Peer scope and cohort confidence in the reputation drill-in

What this is

Every metric tile in the Reputation drill-in shows a franchisee's own number next to a peer number: a median and a percentile drawn from the rest of the network. That comparison only means something if you know two things: who the peer group is, and how much weight the sample size can carry. The peer scope switcher answers the first question. The confidence tier answers the second.

This article covers both, plus why two of the three scope options currently read as pending rather than showing a number.

Verinode HQ never shows you a single franchisee's underlying business records to make this comparison work. The peer median and percentile are aggregate math computed across the network's active members. You see the result of that math, not the inputs.

Where to find it

Open Reputation from the HQ sidebar at hq.verinode.ai/franchise/reputation. Reputation is one of the sections built on the standard HQ franchisee roster: hero metrics up top, then rows of franchisee tiles (below threshold, declining, top performers, most reviewed).

Click any franchisee tile to open its detail slider. The slider header names the franchisee, its location or "Location anonymized," and its membership status (Seeded, Invited, or Active). Directly below that header sits the peer comparison scope row, a small pill group reading Group, Regional, National. This is the peer scope switcher, and every metric tile in the slider (Composite score, Google rating, Google reviews, Yelp rating, Yelp reviews) reads its median and percentile from whichever scope is currently selected.

The three scopes

• Group compares the franchisee against other active members of your own network, the memberships you oversee. This is the only scope with live data today.
• Regional would compare against restoration operators in the same region, drawn from Verinode's broader intelligence layer rather than your own network.
• National would compare against restoration operators nationwide, same source.

The pill label for Group shows the live peer count in parentheses, for example "Group (14)". Regional and National show plain labels with no count, because there is no cohort behind them yet (see "Why Regional and National read as pending" below).

Clicking an available pill re-resolves every metric tile in the slider against that scope, instantly, no reload. Clicking a disabled pill does nothing except show its tooltip on hover, explaining what has to happen before it lights up.

The Group scope and its within-network floor

Group scope draws its peer set from your network's own reputation summary: every other membership marked Active in your roster, with a valid value for the metric in question. A franchisee's own row is always excluded from its own peer set.

Before that peer set can produce a median, it has to clear a floor. Group needs at least three active peers with a usable value for that metric. Below that floor, the Group pill itself goes disabled with the tooltip "Need 3+ active peers in the network for within-network comparisons," and the metric tiles fall back to a "Cohort too small" state instead of showing a number.

This floor exists because a median built from one or two other franchisees is not a benchmark, it is closer to naming names. Even inside your own network, where you already see the aggregate roster, Verinode holds the same anonymity discipline it applies everywhere else: no comparison surfaces until there is a genuine group behind it.

Two things worth knowing about how the Group peer set is built:

• Per-metric peer sets, not one shared list. Each of the five metrics (composite score, Google rating, Google review count, Yelp rating, Yelp review count) builds its own peer list from whichever active franchisees have a non-null value for that specific metric. A franchisee missing a Yelp presence simply doesn't contribute to the Yelp peer set, but still counts toward Google's. This means the group count next to the pill label reflects the composite score cohort specifically. Other metrics on the same slider can have a slightly different peer count behind them even at the same Group scope.
• Status matters, twice. A franchisee only enters the peer set if it is marked active in the group roster. A franchisee that has been seeded or invited but hasn't gone live yet, or one that has left the network, never contributes to anyone else's median.

Reading the confidence tier

Whenever Group scope is showing a real number, the scope-context line under the metric tiles states the tier out loud: for example, "Within-network comparison, Directional, n=11." The tiers, from strongest to weakest signal:

• High confidence (a substantial number of peers behind the median). The most stable read: a franchisee's percentile here is unlikely to move much if one more member joins or leaves the cohort.
• Directional (a real but modest cohort). Read the trend, not the decimal. A gap of a point or two against the median is noise at this size, a five- or ten-point gap is signal.
• Low confidence (the cohort is small enough that the median can swing meaningfully from any single franchisee's number changing). Useful as a rough compass, not a scorecard.
• Hidden (the cohort hasn't cleared the within-network floor at all). No median or percentile is shown; the tile explains why instead.

The tier is driven purely by how many active peers are behind the number, nothing else. Verinode doesn't publish the exact peer-count thresholds that separate one tier from the next inside product copy, by design: the tiers are meant to be read as "how much weight can this number carry," not reverse-engineered to infer how many franchisees are in a given bucket.

Two failure states look similar but mean different things. Cohort too small (shown on the metric tile itself, "Cohort too small") means the scope hasn't cleared the anonymity floor at all, no comparison is being shown. Confidence tier: Low means a comparison IS being shown, it just carries a lighter interpretive weight than a High or Directional read would. If a metric tile shows a number with "Low confidence" next to it, trust the direction more than the magnitude.

