> ## Documentation Index
> Fetch the complete documentation index at: https://docs.gominerva.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Analytics Dashboards Guide

> How to open Screening analytics, arrange your own dashboard widgets, and read the charts that show screening workload and outcomes.

Screening analytics is a set of dashboards your organization arranges itself. A dashboard is a grid of widgets, and each widget answers one question about your screening work: what is waiting, what changed, and how reviews turned out. You can keep a dashboard personal, share one with your team, or start from **Built-in Analytics**.

**Access:** Everyone in your organization can open Screening analytics and read the dashboards shared with them. In the sidebar, go to **Screening Analytics**. Creating, renaming, and deleting team dashboards requires the **Admin** role or above, and so does **Duplicate as team dashboard**. Personal dashboards are visible only to you, and an administrator cannot open or change someone else's personal dashboard.

> [!NOTE]
>
> Screening analytics always reads one workspace. If no workspace is selected,
> the page asks you to choose one first. If your account cannot read analytics
> in the selected workspace, the page says **Screening analytics is not
> available to you**; ask an owner or admin for access, or choose another
> workspace. Changing the address does not grant access.

**Team Dashboard in View Mode, Admin Header**

## Open Screening analytics

Open **Screening Analytics** from the sidebar, or search the command palette for "Screening analytics" (it also matches "analytics", "analytics dashboards", "dashboard widgets", and "screening dashboards").

Screening analytics opens one dashboard for you:

1. The organization default, when an administrator has set one
2. A dashboard shared with the team, when no default exists yet
3. **Built-in Analytics**, when your organization has not created a team dashboard
4. Your own personal dashboard
5. The empty **Build your dashboard** starter, when there is no dashboard to open at all

A personal dashboard is never opened in place of a dashboard shared with the team, so it stays yours until you choose it.

Opening **Screening Analytics** from the sidebar always follows that order, and Minerva does not remember the last dashboard you picked. To come back to a different one, pick it from the switcher.

## Built-in Analytics

Minerva adds **Built-in Analytics** to your organization, and everyone can open it. It cannot be renamed, edited, or deleted, and it cannot be set as the organization default. The page states the rule: **Built-in dashboard, shared with this workspace. Create a copy to make it yours.**

To work from the built-in:

* Select **Create a copy** to get a personal copy you can arrange
* Select **Customize a copy** to get a personal copy and open it ready to edit

Both dialogs open with the same reason: **The built-in dashboard can’t be changed. We’ll create a copy that only you can see.** The dialog behind **Customize a copy** goes on to say it opens the copy so you can customize it.

## Personal and team dashboards

Every dashboard has an audience, and the chip sits beside the open dashboard's name in the header.

| Audience | Chip in the header | Who can open it | Who can change it |
| - | - | - | - |
| Personal | **Only you** | Only you | Only you |
| Team, all members | **Team** | Everyone in the organization | Administrators. Everyone else works from a personal copy. |
| Team, admins only | **Admins only** | Administrators | Administrators. Everyone else works from a personal copy. |
| Built-in Analytics | **Built-in** | Everyone in the organization | Nobody. Create a copy and arrange the copy instead. |

In the switcher, a team dashboard's row sits under **Team dashboards**, and an **Admins only** dashboard carries that chip there as well. **Team** itself appears only in the header, beside the open dashboard.

A dashboard can also carry the **Default** chip. When a default is set, it is the dashboard everyone lands on from the sidebar.

> [!NOTE]
>
> A dashboard set as the organization default opens for everyone, but its
> numbers still come from the workspace you have selected. A workspace is a
> section of your organization with its own screening data. An **Admins only**
> dashboard cannot be the default, because members would open a dashboard they
> cannot read.

A dashboard belongs to your organization, not to one workspace. The same dashboard appears in every workspace you can open, and switching workspace changes its numbers to that workspace's activity. It never combines several workspaces into one report, so check the workspace selector before comparing figures.

