Reading the evidence on a signal tile (coverage grid, trend line)
When Verinode surfaces a signal, wherever a signal renders as a wide Take Action tile, it does not just print a paragraph of colored text. The tile carries a small icon that identifies what kind of…
On this page
What this is
When Verinode surfaces a signal, wherever a signal renders as a wide Take Action tile, it does not just print a paragraph of colored text. The tile carries a small icon that identifies what kind of signal it is, and a dedicated column on the right that shows the evidence behind it as a picture: a grid of dots for "how many of the group this touches," or a small line chart for "how this has been moving." The goal is the same one a fractional COO gives you verbally, "here's the number, and here's the shape of it," instead of a wall of text you have to parse sentence by sentence.
This article is about that icon and that evidence column specifically, the shared building block behind every wide signal tile. For the badge that shows a signal count on an individual vendor, job, client, or carrier card, see signal count badges. For how signals get detected in the first place and where the old Pipeline page moved to, see Signals overview. For the entity-and-lifecycle version of this same evidence column on IQ's own Take Action tiles, see Take Action and Explore on Jobs.
Verinode does not decide anything on your behalf here. It reads your data, a detector notices a pattern, and the tile is Verinode showing you the evidence as plainly as it can. You open the tile, you read it, you act, park, or dismiss it.
Where you'll find it
The tile shape covered in this article, an icon plus a coverage grid or trend line in a divided right-hand column, appears on the wide "Take Action" tiles across Verinode | HQ's network surfaces:
- Network (sidebar Network,
/franchise/network): the network home's Take Action row, one tile per critical cross-network signal. - Signals (sidebar Signals,
/franchise/signals): the Critical & High Severity row.
These tiles only ever appear at the wider, double-width size with the accent-colored left rail (Verinode's "this needs you" treatment). The narrower rows on the same pages, By Domain, Recent, and Resolved, use the plain tile shape: a label, a headline, a caption, and nothing else. No icon, no evidence column, on a standard-width tile there simply isn't room for a divided visual, so Verinode doesn't try to squeeze one in.
IQ's own per-section Take Action rows (Jobs, Vendors, Clients, Team, Equipment, Safety, Compliance, Recruiting, Certifications, Reputation, Margin) use the identical tile shape and the identical icon logic described below, so a signal reads the same everywhere in the product. Their evidence column shows something different, though: the entity's logo, a row of lifecycle dots, and a trajectory word, because an IQ decision is about one entity in your own book rather than a share of a group. That version is covered in Take Action and Explore on Jobs. This article covers the coverage-grid and trend-line version specifically.
How the tile picks its icon
Every signal tile carries a small glyph at the top left, in the tile's accent color, before the severity label. Verinode picks it automatically, it is never something you set. The logic runs in two passes:
- 1Check the signal's specific type first. A handful of keyword hints look for a substring inside the signal's type, and the first one that matches wins: anything about margin gets the Margin icon; anything about cash, pay, or collections gets the Costs icon; anything about a certification gets the Certifications icon; safety or an incident gets the Safety icon; staffing or headcount gets the Team icon; a vendor or a rate gets the Vendors icon; a review or reputation gets the Reputation icon; a stall, a drift, or a conformance gap gets the Process icon.
- 2Fall back to the signal's domain. If nothing in step 1 matched, Verinode looks at the domain the signal was filed under and maps it to that section's own icon, the same glyph you'd see next to that section in the sidebar: Margin, Costs (billing or cash-flow signals), Risk (carrier and weather signals), Clients (client and commercial signals), Jobs (field signals), Team, Safety, Vendors, Equipment, Fleet, Facilities, Reputation, Certifications, Audits (compliance signals), Process, Benchmarks (competitive and growth signals), Recruiting, or Settings.
- 3Otherwise, the generic signal glyph. If neither the type nor the domain matches anything specific, the tile falls back to Verinode's plain signal icon, still colored to the tile's severity, just without a more specific shape.
Note
The icon is a visual hint, not extra information you need to read separately. It exists so a row of tiles reads at a glance, "that one's a vendor thing, that one's a safety thing," before you even read the headline.
