Vendor Score deep-dive: history, trend, and research narrative

What this is

The Verinode Score on a vendor tile is a single number. The Score deep-dive is where that number opens up: how it has moved over time, which dimensions are pulling it up or down and by how much, how much each dimension is even supposed to count for a vendor of this type, and the research-layer narrative explaining why the score reads the way it does. It is a modal that opens on top of the vendor detail overlay, not a separate page, so closing it drops you right back where you were.

Everything in this modal is platform-wide research data: the same the benchmark data scoring engine, the same score history, the same LLM-generated narratives that an IQ operator sees on their own vendor card. Nothing here is specific to your network. One thing is deliberately missing: your own franchisees' private team rating of this vendor. That rating lives in each operator's own business data and never crosses into an HQ surface, even though the panel underneath it is capable of showing one. More on that below.

Where to find it

Open Vendors from the HQ sidebar at hq.verinode.ai/vendors, click any vendor tile to open its detail overlay, then click the Score deep-dive button at the top right of that overlay's header. The button is titled "Score history, per-dimension trend, category weight profile, research narrative" on hover, and the modal that opens is titled with the vendor's name followed by "Score deep-dive," for example "ServPro National · Score deep-dive."

The button only appears when two things are both true: the vendor has cleared your network's minimum-cohort floor (see What HQ sees vs. what stays private: the vendor aggregate boundary), and Verinode's research layer has actually computed a score for this vendor. If either condition isn't met, there's no button, and no separate empty state to click into: a vendor with no research score simply has nothing to deep-dive into yet, and a vendor that hasn't cleared the cohort floor opens to the aggregate-only message described in Opening a vendor: the network detail overlay instead of a full overlay with buttons.

Section by section

The modal stacks a handful of panels top to bottom. Each one hides silently when it has nothing to show, so a thin vendor record produces a short modal, not a modal full of blank boxes.

1. Hero recap

At the top: the vendor's category in plain language (for example "Equipment," "Software," "Subcontractors"), plus its specific catalog category name when one is on file, for example "Verinode Score · Equipment · Water Mitigation Equipment." Beneath that:

• The score itself, one decimal place, on a 1.0 to 10.0 scale.
• A label badge: Strong, Solid, Mixed, or Weak, color-coded from green for Strong down through copper and yellow to red for Weak.
• Coverage, phrased as "N% of signals in." This is the share of the score's total possible weight that actually had usable data behind it. A vendor scored on every applicable dimension shows close to 100%; one with several dimensions excluded for lack of data shows lower. It is a transparency signal, not a confidence grade.
• When the score was last computed, in relative terms: "today," "yesterday," "Nd ago," "Nw ago," or "Nmo ago."

2. Verinode's read on this

When the research layer has generated a cross-signal synthesis for this vendor, a short paragraph appears directly under the hero, under the eyebrow label "IQ's read on this." It weaves together everything Verinode knows about the vendor into one plain-language read: the score, any peer signal, program terms, and recent research together, rather than each fact in its own box. This paragraph reuses the same synthesis component IQ operators see on their own vendor cards, which is why the eyebrow still says "IQ's" even inside the HQ panel; the content itself is the platform-wide synthesis, identical for every viewer. When no synthesis has been generated yet for this vendor, the paragraph is simply absent and the modal moves straight to the next section.

3. Verinode Research

Below the synthesis, a card labeled Verinode Research carries the vendor-level summary narrative: one paragraph explaining, in plain prose, why the score sits where it does. This is generated nightly alongside the scoring run and is the same narrative an IQ operator reads on their own vendor detail card. When no summary narrative exists yet for this vendor, the card is omitted.

4. Why your team's rating isn't here

If you've used the vendor detail overlay or the IQ side of the platform, you may expect a "Your Team's Rating" block here, showing an operator's own headline rating, a team average, and per-dimension survey scores. It never appears in the HQ deep-dive. That block reads from your operator data, a single franchisee's own private business data, and the deep-dive panel is explicitly called with that rating set to nothing when it renders inside HQ. This is not a bug or a missing feature: a franchisee's own opinion of a vendor is their business, not the network's, and surfacing it cross-tenant would breach the same privacy contract that governs every other HQ surface. Your network's aggregated dimension medians (built from the same underlying ratings, but K-anonymity gated and never attributable to one franchisee) live one level up, on the vendor detail overlay itself under "Network team scores," not in this modal.

5. Score Trajectory

A sparkline plots this vendor's score across its recent history, oldest on the left, most recent on the right, with a small dot marking the latest point. Below the chart, the earliest score in the window sits on the left, a count of snapshots in the middle ("N snapshots"), and the most recent score, bolded, on the right.

Empty state. With fewer than two snapshots on file, the chart is replaced with a line reading: "Not enough history yet. The trend line appears as new data comes in."