### How sharing works

A dashboard is shared through its audience, not by inviting people:

* There are no email invitations and no public links. Sending someone the address of your personal dashboard does not share it, and a teammate who opens a team dashboard's address still needs their own access.
* A dashboard's audience cannot be changed in place. To share a personal dashboard, create a team copy; to offer a team dashboard to a different audience, duplicate it.
* An **Admins only** dashboard cannot be copied into a personal dashboard or duplicated as an **All organization members** dashboard. Its audience stays part of it, whatever charts it currently holds.
* A copy is an independent dashboard built from the last saved version of the source. Save the changes you want in it before copying, and expect later changes to the source not to reach the copy.

## Switch between dashboards

Select the dashboard's name in the header (the **Dashboards** control) to open the dashboard switcher. Dashboards are grouped by where they come from:

* **Provided by Minerva** holds the built-in dashboard
* **Team dashboards** contains everything shared with the organization
* **My dashboards** is where your personal dashboards sit

The footer of the switcher offers **New personal dashboard** and, for administrators, **Create team dashboard**.

**Dashboard Switcher**

## Create a dashboard

The **Build your dashboard** starter appears only when there is no dashboard to open, **Built-in Analytics** included. It offers **Add widget**, and administrators also see **Create team dashboard**, described as **Share a dashboard across your organization.** Nothing is saved until you confirm a widget, so opening the starter and leaving it creates nothing.

**Empty Dashboard with Build Your Dashboard**

To create a personal dashboard:

1. Open the switcher and select **New personal dashboard**
2. Enter a **Dashboard name**
3. Select **Create dashboard**

The first widget you add there also creates a personal dashboard, named **My dashboard**, which opens in edit mode focused on that widget.

To create a team dashboard, an administrator selects **Create team dashboard** and then:

1. Enters a **Dashboard name**
2. Chooses **Who can view this dashboard?**, either **All organization members** or **Admins only**
3. Optionally selects **Set as organization default**
4. Selects **Create team dashboard**

Dashboard names can be up to 120 characters, and two dashboards in an organization cannot share a name.

## Rename, copy, and delete

The **Dashboard options** menu in the header lists the actions for the dashboard you have open.

**Dashboard Options Menu**

* **Rename** opens **Rename dashboard**, where you change the **Dashboard name** and select **Save**.
* **Set as organization default** makes this the dashboard readers open first, and the menu states the consequence: **Anyone who has not chosen a dashboard will open this instead of Built-in Analytics.**
* **Create team dashboard** appears on a personal dashboard. It shares a team copy, and the personal dashboard stays personal: **This creates a team copy. Your personal dashboard stays private.**
* **Duplicate as team dashboard** appears on a team dashboard. It creates another team dashboard and leaves the current one unchanged: **This creates another team dashboard. The current dashboard will not change.**
* **Delete** asks you to confirm. For a team dashboard: **This can’t be undone. Everyone in your organization loses access to `<name>`.** For a personal dashboard: **This can’t be undone. Your personal dashboard `<name>` will be permanently removed.**

Two cases follow from who owns the dashboard:

* Members cannot change a team dashboard. **Add widget to my copy** and **Customize a copy** both fork a personal copy first, and the dialog explains why: **Team dashboards can’t be changed by members. We’ll create a copy that only you can see.**
* You cannot delete the current default dashboard. Set another dashboard as the default first, then delete this one.

A copy is saved the moment it is created. Leaving its editor with **Cancel** discards your unsaved layout changes, not the copy itself, so use **Delete** if you no longer want it.

If someone else saves the dashboard while you are working on it, Minerva asks you to review the latest version before making more changes. Nothing you typed is applied over their work.

## Add a widget

Select **Add widget** to open the picker, titled **Add a chart**. The picker is the catalogue of every chart available to your organization, grouped into six families:

* **Profiles**: who you are screening, and where they stand
* **Potential matches**: what screening surfaced, and how it was resolved
* **Work queues**: what is outstanding, and how much has been worked
* **Outcome quality**: whether the screening is finding the right things
* **Screening volume**: how much screening ran
* **Risk assessments and ID verifications**: where these stand, and how they came out

**Widget Picker**

In the picker:

1. Filter the list with the family rail on the left, or select **All**
2. Search with **Search charts** if you know what the chart answers but not its name
3. Select a chart, then set it up in **Set up this chart**
4. Select **Add widget**

A widget is added already configured, so it shows data as soon as it lands on the grid. If the catalogue cannot be read, the picker says **We couldn’t load the chart catalogue.** and offers **Try again**.