What each metric tile shows

Every metric tile in the slider (Composite score, Google rating, Google reviews, Yelp rating, Yelp reviews, plus the standalone BBB rating tile) follows the same layout:

• The franchisee's own value, large, at the top. A rating shows to one decimal (e.g. "4.6"), a review count shows as a whole number, and a missing value shows as a dash.
• Median [value], the peer median at the currently selected scope, formatted the same way as the franchisee's own number.
• · p[N], the franchisee's percentile rank within that peer set, appended right after the median when both a target value and a median exist. A percentile of p75 means the franchisee sits above roughly three-quarters of its peer group on that metric.
• A delta line below the median, reading something like "+12% vs median" or "On par with peers." This compares the franchisee's raw value against the peer median and colors itself by whether that gap is favorable, factoring in whether the metric is "higher is better" (all five reputation metrics are: a higher rating, and a higher review count as a stronger credibility signal, both read as better). A gap under roughly 5% either way reads as "On par with peers" in neutral gray. A favorable gap reads green (Deere Green, the Expand signal color). An unfavorable gap reads yellow (Hard Hat Yellow, Maintain) if the gap is moderate, or red (Ember Red, Analyse) if the gap is large.
• A network target tag, only when your HQ has an active program target set for that specific metric. It reads "✓ Meets target" in green or "✗ Below target" in red, with the target value shown alongside (e.g. "(target 4.5)"). Hovering the tag shows which program declared the target. This tag is independent of the peer comparison, it compares against a target your HQ set, not against the peer median.

When the median isn't available at all, the tile explains why in place of a number: "Cohort too small" (Group scope, floor not cleared) or "Pending benchmark seed" (Regional or National, see below).

The BBB rating tile is the one exception with no peer comparison: it shows the franchisee's letter grade with the caption "Letter grade, external assessor," because BBB ratings aren't the kind of continuous metric a peer median applies to.

Composite score and trend

At the top of the slider, above the per-platform breakdown, sits the Composite reputation section with the Composite score tile and, next to it, the review trend pill in the header. The header pill reads one of:

• Trend: improving (green)
• Trend: stable (yellow)
• Trend: declining (red)

If no trend is available, the header pill falls back to a review-count summary instead: "[N] reviews" in green if there are any, or "No reviews" in neutral gray if there are none. The composite score itself is a single 0 to 100 number blending the franchisee's rating and volume signals; it carries the same Group-scope peer comparison and confidence tier as every other metric.

Why Regional and National read as pending

Both the Regional and National pills are disabled today, with the tooltip "Industry benchmarks not yet seeded for reputation metrics. Coming in the next slice." This isn't a bug and it isn't a paywall, it's an honest gap: those two scopes are meant to compare a franchisee against the broader restoration industry outside your own network, and that requires an industry-wide reputation dataset flowing into Verinode's intelligence layer. That dataset hasn't been seeded for reputation metrics yet.

If you click into a metric tile while Regional or National happens to be selected (which can only happen before you've clicked back to Group, since the pills themselves are disabled), the scope-context footer states this directly: "Regional industry benchmarks pending, reputation metrics will seed via the benchmark data in the next slice." (National reads the same sentence with "National" substituted.) The metric tiles fall back to "Pending benchmark seed" in place of a median.

Once industry-wide reputation data is flowing, Regional and National will resolve the same way Group does today: a real median, a real percentile, and a confidence tier reflecting how many industry peers are behind the number, at whatever region or national level you've selected. Nothing about the Group scope or its floor changes when that ships, Regional and National are additive comparison lenses, not a replacement for the within-network view.

The HQ privacy boundary, held throughout

Every number in this drill-in is aggregate. HQ leadership sees a franchisee's own composite score, ratings, and review counts (these are the franchisee's own data, which HQ has always seen as part of the roster), and it sees peer medians and percentiles computed across the rest of the network. It never sees which specific peer franchisees make up that median, their individual scores, or a ranked list behind the aggregate. The within-network floor and the confidence tiers exist specifically so that even inside your own network, a "peer comparison" can never quietly become "here is exactly how the other two franchisees in this cohort are doing." Franchisees own their underlying data; HQ sees aggregates, rankings, and compliance, not another membership's private numbers.

Best-practice example

Say you open a franchisee's Reputation drill-in and the Group pill reads "Group (11)." The Composite score tile shows the franchisee at 72.4, "Median 68.1, p61," with a green "+6% vs median" delta, and the scope-context line reads "Within-network comparison, Directional, n=11." Read this as: the franchisee is modestly ahead of its network peers on composite reputation, but with eleven peers behind the median, treat the six-point gap as a real signal, not a precise ranking. If the same franchisee's Yelp review count tile instead reads "Cohort too small," that means too few active memberships in your network have a usable Yelp review count for a comparison to publish safely, not that the franchisee itself lacks Yelp reviews. Clicking Regional or National on either tile, once those scopes go live, will add an industry-wide read alongside the within-network one, without changing what Group shows.

Related articles

• /help/network-health
• /help/hq-benchmarks
• /help/hq-standards
• /help/hq-programs
• /help/hq-compliance
• /help/hq-overview