Warming up, the n≥5 floor, and the privacy boundary

The three questions this article answers

Standards Conformance grades every office in your network against the stage-time targets HQ sets (see hq-standards and hq-standards-conformance for the full editor and grid walkthrough). Once leadership starts reading that grid, three questions come up every time:

1. Why does my newest office show "Warming Up" instead of a grade, even though I know it's been running jobs?
2. When HQ reads that grid, is anyone at HQ looking at my actual job files to produce it?
3. If my network is a franchise of independently owned locations, does a slow office get named and shown to every other franchisee?

This article answers all three in one place: the warming-up state and the five-case floor behind it, the aggregate-only doctrine that keeps job-level data out of the grading pipeline entirely, and how office names are anonymized before they ever reach the grid. Read it alongside hq-conformance-grading for the Met / Near / Missed math itself, which this article assumes.

Why "Warming Up" exists

A stage median is only trustworthy once there's enough history behind it. A single case, or even three or four, can swing wildly, one unusually fast job or one unusually slow one and the "median" is really just an outlier wearing a number's clothes. Grading an office against a target on that thin a sample would produce a verdict that looks precise but isn't.

Verinode's answer is to refuse the grade rather than fake it. Any office with fewer than five in-order cases on a given milestone transition (say, Assigned → Started in the Job lifecycle, or Submitted → Responded in Supplement turnaround) carries no median at all for that transition, and therefore no verdict. It reads as Warming Up instead. "In-order" means both milestone dates are on file and the second one doesn't precede the first, same-day completions count fine, but a case with a completion date logged before its start date doesn't count toward the five.

This is the same design principle you'll see stated elsewhere across Network Health: an absence of qualifying data produces an honest empty state, never a number stretched to look complete. A Warming Up cell is not a zero, not a Missed, and not averaged into anything. It's simply left out of the count until there's enough behind it to grade honestly.

The n≥5 floor in practice

Five in-order cases is the floor for every milestone transition in every registered process, jobs, supplements, recruiting, safety, and any other process your network has milestones for. It is not configurable per standard or per office; it is the same bar everywhere the nightly aggregate refresh computes a stage median.

Here's where you'll see it stated directly, in the exact product copy:

• On the Conformance tab's office list (Network page, sidebar Network, the Standards Conformance tile, then the Conformance tab): a warming-up office's row reads "Warming Up · N of 5 [cases]", for example "Warming Up · 3 of 5 jobs" or "Warming Up · 1 of 5 supplements." The noun changes to match the process, a job, a supplement, an applicant, an incident, a purchase, a lead, a review, so the count always reads naturally.
• When no office on a transition has cleared the floor yet, the whole transition card reads: "Office grades appear once an office clears five in-order [cases] on this stage."
• On the Process Standards editor (hq.verinode.ai/standards/process, reached from Compliance in the sidebar, then the Process Standards tile), a transition with no network-wide median yet shows "No Network Median Yet" in place of the median line, since the floor applies to the network's own pooled median too, not just individual offices.
• On the Process Standards hero panel, if standards exist but nothing anywhere has scored yet, the status pill reads Warming Up (neutral tone) rather than a percentage.
• On the Standards Conformance tile on the Network page, before anything has scored, the subtitle reads "N Standard(s) · Warming Up" instead of a percentage.

A worked example: your network sets a 3-day target for Assigned → Started. Office A has run 12 jobs through that transition with clean dates and a 2.5-day median, it grades Met. Office B just opened three months ago and has 2 jobs through that transition so far, its row reads "Warming Up · 2 of 5 jobs", no grade, no color, nothing implied about whether it's fast or slow. Once Office B logs its 5th in-order job on that transition, the next nightly rollup computes its median and it grades like everyone else, Met, Near, or Missed, on the same bar as Office A.

Warming-up cells never count toward Conformance % in either direction. The percentage is met / (met + near + missed), warming-up cells are excluded from that denominator entirely. A network where every office is still warming up on every transition reads as a dash (null), not zero, because zero would wrongly say the network is failing a bar it hasn't been measured against yet.

Grading rides on stage medians only, never job-level records

The five-case floor is a statistical safeguard. The privacy boundary underneath it is structural, and it's worth being precise about the difference: even with a perfectly sized sample, HQ's read side of Standards Conformance never has access to any individual job, invoice, adjuster note, or case file to produce a grade.