Two charts record individual reviewer activity: **Matches reviewed by person** and **Outstanding by assignee**. They can only be added to a dashboard whose audience is **Admins only**, never to a personal dashboard, an administrator's included, and the picker explains why: **These charts need a dashboard set to Admin only. Check your local requirements before sharing performance data.** **Outstanding by assignee** cannot be added yet even there; see [Troubleshooting](#troubleshooting).

## Arrange the layout

Select **Customize** to open the editor on a dashboard you can change: your own personal dashboards, and team dashboards if you are an administrator. The header then shows **Add widget**, **Undo**, **Redo**, **Cancel**, and **Save**, and **Save** stays unavailable until the layout actually changes.

**Team Dashboard in Edit Mode**

Charts stop refreshing while you arrange them, so a widget keeps its data until you leave the editor.

* Drag a widget to move it, and drag from its edge to resize it.
* Focus a widget and press **Enter** or **Space** to pick it up, the arrow keys to move it, and **Shift** with the arrow keys to resize it.
* Press **Space** or **Enter** again to place the widget, **Esc** to put it back, and **R** to switch between moving and resizing.
* Press **Home** or **End** to jump to the start or the end of a row, and **PageUp** or **PageDown** to jump to the top or bottom of the grid.
* Open a widget's menu and select **Remove** to take it off the grid.

Select **Cancel** to leave the editor. With unsaved changes Minerva asks **Leave without saving?** and offers **Keep editing**, **Discard**, and **Save**. Escape does not leave the editor.

On a narrower screen you can still add, configure, and remove widgets, but not move or resize them. The editor says so: **To move or resize widgets, use a wider screen.**

Renaming a widget is separate from this draft, and it is offered in view mode only, on a dashboard you can change. Double-click a widget's title, or use the pencil beside the title, then press **Enter** to save the new name. A rename saves on its own, so **Cancel**, **Discard**, and **Undo** do not undo it.

## Configure a widget

Each widget's controls are phrased as questions rather than settings, such as **Show as**, **Break down by**, **Filter to**, and **Group time by**. Leave a filter empty and it applies to the whole workspace, which the control states as **All statuses**, **All item types**, or **All channels**.

To change a widget that is already on the grid, select **Customize** to open the editor, open the widget's options menu, and select **Configure**. Change the controls and select **Apply**; the dialog reminds you to **Save the dashboard to keep these changes**.

**Widget Configuration**

## Set the time range

The **Time range** control in the header sets the period every chart on the dashboard uses. The presets are **Last 24 hours**, **Last 7 days**, **Last 30 days**, **Last 90 days**, and **Last year**.

The range you pick applies to your view of the dashboard. It is not saved to the dashboard, and it does not change what other readers see.

Some widgets take a range of their own. Each card has its own **Time range** control beside its title, showing **Dashboard default** while the chart follows the header. Choose a preset there to give that chart its own range, and choose **Dashboard default** to put it back in step with the dashboard.

**Group time by** sets the size of each period on a chart that offers it: **Hourly**, **Daily**, **Weekly**, or **Monthly**. Only **Profiles by status over time**, **Monitored profiles over time**, **Profiles added over time**, and **Queue backlog over time** offer **Hourly**, and **Queue backlog over time** switches to daily periods for any range older than 90 days, because hourly detail is kept for 90 days. A chart with no grouping chosen uses daily periods. Grouping is not a second time range.

Periods are counted in UTC, while chart labels use your browser's local time. A period can therefore start on a different local date than you expect, and charts do not follow a workspace timezone.

