> For the complete documentation index, see [llms.txt](https://guides.tability.io/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://guides.tability.io/docs/playbooks/build-an-exec-dashboard.md).

# Build a scorecard

Turn the measures that matter to your team into a reusable scorecard in Tability, with clear targets, owners, and a review rhythm.

Build a scorecard your team can use every week or month to compare results with targets, spot changes, and decide what needs attention. In Tability, your scorecard is a KPI View: an ordered collection of KPIs with actuals and targets over time.

To start, share this page with Claude, ChatGPT, Grok, or another AI assistant with Tability connected through MCP, and say:

> Run this playbook

This playbook runs in your connected assistant, outside Tability AI Mode. It creates or updates a KPI View and a companion file with metric definitions and review notes. KPIs and KPI Views require Premium, and creating or updating records requires the appropriate edit access. Other connected tools or files can supply the underlying business data.

The instructions below are for your assistant. You don't need to read or copy them.

***

### Instructions for AI assistants

If the user shares this page and asks you to run it, follow the steps below from start to finish. If they only ask you to read, summarize, or edit it, do that instead.

#### Ground rules

* **Ask one short question at a time.** Reuse known context. Produce a first scorecard proposal within five discovery questions, with assumptions and gaps clearly marked.
* **Confirm the business, team, and audience.** Don't infer them from the publisher or connected accounts. Resolve the user-supplied Tability workspace URL to its name and ID, and confirm scope unless already authorized.
* **Choose measures for decisions.** Each KPI must help the audience assess a result or decide what to do. Keep the scorecard small enough to review regularly.
* **Use real evidence.** Include periods, source links, and freshness. Don't invent benchmarks, targets, causes, or missing values.
* **Check the available MCP actions and access early.** Confirm KPI availability and edit access before promising setup. Use the connected KPI and KPI View tools. If setup is unavailable, deliver the scorecard proposal and the exact manual next steps.
* **Show the scorecard before writing.** Get approval for the selected KPIs, their order, reporting cadence, View access, and any target or ownership changes. Reuse that approval for implementation and verify the result. Don't send reports or change access without a specific request.

### 1. Understand the review questions

Start with:

> Who will use this scorecard, and what decisions should it help them make?

Then establish the business model, the part of the organization the user belongs to, the reporting scope, and the review cadence. Ask for an existing scorecard or team report if available.

Turn broad requests such as “visibility” into concrete questions: Are we likely to hit the quarter's commitments? Where should we put more resources? Is growth creating a service problem? What needs intervention this week?

Separate confirmed priorities from assumptions. If the user isn't sure which questions matter, suggest a short list based on the context and ask them to choose. Don't contact other people yourself.

### 2. Choose what the scorecard should cover

Offer 3–5 categories that combine the business model with the team's responsibilities. Treat these as a menu, not a mandatory template or a source of benchmark targets.

| Business context            | Categories to consider                                                                      |
| --------------------------- | ------------------------------------------------------------------------------------------- |
| Subscription business       | Recurring revenue, retention, acquisition efficiency, product adoption, service reliability |
| Ecommerce or retail         | Sales and margin, conversion, repeat purchasing, inventory, fulfillment                     |
| Professional services       | Bookings, project margin, capacity, delivery, client retention                              |
| Marketplace                 | Demand and supply, successful matches, repeat usage, unit economics, trust and quality      |
| Manufacturing or operations | Throughput, quality, delivery reliability, cost, safety                                     |
| Nonprofit or public service | Beneficiary outcomes, reach and access, service quality, funding, resource use              |

Then narrow the categories to the user's part of the organization:

* **Sales and marketing:** demand, conversion, revenue contribution, and acquisition efficiency.
* **Product and engineering:** adoption, customer outcomes, delivery reliability, and product quality.
* **Customer success and support:** retention, customer health, resolution, and service quality.
* **Operations:** capacity, throughput, quality, cost, and delivery commitments.
* **People and finance:** workforce capacity and health, financial performance, cash, and forecast risk, as relevant to the team.

Explain why each chosen category belongs. Include only results the audience needs and the team owns or materially influences. If the team has no useful KPI definitions yet, use [Find and track the right KPIs](https://guides.tability.io/docs/playbooks/find-and-track-the-right-kpis.md) before creating new measures.

### 3. Gather and check the evidence

Start with existing Tability KPIs, their definitions, dated values and targets, owners, and KPI Views. Reuse an existing View when its purpose and audience match. Read relevant plans, outcomes, check-ins, and initiatives where they help explain the results. Use connected finance, CRM, analytics, support, or project sources where they answer a selected question.

For every candidate measure, record its formula, unit, business scope, reporting period, comparison period, target, source, owner, and freshness. Resolve conflicting definitions before combining figures. Follow pagination and disclose any incomplete coverage.

Keep KPI performance separate from goal progress, check-in confidence, and data freshness. A goal marked on track is not proof that its business metric met target. A metric without recent data must not appear healthy by default.

### 4. Propose the scorecard

Choose the smallest useful set of KPIs, with 6–10 headline measures at most. Explain the decision each measure supports and why it belongs. Use the categories to choose and order the KPIs; the saved View is an ordered list of KPI rows.

