Reading a result

Summary#

Every tool result in Metric Vault is built from the same set of parts, in the same order: a source badge, a summary, then tabs that group the detail by what it is for. Once you can read one report you can read all of them. This page explains each part, what the tabs mean, and how to tell measured data from an AI reading of that data.

Purpose#

A platform with seventy tools can easily become seventy report formats. That costs you time on every single run, because you have to find the important number again each time. So results here are not laid out by tool; they are laid out by a shared result shell that every renderer feeds into.

The shell also solves a second problem. A full report can be very long. Putting everything on one scrolling page buries the one number you opened the tool for. The category tabs put a short summary first and move the detail behind labelled tabs, so the answer is at the top and the evidence is one click away.

Overview#

A finished result reads top to bottom like this:

1. Source badge. A colored strip stating where the numbers came from.
2. Summary. The headline verdict, the KPI tiles, and the chart that matters most.
3. Category tabs. A row of pills that switch between groups of detail.
4. Export row. PDF, Excel, CSV, JSON and a public share link.
5. Recommendation panel. An offer to turn the result into an action plan.
6. Explainer. The illustrated guide to the tool, below everything else.

Before the result exists, the same area shows one of three states.

How it works#

The source badge#

The first thing in every result is a badge saying how the numbers were produced. This exists because a report that mixes measured data with generated narrative is untrustworthy unless it tells you which is which.

Where a run carries a note from the provider, it appears at the right of the same strip. See Where the data comes from for who the providers are.

Summary first#

Directly under the badge sits the summary: the KPI tiles, the headline chart and the verdict. KPI tiles are a label, a large value and a small qualifier underneath. They are always the tool's own primary metrics, so they are the fastest place to see whether anything has changed since your last run.

The category tabs#

About a tenth of a second after a result renders, its sections are sorted into categories and a tab bar appears. The first tab is always Summary. The rest appear only if the result actually contains that kind of content, in this fixed order:

A section whose title does not fit any category keeps its own title as its tab label. Tab labels are trimmed to 34 characters so the bar stays on one line.

Right after the last tab come two buttons: View Full Report, then Get recommendations (the same action as the panel below the result), each separated by a thin divider.

When a result is long but has no sections#

If a result is too long to read comfortably but does not divide into sections, it is clamped to a readable height with a soft fade at the bottom and a bar appears offering Expand here and View Full Report.

View Full Report#

On a tabbed result, View Full Report does two things at once. It reveals every section of the report in place, in order, and lights up like a selected tab, and it opens a clean, standalone copy of the whole report in a new browser tab. Click any tab to go back to a single category.

The in-place state decides what the export buttons cover: on a tab they export that tab's content, in the full-report view they export everything. The new tab is built the same way as a public share link, so it needs a signed-in session and a finished result. See Exporting results and Sharing a result by link.

On a long result with no tab bar (the clamped kind), the View Full Report button in the bar at the bottom instead builds a clean, standalone copy of the whole result and opens it in a new tab — the same mechanism as a public share link, so it needs a signed-in session and a finished result. See Sharing a result by link.

Empty results are honest, not blank#

Thirty of the data tools are marked real-data tools. If the provider has nothing for your query, they say so instead of inventing a plausible number:

Several tools go further and explain the specific limit. Social Analytics names the platforms it can and cannot see. Advertising Research reports $0 with the note that no paid advertising was detected, rather than guessing a spend. Influencer Analysis explains that influence is measured by citation count and page authority, never by follower counts it cannot verify. Sections inside an otherwise full report use the same convention, with titles such as "No paid ads found" or "No related keywords surfaced".

This is deliberate. A number you cannot trust is worse than no number, because you cannot tell which is which after the fact. See A tool returned no data for what to try next.

Exports are disabled until there is something to export#

Export buttons stay dimmed, with the tooltip Run the analysis first, until the result area holds a finished report. Starting a new run dims them again. If you click an export before a result exists you get Run the tool first — there is no result to export yet.

The export format is bound to the button's color rather than its label, so exports work identically in all six interface languages.

Charts#

Charts are drawn in the page and are captured as images when you export to PDF or create a share link. A chart with no data is dropped from the copy along with its heading, so a shared report never contains an empty frame. See Charts and visualizations.

Screenshots#