Charts that report the current state rather than a period do not follow the range at all, and the card says so: **Current state. The dates above do not change this.** Under its title, **Profile headline** carries **Current totals, not a date range.**

## Read the widgets

Every widget carries the question it answers, and the table below lists what each one measures.

| Widget | Family | What it measures |
| - | - | - |
| **Profiles by status over time** | Profiles | How many profiles sat at each stage of screening, over the period |
| **Profiles by status now** | Profiles | How many profiles are waiting at each stage right now |
| **Monitored profiles over time** | Profiles | How many profiles were being monitored, over the period |
| **Profiles with risk by category now** | Profiles | How many profiles currently carry sanctions, PEP or adverse media risk |
| **Profiles added over time** | Profiles | How many new profiles were added in each period, excluding the running total |
| **Profile headline** | Profiles | A live count of your profiles and where they currently stand |
| **Matches by status over time** | Potential matches | How potential matches were resolved, over the period |
| **Matches by risk category over time** | Potential matches | Which kinds of risk the potential matches came from |
| **Unresolved matches now** | Potential matches | How many potential matches are waiting for a decision |
| **Matches by channel over time** | Potential matches | Which channel the potential matches came through |
| **Matches arriving against matches resolved** | Potential matches | Whether the team is keeping up: work entering review against work resolved |
| **New alerts over time** | Potential matches | How many new potential matches arrived in each period, each counted once |
| **Queue backlog now** | Work queues | How much work is outstanding right now, by kind of item |
| **Queue backlog over time** | Work queues | Whether the outstanding pile is growing or shrinking |
| **Outstanding by age** | Work queues | How long outstanding work has been waiting |
| **Oldest outstanding item** | Work queues | How long the longest-waiting item has been outstanding |
| **Items worked over time** | Work queues | How much work the team got through in each period |
| **Risk categories in queue now** | Work queues | What kinds of risk are sitting in the outstanding work |
| **Risk categories in queue over time** | Work queues | Whether the mix of risk entering review is changing over time |
| **Matches reviewed by person** | Work queues | How many potential matches each person reviewed |
| **Match outcomes over time** | Outcome quality | How review decisions broke down over the period |
| **Match outcomes, latest period** | Outcome quality | How review decisions broke down in the most recent period |
| **Profiles with a True Match now** | Outcome quality | How many profiles have at least one confirmed True Match |
| **Profiles with a True Match over time** | Outcome quality | How the number of profiles with a confirmed True Match changed |
| **Screening precision** | Outcome quality | What share of reviewed matches were confirmed as True Matches |
| **Review effort per True Match** | Outcome quality | How many matches a reviewer reads for each confirmed True Match |
| **Automatic disposition review** | Outcome quality | How often Minerva's suggested decision matched the reviewer's |
| **Prediction and confidence distribution** | Outcome quality | How the model scored the potential matches it was asked about, split by hints against automatic policy |
| **Screens over time** | Screening volume | How much screening the workspace ran in each period |
| **Screening hit rate by feed** | Screening volume | Of the searches that queried each feed, how often a risk feed flagged |
| **Risk assessments by status over time** | Risk assessments and ID verifications | How risk assessments moved through their statuses |
| **ID verifications by outcome over time** | Risk assessments and ID verifications | How ID verifications came out |

The outcome quality charts carry a note on the card when a reading needs one, for example **Not enough reviewed matches to show a rate.** or **Do not add categories: a match can count in more than one.**

### Setup options by widget

The catalogue holds 33 charts. The table lists each one with how it treats time and the setup options saved with the widget. Plural filters such as Statuses, Risk categories, Channels, Item types, Queues, and Age bands accept several values, and an empty filter means every eligible value. A chart can be grouped by a period only where the table lists **Group time by**.