Show a draft table with each KPI's name, existing Tability link or proposed new record, owner, actual and observation date, agreed target or “no target,” and prior comparable result. Label missing or provisional data. Keep definitions, source links, calculated comparisons, and review notes in the companion file; these are not extra configurable columns in the KPI View.

Agree the View name, KPI order, display mode, and date range. Choose a cadence that matches the observations and decisions: daily for frequent operational reviews, weekly for team reviews, or monthly for business reviews. Use a supported range: day — 1 or 3 months; week — 1, 3, or 12 months; month — 1, 3, 12, or 24 months. Record the workspace timezone. The display cadence does not schedule data updates.

Include the intended audience and access in the proposal. Views created through the current MCP tools default to “Everyone can edit”; read-only workspace members remain read-only, and underlying KPI permissions control which rows each person can see. Sharing administration is UI-only. If restricted View access is needed, have the user create the View with the required access in Tability first, then reuse it. Preserve access settings on existing Views.

Save `{team-name}-scorecard-review.md`, marked Draft, with the proposed rows, definitions, sources, View settings, access, and any record changes. Ask the user to approve the concrete setup.

### 5. Build the scorecard in Tability

After approval, search again for the selected KPIs and View to avoid duplicates. Reuse approved records and create only missing KPIs. Follow [Find and track the right KPIs](https://guides.tability.io/docs/playbooks/find-and-track-the-right-kpis.md) for new definitions, dated observations, approved targets, and owners. Keep a KPI without an agreed target targetless; never turn a measured baseline into a target.

Create or update the approved KPI View, add the selected KPIs, and put them in the agreed order. A KPI can belong to several Views without duplicating its values or targets. When updating an existing View, retain other KPIs unless their removal was explicitly included in the approved setup. Removing a KPI from a View does not delete the KPI.

If the connected tools cannot complete setup, provide the manual steps: KPIs → Views → Create view → choose name, display mode, range, and access → Create → Add KPI. For an existing View, use More → Edit settings and reorder the KPI rows as agreed. See [KPI Views](https://guides.tability.io/docs/become-a-tability-power-user/features/kpis/kpi-views.md) for the product workflow. Mark any unfinished setup clearly.

A KPI View displays the latest recorded actual in each period and the active target. It does not sum or average observations. Confirm that each recorded value represents the intended measure before calling it a weekly or monthly result. Missing observations remain missing, and targetless KPIs have neutral actuals. Edit values and targets on the underlying KPI through the supported tools or its KPI panel.

Read back the View's settings, KPI membership, order, and links. Check the periods returned by the data tools rather than assuming they match the saved display settings. If the required periods are unavailable through the View tools, use dated KPI history for the review and verify the display in Tability where possible. Add the View and KPI IDs and links to the companion file.

### 6. Run the first scorecard review

Use the scorecard to answer the original review questions. Identify the measures that met or missed their agreed targets, meaningful changes from a comparable period, and missing or stale observations. Trace at least one result and comparison back to the underlying source. Keep target attainment separate from freshness; an old value that met target is still stale.

Write a short review in the companion file: what changed, which results need attention, and the proposed actions or decisions with owners and dates. Distinguish documented explanations from hypotheses. Keep these notes alongside the scorecard; the KPI View contains the metric timeline.

Verify audience access where possible: access to a View does not guarantee access to every KPI. Report any missing rows or access limitations without widening permissions automatically.

Agree who updates each KPI and who runs the review. Reuse native Tability data connectors where suitable; connector setup is currently UI-only. Use the update-method guidance in [Find and track the right KPIs](https://guides.tability.io/docs/playbooks/find-and-track-the-right-kpis.md) for manual entry or external assistant jobs, and avoid competing writers for the same KPI.

Only create a recurring assistant review when requested and supported. Include the View and KPI IDs, sources, formulas, timezone, cadence, freshness rules, and permitted write scope; verify its next run. A review job should read the scorecard and prepare notes unless data updates were separately approved. Otherwise, provide this prompt with the actual links filled in:

> Review \[KPI View] for \[period] using \[definitions and sources]. Compare actuals with agreed targets and the prior comparable period. Flag stale or missing data, explain meaningful changes using evidence, and propose actions with owners and dates. Save the review notes in \[companion file]. Propose any KPI or target changes before writing them.

### 7. Wrap up

Share the KPI View link, companion file, questions the scorecard answers, first review findings, and update and review owners. Clearly list any manual setup or inaccessible data. After the first few reviews, suggest removing measures that do not help the audience make decisions, without changing the View automatically.

***

### Scorecard review file outline

```markdown
# {Team}: scorecard review
Status · Business · Audience · Workspace · Review cadence · Timezone

## Review questions and decisions
## Selected categories and KPI order
## Metric definitions, sources, targets, and owners
## View settings and audience access
## Actuals, comparable periods, and freshness
## Review findings, proposed actions, and decisions
## Tability links and verification
## Update methods, review prompt, and remaining setup
```

**END OF PLAYBOOK**
