Reputation Findings and decisions

What this page shows

Reviews and standing tell you where your reputation sits today. Findings is where Verinode turns that picture into something to act on. As reviews land, response times slip or hold, carrier scorecards diverge from what customers are actually saying, Verinode's detector reads the pattern and writes it up as a decision: what's happening, what it's costing you (in dollars when it can be priced, in plain risk language when it can't), and a recommended next step.

Verinode does not decide anything for you. It surfaces the pattern, prices the impact where it can, and recommends a move. You read it, and you click Act, Not now, or Ignore.

This article covers the two places reputation decisions actually show up: the Take Action row on the Reputation home screen, and the Findings tab inside the Reputation card slider, plus what happens when you open one. For the Trust Score and the platform-by-platform breakdown, see Your Trust Score and How you're doing. For the review list itself, see The Reviews list.

Where to find it

Open Reputation from the sidebar at iq.verinode.ai/reputation. The home screen opens with rows stacked top to bottom, and the first content row is Take Action, the row this article starts with.

Further down, the Overview row's tiles open the Reputation card slider, a three-tab deck: Reviews (violet), How you're doing (green), and Findings (copper). You can reach Findings directly from a decision tile in Take Action, or by clicking through to the slider and switching to the Findings tab yourself.

The header carries two standard controls on every Reputation page: Manage profiles (connect or disconnect Google, Yelp, BBB, Facebook, or Trustpilot) and Send data, for forwarding anything reputation-related straight to IQ. See Forwarding documents for how that works.

The Take Action row

This is the first row on the Reputation home screen, and it always opens with two fixed tiles before any decisions:

1. An agent-conversation tile. A single tile that opens IQ's chat panel with a reputation-specific prompt already loaded, so you can ask a question or describe a problem instead of waiting for a decision to surface on its own. It retires once you've used it.
2. Profile Bio. A generator tile that drafts an "About your company" bio you can paste onto your Google or other review profile, built from your company name, service mix, city, state, and years in business. It's a writing tool, not a decision, it doesn't count against the decisions described below.

After those two, the row fills with your actual reputation decisions, each rendered as a tile built from the same shared decision-tile component every section's Take Action row uses. A decision tile shows:

• Recommended, a small eyebrow label. Every decision the row shows you is a live recommendation, there's no other posture at this stage.
• The headline. When the decision carries a priced dollar impact, the headline is that dollar figure with its period (/mo or the matching annual suffix), and a line underneath names what the number is about (for example, "Contractor Connection: response time drifting from peers"), so a bare dollar figure is never left floating without context. When there's no dollar figure yet, the headline is the action itself in plain words, for example "Respond to negative review on Google," instead of a raw metric name.
• The entity column, to the right: a small logo or initials badge for the platform or carrier the decision concerns, its name, a row of lifecycle dots (more on those below), and, when the underlying metric has a direction, a trajectory word (Improving, Stable, Declining) with a matching arrow.
• A trajectory arrow in the corner badge when the metric is moving, colored to match: Ember Red trending down, Deere Green trending up.

Clicking a decision tile opens it directly as a full-screen decision detail, the same view described under "Drilling into a decision" below.

The four empty states

The Take Action row never shows a blank space. Depending on where you are, it shows one of four honest states:

• Loading. Three pulsing placeholder tiles while Verinode fetches your decisions.
• No connected platform yet. "Connect a review profile to start." "Click a platform tile below to connect your Google Business Profile (or paste a URL for Yelp / BBB / Facebook / Trustpilot). Decisions surface as reviews land + nightly detection runs."
• You're caught up. Once you've worked through at least one past signal and nothing new is open, "Reputation is in good shape." "You've worked through [N] signals. New ones surface here as the detector finds them, until then, nothing needs your attention." (Singular "signal" when the count is one.)
• Still building a picture. Platforms are connected, but nothing has surfaced yet and nothing has been resolved before either: "Still learning your reputation." "As reviews accumulate and the detector analyzes per-platform trends + carrier scorecard divergence, top decisions will appear here."

None of these mean something is broken. They mean the detector hasn't found anything worth flagging yet, or you've already cleared what it found.

The Findings tab

Switch to the copper Findings tab inside the Reputation slider (or open it from a decision tile) and you get the full gallery: every open reputation decision, not just the handful the home screen has room for.

The header line reads "[N] decisions to review" (or "1 decision to review" when there's exactly one). Below it, a responsive grid of tiles, one column on a phone-width screen, two on a tablet, three on desktop.

Empty state. "No decisions for Reputation right now. As your agent finds patterns, they'll appear here."