| Catalogue title | Time behavior | Saved setup options |
| - | - | - |
| Profiles by status over time | Time range; hourly, daily, weekly or monthly grouping | Statuses, Show as, Group time by |
| Profiles by status now | Current state | Statuses, Show as |
| Monitored profiles over time | Time range; hourly, daily, weekly or monthly grouping | Group time by |
| Profiles with risk by category now | Current state | Risk categories |
| Profiles added over time | Time range; hourly, daily, weekly or monthly grouping | Group time by |
| Profile headline | Current state | No saved setup options |
| Matches by status over time | Time range; daily, weekly or monthly data | Statuses, Risk categories, Show as, Group time by |
| Matches by risk category over time | Time range; daily, weekly or monthly data | Risk categories, Which matches, Show what auto mode would suppress, Show as, Group time by |
| Unresolved matches now | Current state | Channel, Risk category, Age band, Queues |
| Matches by channel over time | Time range; daily, weekly or monthly data | Channels, Group time by |
| Matches arriving against matches resolved | Time range; daily, weekly or monthly data | Group time by |
| New alerts over time | Time range; daily, weekly or monthly data | Show as, Group time by |
| Queue backlog now | Current state | Item types, Queues |
| Queue backlog over time | Time range; hourly, daily, weekly or monthly grouping | Item types, Break down by, Show as, Group time by, Queues |
| Outstanding by age | Current state | Age bands, Item types, Queues |
| Oldest outstanding item | Current state | Item types, Queues |
| Items worked over time | Time range; daily, weekly or monthly data | Item types, Break down by, Group time by |
| Risk categories in queue now | Current state | Risk categories, Item types, Break down by, Queues |
| Risk categories in queue over time | Time range; daily, weekly or monthly data | Risk categories, Show as, Group time by |
| Outstanding by assignee (Admins only) | **Currently unavailable:** cannot be added | Catalogue-defined Item types option is not currently usable |
| Matches reviewed by person (Admins only) | Time range; daily, weekly or monthly data | No saved setup options |
| Match outcomes over time | Time range; daily, weekly or monthly data | Measure, Show as, Group time by |
| Match outcomes, latest period | Time range; daily, weekly or monthly data | Period |
| Profiles with a True Match now | Time range; daily, weekly or monthly data | Outcome |
| Profiles with a True Match over time | Time range; daily, weekly or monthly data | Outcome, Group time by |
| Screening precision | Time range; daily, weekly or monthly data | Metric, Channels, Group time by |
| Review effort per True Match | Time range; daily, weekly or monthly data | Group time by |
| Automatic disposition review | Time range; daily, weekly or monthly data | Metrics, Group time by |
| Screens over time | Time range; daily, weekly or monthly data | Break down by, Show as, Group time by |
| Screening hit rate by feed | Time range; daily, weekly or monthly data | Channel, Group time by |
| Risk assessments by status over time | Time range; daily, weekly or monthly data | Group time by |
| ID verifications by outcome over time | Time range; daily, weekly or monthly data | Group time by |
| Prediction and confidence distribution | Time range; daily, weekly or monthly data | No saved setup options |

### Read the setup options precisely

* **Profiles by status now** offers Potential matches, In review, and Escalated. The historical profile-status chart supports a wider set of statuses. The current risk-category profile chart is limited to Sanctions, PEP, and adverse media.
* **Age bands** are Up to 1 day, 2 to 7 days, 8 to 30 days, and Over 30 days. They describe waiting time, not calendar dates.
* **Queue backlog over time** and **Items worked over time** can break down by Item types or Status groups. Queue selection is supported on the backlog chart, not the items-worked chart.
* **Risk categories in queue now** can break down by Risk categories or Item types, and supports Profiles or Potential matches as item types.
* **Screens over time** can break down by Channels, Subject types, Tag groups, or no breakdown.
* **Profiles by status now** offers Donut, Horizontal bars, or Tiles. Other charts offer their supported stacked shape or Lines; not every chart offers every presentation.
* **Match outcomes over time** can show counts or rates. **Match outcomes, latest period** selects the latest day, week, or month rather than aggregating the whole range.
* **Profiles with a True Match now** and its over-time counterpart can select either True Match or False Match. The default title does not prevent choosing the other outcome.
* **Screening precision** can show Screening precision or False discovery rate. These are measures of eligible human-reviewed decisions, not the share of all screens that produced a hit.
* **Automatic disposition review** offers Model precision, Model recall, Model F1, Agreement rate, Override rate, No call rate, Coverage, and Measurable share. Precision asks how often a positive model suggestion was correct; recall asks how many human-confirmed matches it identified; F1 balances precision and recall. Agreement and override describe consistency with human decisions, while no-call, coverage, and measurable-share figures explain which population could be evaluated. Read the chart's sample and population notices before drawing conclusions.
* **Queues** means all queues or selected queues in the active workspace, only on charts that offer it. A previously selected queue that no longer resolves stays visibly unavailable; remove that selection deliberately rather than assuming it has silently become all queues.