6. Dimension Breakdown

This is the core of the deep-dive: every scoring dimension that applies to this vendor's category, ordered so the dimensions that actually moved the score lead, and the ones that didn't sink to the bottom. Each row shows:

• The dimension name (Market Trust, Peer Intelligence, Cost Position, Integration & Ecosystem, Switching Cost, AI & Innovation, Risk & Compliance, Industry Alignment, Operational Fit, Vendor Trajectory, ESG, or Feature Depth, depending on the vendor's category).
• Its weight, as a percentage of the total score, next to the name. A dimension with zero weight for this vendor's category (present in the model but not counted toward the score) shows a Context only tag instead of a percentage, so it never reads like a weakness.
• An Excluded tag, when this specific vendor didn't have enough data on this dimension for it to contribute. Its weight was redistributed across the remaining dimensions. Where Verinode has a reason on file, it appears underneath the row in italics.
• The dimension's own score, one decimal, out of 5, with a colored bar underneath: green at 4 and above, copper from 3 up to 4, red below 3. Excluded dimensions show a flat grey bar instead.

Rows with indicator detail carry a small disclosure arrow, and a header note across the top of the section reads "Tap a row to see indicator-level breakdown." Click any such row to expand it in place. An expanded row can show, in order:

• A per-dimension narrative, when the research layer has written one specifically for this dimension, explaining what's driving that particular score.
• A Trend mini-chart, when at least two historical snapshots exist for this dimension: a small sparkline plus "first score → last score" read left to right.
• Up to 8 indicators, the raw data points behind the dimension's score. Each indicator shows its label, a confidence pill (Verified, Observed, Reported, Estimated, or Stale), its own contribution score out of 5, the raw underlying value when there's something displayable (a number, a Yes/No, a short string, or an item count), any operator-facing note Verinode has attached to it in italics, and, when known, how recently that data point was refreshed ("Refreshed today," "Refreshed 3d ago," and so on). If more than 8 indicators exist, a line underneath reads "+ N more indicators."
• If none of the above exists for a dimension (no narrative, no indicator detail, fewer than two history points), the expanded row simply reads: "No indicator detail captured for this dimension yet."

7. How Much Each Dimension Counts For [category]

A reference panel, not a rating: a horizontal bar for every dimension in this vendor's scoring category, showing its nominal weight as a percentage. This is the model itself, not this vendor's outcome. It's here so you can read a low dimension score in context: a dimension carrying real weight (Cost Position on an equipment vendor, say) is worth acting on if it's weak; a dimension the category barely counts (say ESG on a subcontractor) is not. For SaaS vendors specifically, Feature Depth appears in this list even though it's computed outside the standard weighted dimension set, because it's the single dimension that matters most for evaluating software.

8. How The Score Is Computed

A closing methodology note, present on every deep-dive regardless of what data exists above it: the Verinode Score combines weighted dimensions tailored to each vendor category on a 1.0 to 10.0 scale, dimensions with insufficient data are excluded and their weight redistributes across the rest, coverage for this vendor (the same percentage shown in the hero) reflects how much of the total possible weight actually contributed, and scores update as new data comes in.

How to read a deep-dive in one pass

1. 1Start with the hero: the score, the label badge, and the coverage percentage. Low coverage is a reason to weight the number less heavily, not a sign of anything wrong with the vendor.
2. 2Read the Verinode Research narrative (and the synthesis paragraph above it, when present) before the dimension rows. It tells you in plain language why the score sits where it does, which makes the rows below easier to scan quickly.
3. 3Open the Score Trajectory chart. A vendor trending up is a different conversation than one that's been flat or slipping, even at the identical current score.
4. 4Scan the Dimension Breakdown from the top, since it's already sorted by what's actually driving the score. Tap into any dimension that looks surprisingly low, especially one carrying real weight in the category profile below it.
5. 5Check the category weight profile if a low dimension score gave you pause. A weak dimension the category barely weights is background noise; a weak dimension carrying real weight is worth a conversation with the vendor or a look at alternatives.

Best-practice example

Say you open the deep-dive on a materials supplier scoring 6.8, labeled Solid, with 78% of signals in and last computed two days ago. The trajectory chart shows the score climbing steadily over eight snapshots, from 6.1 up to 6.8. The Dimension Breakdown leads with Cost Position at 32% weight, scoring 7.9, and Market Trust at 24% weight, scoring 7.2, both strong and both weighted heavily for a materials vendor. Further down, Risk & Compliance carries an Excluded tag with the note "No compliance certification data on file for this vendor," and ESG sits at the bottom tagged Context only. Read together: this is a vendor genuinely improving on the two things that matter most for a materials supplier, with one real data gap (compliance certification) worth asking the vendor to close, and nothing else in the breakdown worth acting on.

Related reading

• Opening a vendor: the network detail overlay: the overlay this modal opens from, including the Score deep-dive button's location
• What HQ sees vs. what stays private: the vendor aggregate boundary: the full privacy mechanics, including why the operator's team rating never appears here
• Vendor detail: score and composition: the always-visible dimension breakdown on the overlay itself, and how it differs from this modal
• Comparing vendors across your network: the Compare alternatives button that sits next to Score deep-dive in the overlay header
• Vendors: your network's spend and procurement leverage, in aggregate: the Vendors home every vendor tile opens from
• HQ Benchmarks: how the same research-layer scoring surfaces at the network level

Data sources