margin.

The score is a weighted blend of nine dimensions, Margin, Cashflow, Operations, Compliance, Team, Reputation, Sales, Process, and Vendor stack. Each is scored on its own 0–100 read from your data, then weighted and mapped into the 580–820 display range. The score is not a judgment Verinode hands down, it is a mirror of the data flowing in, and it enriches as more of it arrives. Verinode acts here as your Co-COO: it reads the whole business at once so you can see where to spend the week.

The gauge at the top of Margin

The gauge shows four things at a glance:

• The big number, your score, followed by / 820.
• A band label, a plain-language read of the number (see the band table below).
• A "% of dimensions active" line, how much of the score is currently backed by real data, with a breakdown → hint.
• A 12-month trajectory sparkline and a peer cohort rail beneath it.

Clicking anywhere on the gauge opens the full Score breakdown.

What the band means

The color and the band label are driven purely by the number:

| Score | Band | Color | |---|---|---| | 780–820 | Excellent | Green (Expand) | | 730–779 | Strong | Copper | | 680–729 | Fair | Yellow (Maintain) | | 630–679 | Needs work | Red (Analyse) | | 580–629 | At risk | Red (Analyse) |

When your score is still partial (see below), the gauge greys out, shows a , instead of a number, and the band reads "Partial, keep contributing."

The nine dimensions

Each dimension is scored 0–100 from your own data, then carries a base weight into the composite. The weights below are the starting weights; they renormalize as dimensions come online (see "Cold start" below).

1. 1Margin, weight 25%. Your collection rate: total collected divided by total billed, across every job with a billed amount. Needs at least 5 invoices. 95%+ collection scores 90; 90–95% scores 75; 80–90% scores 55; below 80% scores 30. The row reads like 92% collection rate across 41 jobs.
2. 2Cashflow, weight 15%. Average days-to-pay across your paid jobs. Needs at least 5 paid jobs. 30 days or fewer scores 95; up to 45 days scores 80; up to 60 scores 60; up to 90 scores 40; over 90 scores 20. Reads like 48 day average across 22 paid jobs.
3. 3Operations, weight 15%. Average cycle time, days from assigned to completed, across finished jobs. Needs at least 5 jobs with both dates. 8 days or fewer scores 90; up to 12 scores 75; up to 18 scores 55; up to 30 scores 35; over 30 scores 20. Reads like 10.4-day average cycle across 33 jobs.
4. 4Compliance, weight 15%. How current your team's certifications are. Needs at least 1 certification. It scores the share of credentials that are current, then subtracts a penalty for anything expired (expired credentials are weighted harder than ones expiring soon). Reads like 6 current · 2 expiring · 1 expired.
5. 5Team, weight 10%. Roster size plus review coverage. Needs at least 1 team member. A size component (up to 40 points, growing with headcount) plus a coverage component (up to 60 points, based on the share of the team with a current performance review). A small shop still scores well if its few people have current reviews. Reads like 8 on roster · 5 reviews this year.
6. 6Reputation, weight 10%. Average client survey rating over the last 90 days, on a 1–5 scale mapped to 0–100 (an average of 4.5 lands near 90; 3.0 near 50). Needs at least 1 survey response. Reads like Avg 4.4 / 5 across 12 survey responses (last 90d).
7. 7Sales, weight 10%. This one is peer-relative: there is no universal "good" average job value, so it scores where your trailing-year average sold-job value sits against operators like you. Needs at least 8 billed jobs in the last year and a formed peer cohort (at least 3 peers). Top quartile scores 90; above median 75; above the 25th percentile 55; below that 35. Reads like $5,200 average sold job, P68 vs peers (14 peers).
8. 8Process, weight 5%. SOP coverage and quality. Needs at least 1 SOP. Combines the share of your SOPs that are active (weighted 60%) with their average lean score (weighted 40%). Reads like 4/6 active SOPs · avg lean 72.
9. 9Vendor stack, weight 5%. The average Verinode score across your active vendor relationships. Needs at least 1 relationship. If vendors are on file but none are scored yet, the dimension sits at a neutral mid-band and reads 3 vendors on file · awaiting Verinode scores.

Inside each dimension, its own 0–100 read is labeled Healthy (75+), Watching (50–74), or Exposed (below 50). A dimension with too little data reads No data.