Reading a tile

Each tile in the grid carries a compact version of the same information:

• A colored top bar naming the decision's status: Ember Red for Urgent, Hard Hat Yellow for Pending, Deere Green for Acted or Resolved, IQ Teal for Parked, and grey for Ignored. A matching small status pill sits in the top-right corner for anything that isn't Acted or Resolved.
• A faint watermark icon in the background, unique to the Reputation domain, so the tile reads as reputation-related at a glance even in a mixed gallery.
• The dollar impact (when there's one to show), the biggest thing on the tile, with its period underneath. Under it, the decision's title, the specific finding in plain language.
• When there's no dollar figure, a small risk-kind and urgency chip instead ("Compliance risk," "Business risk," paired with "Urgent," "Act soon," or "When you can"), then a short consequence line, then the title.
• An ACTED or DONE stamp, rotated and rubber-stamped across the tile once you've committed to a decision, so a worked-through tile reads as finished at a glance. While the agent is actively drafting a plan for a decision you just acted on, the stamp reads Planning instead, with a pulsing dot, so you can see the work is in progress.
• A button row along the bottom, matched to the decision's status: a pending decision shows Act, Not now, and Ignore; a parked one shows a single Resume; an acted decision shows Edit plan; a resolved one shows Review outcome. Clicking the tile itself, anywhere outside that button row, opens the full decision.

Drilling into a decision

Click a tile, from either the Take Action row or the Findings tab, and the full decision opens as its own screen inside the slider.

The back link. A small "← All decisions" link at the top returns you to the Findings gallery. You can also page between decisions with the left/right arrow keys, a trackpad swipe, or the floating side arrows the slider provides, without going back to the gallery first. An "← / → or swipe to navigate decisions · Esc to go back" hint sits at the bottom of the card.

The header row. A small dot colored by trajectory (pulsing Ember Red if declining, Deere Green if improving, Hard Hat Yellow if stable), a status label (Declining, Needs context for low-confidence findings, or plain Decision), the entity name if the decision is about one, and a confidence pill on the right (high, medium, or low).

The impact number. When Verinode has a priced dollar figure, it's the largest thing on the screen: the annualized dollar amount at stake and its period, with a small tag underneath reading Calculated (built directly from your own operator data) or Estimated (a working assumption pending more data), sometimes followed by a refinement note like "Peer cohort" or "Industry baseline" showing what kind of estimate it is. A short sentence beside the tag explains, in plain language, how Verinode got the number.

When there's no dollar figure to show (typical for compliance-flavored reputation findings), you instead see a risk-kind and urgency chip, the same "Compliance risk" / "Business risk" plus "Urgent" / "Act soon" / "When you can" pairing from the tile, followed by a plain consequence sentence.

The title. The finding itself, in one bold sentence, for example a specific platform, metric, and direction.

Lifecycle dots. Four stages, Flagged → Planned → Acting → Resolved, showing exactly where this decision sits: freshly surfaced, a plan exists but hasn't started, steps are underway, or every step is done.

The recommendation tile. When Verinode has a concrete next step, a copper-tinted box labeled Recommended action states it plainly, followed by "If you don't: [consequence]" when there's one to show. If the decision's confidence isn't high yet, a note below it, "Refine this estimate: [what to send]," tells you exactly what would sharpen the number, forwarding a specific document or answering the agent's next question, not a vague "add more data" nudge.

When Verinode can see a pattern but doesn't have enough of your own context to recommend a specific move, the tile instead reads Needs your context, with a note that clicking Act lets the agent ask you two or three quick questions to shape a real recommendation.

Peer decision path. Below the recommendation, when there's enough peer coverage behind this specific decision, a short block tells you how many operators in your comparison cohort faced the same pattern in the last 90 days, what share of them acted on it, and, of those, what share saw the metric improve afterward. This section disappears entirely when there isn't enough peer coverage yet, it never shows a placeholder or a zero.

Acting on it. Three controls sit at the bottom: Act (commits you to the decision and opens the full action-plan workspace at /decisions/[id], where the agent drafts a step-by-step plan), Not now (opens a short reason picker, "Too busy," "Need more info," "Not convinced," or "Other," and parks the decision to resurface later), and Why? (expands a proof drawer with the specific data points and root-cause chain behind the finding, when there are any to show).

Once you've acted. The card changes shape entirely. Instead of the recommendation tile, you see the actual plan: a checklist of steps with due-date labels ("Today," "Tomorrow," "In 3d," or "[N]d overdue"), a running count ("[N]/[M] done"), and a highlight on whatever step is next. Checking a step off updates instantly, no page reload, and clicking anywhere else on the card opens the full workspace. If the agent hasn't finished drafting the plan yet, the card reads "Verinode IQ is preparing your action plan, open it to watch it come together" instead of an empty checklist.

Best-practice example

Say a Findings tile shows "$420/mo · Trustpilot response time drifting from peers," tagged Estimated, Industry baseline. Open it. The impact number carries the note "based on your median reply time vs. the industry baseline for this platform," confidence is medium, and a "Refine this estimate" line asks you to keep replying so Verinode can build your own peer-cohort comparison instead of leaning on an industry baseline. Click Act: the workspace opens with a starter plan (respond to the oldest unanswered reviews first, then set up review notifications so nothing sits unanswered again). Work the checklist, and the next time this pattern comes up, the estimate should be built from your own trajectory, not an industry number.

Data sources

Related: Your Trust Score and How you're doing, The Reviews list, The decision workspace, Forwarding documents.