Here's the actual pipeline. Each night, a cron reads job and milestone dates from every operator's own PII database (the same database that holds their invoices, adjuster correspondence, and everything else in their own IQ account), mines the in-order case durations per stage transition, and writes exactly one row per office per transition to the network data: a sample count (n), a median (median_days), and the p25/p75 spread. That write happens through the sanctioned pii-to-core bridge, the one place in the platform allowed to touch both databases in the same process. Standards Conformance's own read side (the page you actually look at) only ever queries the network data and the network data, both the network layer schema tables that HQ's database role can reach. There is no code path from the Conformance grid back to a location's job list, no drill-in, no "see the underlying cases" link, because the grid was never built from individual cases in the first place, it was built from a number the cron already reduced to a median before HQ's side of the system ever ran.

Practically, this means:

• A Missed verdict tells you an office's median for that stage is running slow. It does not, and cannot, tell you which specific job caused it, what an adjuster said, or what a customer complained about. That context comes from the location itself, in whatever conversation follows a Missed reading.
• The Network Median row on each transition card, the network-wide figure shown alongside every office's own row, is computed the same way: a pooled median across every office's raw cases, weighted by case count, not an average of office averages. It is graded against the same standard, using the same Met / Near / Missed math, so you can read whether the network as a whole is keeping its own promise.
• Nothing about conformance grading changes what a location owns. Franchisees keep their operational data in their own IQ account; what crosses the boundary into the network data is a median and a count, nothing else.

Office names and the group's anonymization model

The office name sitting next to each row on the Conformance grid follows the exact same anonymization rule as every other named surface on Network Health, Location Directory, Leaderboard, Best Practices, all of it. Standards Conformance doesn't have its own separate privacy setting; it inherits your group's entity model.

• Independent operators (a franchise network of separately owned locations, or an association of independent member businesses): this is the default posture for every new network. Each office's real name is replaced with a stable label, "Franchisee #A1B2" style, derived from the last four characters of that office's internal ID. The same office always gets the same label, so leadership can track a specific office's conformance over time or notice the same label showing up as Missed on several transitions, without that label ever revealing which business it is to anyone reading the grid.
• Same entity (a commonly owned multi-location operation, or a PE-backed roll-up where every location is the same legal business): real location names show through everywhere, including on this grid, because there's no second party to anonymize from. It's one business looking at its own branches.

This is a network-level setting HQ configures once, not something that can be toggled per page or per user. If your network runs as independent operators, you will never see a real name on the Conformance grid no matter which transition or which office you're looking at; if it's same-entity, you'll never see the anonymized placeholder. Either way, the underlying grading math is identical, only the label changes.

A useful way to hold both halves of the privacy story together: the anonymization rule protects who an office is; the aggregate-only pipeline described above protects what's in an office's business. A Missed verdict under an anonymized label still tells you exactly which stage is slow and by how much, it just doesn't hand that information to anyone outside HQ leadership as "here's a name to go with a weak spot." Franchisee-to-franchisee visibility of each other's conformance standing was never part of the design; this grid is an HQ leadership view, not a peer leaderboard.

Where you'll see all of this together

• Network page (hq.verinode.ai/network, sidebar Network): the Standards Conformance tile leads the Network Flow row once at least one standard is defined. Clicking it opens the Conformance tab on the card slider, where every standardized transition is broken out office by office, each office row showing either a scored verdict or the "Warming Up · N of 5" line, and the office name resolved per your entity model.
• Process Standards (hq.verinode.ai/standards/process, reached from Compliance in the sidebar, then the Process Standards tile under Standards & Audits): the editor where targets are set, and where the hero panel's status pill reads Warming Up until anything scores.

Empty states, verbatim

• No standards defined anywhere. The Conformance tab reads: "No stage-time standards set yet. Define target days per stage under Standards → Process Standards; every office is then measured against the same bar, and the grades land here." The Standards Conformance tile doesn't render on the Network page at all until this changes.
• Standards are set, but no snapshot has landed yet. The Conformance tab reads: "Standards are set. Office grades appear after the next nightly rollup of stage medians."
• A transition has a standard but no office has cleared the floor. That transition's card reads: "Office grades appear once an office clears five in-order [cases] on this stage," with the noun matched to the process.
• No standards set yet, on the Process Standards hero. The status pill reads Not Set Yet (neutral), distinct from Warming Up, which only appears once standards exist but haven't scored.

Related articles

• hq-conformance-grading: the Met / Near / Missed math itself, including the 20% tolerance band and the Conformance % formula.
• hq-standards-conformance: the full Conformance tab and tile walkthrough, column by column.
• hq-standards and hq-process-standards-editor: setting, adjusting, and clearing a target-days standard, and adopting network medians in bulk.
• hq-network-privacy-boundary: the full HQ privacy model, including the separate k-anon floor that protects the cross-network industry reference line.
• network-health: the Network page these tiles live on, and how Missed conformance feeds the cross-network signal detector.
• hq-compliance: the Compliance hub that Process Standards lives inside.