"Score loading" and no-data states

You will see one of three states before the score is fully live:

• Score loading…, the gauge shows a pulsing placeholder with the kicker "Operator Health" while the number is still being fetched.
• Partial (cold start), if fewer than half of the dimensions have enough data to score, the gauge greys out, shows , , and the band reads "Partial, keep contributing." The subtitle tells you how much is covered, for example 40% of dimensions active. This is deliberate: a score built on one or two dimensions would mislead more than it informs, so Verinode holds the number back until it is credible.
• Each dimension has its own unlock nudge. Until a dimension crosses its input threshold, the breakdown shows a plain prompt for what to add, for example "Forward 5+ invoices to unlock the margin dimension," "Add a team certification to unlock compliance," or "Promote your first decision plan into an SOP to unlock process."

As you add data, forward invoices, connect your books, log certifications, run surveys, dimensions come online one by one and the score stabilizes. Nothing you add ever penalizes the number for having been missing; a dimension with no data is simply left out until it is ready.

The peer percentile

Beneath the number, the peer cohort rail shows where your score ranks against operators like you. It reads as a percentile, for example the 68th, with the cohort size (n=) beside it, and a rail marking the 25th percentile, the median, and the 75th percentile so you can see your position at a glance. A line of plain copy states it in words: "You're in the 68th percentile of mid-size operators in TX."

How the cohort is chosen, in order, falling back only when a tier is too thin:

1. Your size band and state, the closest match.
2. Your size band nationally, if the state cohort is too small.
3. National, the last resort.

Two floors protect this number. A privacy floor (k-anonymity) means a cohort only resolves once it has enough distinct peers that no single operator can be reverse-identified, below that, Verinode widens to the next tier or shows nothing. A meaningfulness floor means a to-the-percent rank only appears once the cohort has at least 8 peers; below that you see an honest qualitative read ("You're ahead of the typical peer among 4 mid-size operators. A firm ranking will surface as more peers contribute.") instead of false precision.

If no honest cohort exists yet, the rail reads: "Cohort row not seeded yet for your size + region. Ranking will surface as more peers contribute." For a partial score, it reads: "Add more data to unlock peer ranking. Cohort lookup needs at least half your dimensions seeded." The percentile is suppressed entirely for partial scores, the number isn't credible enough to rank. See how benchmarks work for how the network builds these cohorts.

The 12-month trajectory sparkline

To the right of the number, the 12-month trajectory plots your score over the last year at five anchor points: 1 year, 6 months, 3 months, 1 month, and today. A weekly snapshot cron records your score over time; each anchor picks the recorded snapshot closest to that point (within about five weeks), and today's live read is spliced in so the freshest number always anchors the right edge. Beneath the line it summarizes the move, for example +18 pts over 1y.

Until at least two snapshots exist, the sparkline shows a dashed midline with the note "Trajectory fills in as your weekly snapshot accumulates." History builds automatically from here, there is nothing to turn on.

What clicking the gauge opens

Clicking the gauge opens the Score breakdown overlay. It keeps the menu bar and the IQ agent panel visible, so you can ask IQ about anything you see mid-flow. Inside:

• A header reading "Score breakdown" with the count of active dimensions (or "Partial, X% of dimensions active"), and the "as of" date.
• If your score is partial, a copper banner: "Score is partial. Fewer than half of the dimensions have enough data to score confidently. Adding the missing inputs (P&L, jobs, certs, payroll) unlocks the full picture and stabilizes the YoY trend line."
• A Peer position panel with the percentile and its plain-language copy.
• An "Operators like you" section comparing your margins to operators in similar shape (health band) to yours: Gross margin, Net margin, Labor % of revenue, and Days to pay. This section is hidden when there isn't genuine peer data to show.
• A table of all nine dimensions with columns: Dimension, Raw 0–100, Weight (the effective weight, with the base weight in parentheses when they differ), Contribution to the composite, Status (Healthy / Watching / Exposed / No data), and Detail (the read plus any recommended action in copper).
• A link to "How this is calculated" for the full methodology and weights.

Every dimension row is a link. Clicking a row closes the modal and drops you on the section that owns that signal, Margin and Cashflow go to /margin (Cashflow to its Cash Flow tab), Operations to /jobs, Compliance to /compliance, Team to /team, Reputation to /reputation, Sales to the growth benchmarks, Process to /processes, and Vendor stack to /vendors. Hovering a row shows an "Open →" cue.

Related

• Understanding your margin, Margin is the heaviest-weighted dimension in your score.
• How benchmarks work, how the peer cohort and percentile are built.
• Cash flow and runway, the Cashflow dimension and the runway gauge beside your score.

Data sources