For **Matches by status over time**, use **Split by feed** to compare the risk feeds separately instead of hiding their differences in a combined chart.

For **Screening hit rate by feed**, the saved **Channel** is **Direct API**, **Onboarding**, or **Risk assessment**. **Trend** shows the hit rate through time, and **Prior window** inside it adds the same-length period just before the range for comparison. Feed rates are not additive, because a search can query or flag more than one feed.

For **Matches by risk category over time**, **Which matches** changes the population: **All** includes recorded decisions, including automated handling; **At review gate** excludes earlier suppression but includes automatic disposition; **Outstanding** focuses on what still needs a person. **Show what auto mode would suppress** is a hypothetical overlay, not proof that those matches were actually suppressed. Read the population subtitle when comparing this chart with its neighbors.

For **Prediction and confidence distribution**, the on-chart **Model output** selector offers **All outputs**, **Confirmed**, **Dismissed**, **No prediction**, and **Undetermined**. **Policy mode** compares hint and automatic modes where the widget has room; a compact view shows one mode at a time. These are local viewing controls, not saved widget configuration. Read **Details** for the underlying cells, including values that are not valid confidence scores.

**Trend** and **Prior window** on **Screening hit rate by feed**, and **Model output** and **Policy mode** on **Prediction and confidence distribution**, are viewing controls on the card. They change what you see, are not saved to the widget, and do not change what other readers see.

**True Match** means a reviewer accepted the match as real; **False Match** means they rejected it as not real. **No Material Change**, **Not found**, and **Closed** are other recorded outcomes, not acceptance or rejection decisions.

### What links through

Two charts open the matching list of work when you select a value, and the status figures on **Profile headline** do the same:

* **Profiles by status now** links each status to the profiles in that status.
* **Profiles with risk by category now** links each risk category to the profiles carrying it.
* **Profile headline** links its status figures the same way as **Profiles by status now**.

Every other figure explains why it cannot open a list, rather than opening a list that holds something different. The explanations are:

* **Nothing in Profiles or Work Queues shows this set of items.**
* **Profiles and Work Queues cannot be limited to these dates, so no list holds exactly these items.**

## Widget states and data freshness

A widget shows one state at a time, and the card names the state rather than leaving the reader to guess.

**Widget States: Loading, No Data, and Error**

The figure shows three cards side by side: one still loading, one with no data in the selected period, and one whose source could not be read.

* While a widget is loading it shows a placeholder in the shape of its content
* A chart with nothing in the period reads **No data in this period**
* A chart whose source has never produced data in this workspace reads **No data in this workspace yet.** Some charts name the source that will fill them
* A chart that cannot be read reads **We could not load this widget.** and offers **Try again**
* A chart that is too small to draw reads **This chart needs more room.** and offers **Resize**, which grows it in one step
* A chart that is still collecting its first data reads **Collecting data. The first update runs at `<time>`.**
* A chart that records reviewer activity reads **This chart shows individual reviewer activity and is only visible to admins.** on an account that cannot see it

A chart can also carry up to two notice lines above its body, with any further notices behind **+N more**. Notices explain how to read the figure, for example **This chart counts matches entering review rather than matches arriving. Nothing records when a match first arrives.**, **A queue used by this chart is no longer available. Fix in settings.**, or **This chart is shown by day: hourly detail is only kept for 90 days.**

Each widget footer shows when its data was last refreshed, as **Updated 14:05 · Daily** in your local time. On a short card the same timestamp moves into the info icon beside the title.