The evidence column
On a wide Take Action tile, the body splits into two: your text on the left (the severity label, the signal's title as the headline, a caption line, and a detected-date footer), and a divided visual column on the right, separated by a thin vertical line in the tile's accent color. That right column is where the evidence lives, and it renders one of two ways depending on what the signal actually carries.
The coverage grid: "X of Y affected"
Most cross-network signals carry an affected count and a total count, how many members of the group the pattern touches, out of how many were checked. When both numbers are present, Verinode renders a grid of small dots: one dot standing in for a share of the group, with the affected share filled in the signal's severity color (Ember red for critical, Hard-hat yellow for a warning) and the rest of the dots dimmed in the tile's base accent. Under the grid, a caption spells the same thing out in text: "{affected} of {total} · {percentage}%", for example "6 of 22 · 27%". For a very large group, the grid caps at a fixed maximum size for legibility, past that point each dot represents a proportional share of the group rather than exactly one member, but the "X of Y" caption above it always states the real numbers.
On the Network home's Take Action row, the caption line above the tile's evidence column also states this in plain language, for example "6 of 22 franchisees affected." When a signal doesn't carry an affected/total pair, that line falls back to naming the signal's domain instead.
If a signal has no affected count, no total count, or a total of zero, the coverage grid does not render at all, Verinode does not draw an empty or misleading grid. On tiles where a trend is available as a fallback (see below), the tile shows that instead. Where neither is available, the tile simply has no evidence column: text on the left, nothing on the right, exactly as if the divided column were never there.
The trend line
Where a signal is about something drifting over time rather than a share of a group, a labor-margin trend worsening, a metric moving off your own historical norm, the evidence column shows a small line chart instead: a filled sparkline with a single dot marking the most recent point. This shape only appears for signals that fire on a worse-direction drift, so the line always renders in Ember red, it is never used to show a positive trend.
Under the line, a caption states the window it covers, for example "8-week trend." When Verinode cannot compute the exact span from the signal's evidence, the caption reads simply "Recent trend."
The trend line is a fallback: on the tiles where Verinode tries both evidence shapes (the Network home's Take Action row), a coverage grid is used whenever the signal has a valid affected/total pair; the trend line only appears when that pair is missing but the signal carries at least two data points to plot, a stored series when one was recorded, or a two-point line from a prior value compared against a recent one. On the Signals page's Critical & High Severity row, only the coverage grid is used, there is no trend fallback there, so a signal without an affected/total pair on that page shows no evidence visual at all rather than a line chart.
Reading it together
A tile's icon tells you what kind of thing you're looking at before you read a word. Its evidence column tells you the shape of the proof: a dot-grid means "here's how widespread this is across the group," a red trend line means "here's how this has been getting worse." Put together with the headline and the detected-date footer, the tile is built so you can decide whether to open it in the time it takes to scroll past it.
Best-practice example
On the Network home, a Take Action tile reads Critical, headline "Contractor Connection program cost drifting above peer rate," with a Vendors icon in the header. The evidence column shows a coverage grid, filled dots read "6 of 22 · 27%", telling you this isn't one office's fluke, more than a quarter of your network is paying above the peer rate on the same program. Next to it, a second tile reads Watching, with a Margin icon and a red trend line captioned "6-week trend," no affected/total pair this time, just a worsening line. The first tile tells you to negotiate the program rate network-wide; the second tells you to watch a metric that's sliding before it becomes a network-wide pattern too.
Empty states
- No evidence column at all. The tile still renders, label, headline, caption, footer, just without a right-hand visual. This happens whenever the signal carries neither a valid affected/total pair nor a plottable trend series.
- No signals in a row. The row itself explains why in plain language rather than showing an empty tile: the Network home's Take Action row reads "No open signals. Cross-network warnings appear here when ≥30% of active franchisees hit the same pattern." The Signals page's Critical & High Severity row reads "No critical or high-severity signals currently open across the network."
Neither is a broken screen. It means the aggregator hasn't found a pattern that clears its threshold yet, not that Verinode has stopped watching.
Data sources
- 1.Your operator signals, aggregated to the network level. Your business.
- 2.Cross-network detection pass (aggregator threshold on shared patterns). Verinode reference data.