Job types, perils, and service lines

Overview

Every job in Verinode carries a category, the kind of work it is. But behind that one word sit two separate questions Verinode answers about each job: what caused the loss, and what work you actually did. Those two axes, the peril and the service line, are what let Verinode compare your water jobs against other water jobs, your reconstruction margin against other reconstruction margin, and never mix the two into one meaningless average.

This is the quiet machinery under margin by job type, your peer benchmarks, and cohort comparisons. None of it changes what you see on a job. Your own category text is shown exactly as you typed it. The classification happens alongside it, automatically, as your data flows in.

The category you see is your own words

Verinode does not decide what your job categories are. If you call a job "Cat 3 water," "WATER," "Roof tarp install," or "Pack out," that is what shows on the job tile, the card, and the profile, verbatim. There is no master list of approved categories, and there is no relabeling. What you typed is what you see.

That matters because operators run different kinds of work and name it differently. One shop's "water damage" is another's "flood mitigation." Both are valid, and Verinode treats them as the operator's own truth rather than forcing them into a house taxonomy.

So how does Verinode benchmark a "Cat 3 water" job against a "flood mitigation" job when the words are different? By reading each one into a shared, controlled vocabulary used only for grouping, never for display.

Peril: what caused the loss

The peril is the cause of loss. It is a locked, six-token spine, and every job is read into one of these (or none):

• Water Mitigation (water)
• Fire & Smoke (fire)
• Mold Remediation (mold)
• Contents (contents)
• Reconstruction (reconstruction)
• Biohazard & Trauma (biohazard)

When the wording is specific enough, the peril carries a subtype for finer grouping. If a subtype cohort is too small to compare on its own, it rolls up to its parent token:

• Water: Category 1 (Clean), Category 2 (Gray), Category 3 (Black)
• Fire: Structure Fire, Smoke & Odor Only, Wildfire
• Contents: Pack-out, Electronics, Documents, Textiles
• Biohazard: Trauma / Crime Scene, Unattended Death, Hoarding, Sewage / Infectious
• Mold and Reconstruction have no subtypes; they group at the token.

Service line: what work you did

The service line is orthogonal to the peril. The peril is the cause; the service line is the work you performed and billed for. One fire loss can produce mitigation, reconstruction, and contents revenue all at once. A mold job is specialty remediation work. The two axes are read separately on purpose, so both cohorting and revenue-mix stay honest.

There are seven service lines:

• Mitigation (emergency, drying, demo)
• Reconstruction (interior build-back trades: drywall, paint, flooring, framing, cabinets, trim, insulation)
• Exterior & Roofing (roofing, siding, gutters, windows, exterior paint)
• Contents (pack-out, electronics, documents, textiles)
• Specialty Remediation (mold, asbestos, lead, biohazard, sewage, IAQ / air-quality)
• Cleaning (carpet, duct, commercial, disinfection)
• Other (consulting, inspection, equipment rental, non-loss remodel)

Why the classification matters

Two jobs are only comparable when they are the same kind of work. Grouping is the whole point:

• Margin by job type ranks your gross burdened margin per service line so you can see which work pays off and which quietly loses money. See margin by job type.
• Peer benchmarks put your water margin next to other operators' water margin, not against a pool that mixes water, fire, and reconstruction into one flat figure.
• Cohort comparisons group like with like: a commercial fire job is benchmarked against other fire jobs, not against residential water.

Without the peril and service-line axes, every benchmark would be a blended average that hides exactly the differences you need to act on.

How a job gets classified

Classification runs on the data that flows in, no manual tagging required.

1. 1Verinode reads the words on the job. For the peril, it reads your verbatim category or peril text. For the service line, it reads scope labels and line-item descriptions.
2. 2The text is normalized to a comparable form (lowercased, accents stripped, punctuation collapsed) so "Cat 3 Water" and "cat3 water" land the same.
3. 3An ordered rule table decides, first match wins. The order encodes precedence: specific, cert-gated terms are tested before generic ones, so "asbestos abatement" reads as Specialty Remediation rather than a stray "cleaning," and "Cat 3 water" resolves to Water / Category 3 rather than a bare water token.
4. 4A subtype is attached when the wording supports one, otherwise the job groups at the parent token or line.
5. 5If a service line cannot be read from the text, Verinode falls back to the peril. A mold or biohazard job defaults to Specialty Remediation; a water or fire job defaults to Mitigation, the first-response phase almost every such loss begins with.

Sharpening the classification

The more structured your data is, the more precise the grouping. A job that arrives with clear scope text and line items classifies to a subtype; a bare one-word category groups at the token. As invoices and line-item costs flow in, service-line attribution moves from the job's primary work phase toward the true dollar split of work within the job. Nothing here needs your attention, but richer data in means sharper cohorts out.

Data sources

Every job is classified from your own data, and the shared taxonomy exists only to group like work with like for benchmarking. Verinode is an independent data trust; your operator data is never sold to carriers.