## Practical dashboard recipes

> [accordion] Daily screening triage
>
> Add the profile headline, profiles by status now, unresolved matches, and
> outstanding by age. Put the headline at the top, then the widgets that
> explain where to act. Current-state charts do not need a historical range to
> show today's outstanding work. Use Overview or an available drill-through to
> move from monitoring to action.

> [accordion] Weekly workload review
>
> Add new alerts over time, matches arriving vs resolved, queue backlog over
> time, and items worked over time. Choose **Last 30 days** to give a week's
> changes context. Use compatible grouping and item filters, and keep arrival
> counts separate from the outstanding backlog.

> [accordion] Monitoring coverage review
>
> Compare profiles added over time with monitored profiles over time and
> profile status levels. Use a longer dashboard window, then override a single
> widget to **Last 7 days** to inspect a recent change. Reset it with
> **Dashboard default** before presenting a like-for-like comparison.

> [accordion] Restricted team activity review
>
> An authorized owner or admin creates an **Admins only** team dashboard, then
> adds **Matches reviewed by person**. Pair it with team-wide backlog and
> outcome charts. **Outstanding by assignee** is catalogue-defined but cannot
> currently be added. Keep the dashboard restricted and avoid treating
> individual counts as a standalone productivity ranking.

> [accordion] Review quality and model suggestions
>
> Combine match outcomes, screening precision, review effort per True Match,
> automatic disposition review, and prediction and confidence distribution.
> Check their populations before comparing them: screening hits, reviewed
> outcomes, model suggestions, and confidence bands answer different
> questions.

## Troubleshooting

**A dashboard will not let me change anything.**
It is a team dashboard you do not administer. Use **Add widget to my copy** or **Customize a copy** to work from a personal copy.

**A chart I want is disabled in the picker.**
It records individual reviewer activity, so it needs a dashboard whose audience is **Admins only**. A personal dashboard cannot hold it, even when you are an administrator.

**A chart in the picker cannot be added anywhere.**
**Outstanding by assignee** has no data behind it yet. On an **Admins only** dashboard the picker reports it as **Not available yet. This chart needs a change to how profile counts are scoped to a workspace.**, and it stays unavailable until that change ships.

**Changing the time range does not change a figure.**
The chart reports the current state, or the widget has its own **Time range** set. Choose **Dashboard default** on the card to put it back in step with the header.

**Counts do not add up.**
Compare what each chart counts, and its filters, range, and grouping. Profiles, potential matches, queue items, and screens are different units, and a profile or match can count in more than one risk category or feed.

**My teammate cannot see my layout changes.**
Select **Save** in the editor. **Save** stays unavailable until the layout changes, and leaving the editor discards unsaved changes unless you save them.

**I see a message that the dashboard changed.**
Someone else saved the same dashboard first. Select **Review latest** to load their version, then apply your changes again.

**A chart says the first update is still to come.**
New workspaces collect data on a schedule. The card names the first update time, and the chart fills in from then on.

**A queue used by a chart is gone.**
The queue was removed or renamed in [Work Queues](https://docs.gominerva.com/work-queues-guide.md). Select **Customize** to open the editor, then open the widget's menu, select **Configure**, and choose an available queue. The row that lost its queue reads **Queue no longer available**.

**Team actions are temporarily unavailable.**
Minerva could not read your team role. Try again in a moment, and contact [Minerva Support](mailto:support@gominerva.com) if it persists.

## Related guides

* [Dashboards Guide](https://docs.gominerva.com/dashboards-guide.md)
* [Work Queues Guide](https://docs.gominerva.com/work-queues-guide.md)
* [Screening Guide](https://docs.gominerva.com/screening-guide.md)
* [Profiles](https://docs.gominerva.com/minerva-profiles.md)
* [Exports Guide](https://docs.gominerva.com/exports-guide.md)
* [Match Scoring Guide](https://docs.gominerva.com/match-scoring-guide.md)
* [Risk Assessment Flow](https://docs.gominerva.com/risk-assessment-flow.md)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.