> 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/audit-and-fix-your-okrs.md).

# Audit and fix your OKRs

Review an existing OKR plan in Tability, find the goals that won't give you useful updates, and fix them.

To start, share this page with your AI assistant and say:

> Run this playbook

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

### Instructions for AI assistants <a href="#instructions-for-ai-assistants" id="instructions-for-ai-assistants"></a>

If the user shares this page (or an article linking to it) 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 <a href="#ground-rules" id="ground-rules"></a>

* Ask one short question at a time. Reuse anything the user has already told you.
* Confirm before touching a Tability workspace. Only use a workspace the user gives you by URL. Show its name and ID and get a yes before reading from it or writing to it.
* Read before you judge. Base every finding on what's in the plan and its check-ins, and quote the goal you're flagging.
* Don't make up numbers. If a baseline or target is unknown, say so. Label figures as measured, estimated, or proposed.
* Change nothing without approval. Propose every fix first. Apply only the ones the user approves, one batch at a time.
* Don't reassign people silently. Ownership changes are proposals until the user confirms them. Don't message anyone else.
* Never delete goals. If a goal should go, propose archiving or moving it, and let the user do it.
* Verify what you write. Read records back before saying they changed.
* If something is blocked, keep the work done so far, tell the user the exact next step, and carry on with what you can.

#### 1. Choose the plan <a href="#id-1-choose-the-plan" id="id-1-choose-the-plan"></a>

Start with a brief intro and ask:

> Which Tability plan should I review? Share the workspace URL and the plan's name or link.

If they share only a workspace, list its current plans and ask which one. Confirm the workspace and plan, then ask only what you still need, such as:

* Whether to audit the whole plan or specific objectives.
* Who the plan is for (company, department, or team), so owners can be checked against it.
* Any goals that are off limits for changes.

#### 2. Read the plan <a href="#id-2-read-the-plan" id="id-2-read-the-plan"></a>

Read the full plan content: every objective, key result, and initiative, with its owner, baseline, target, start and due dates, progress type, and any dependencies or alignments.

Also read:

* Recent check-ins for each key result, noting when the last one was and whether it included a comment.
* Workspace members and their reporting lines, so you can tell senior leaders from the people doing the work.

If the plan is large, read objective by objective and keep a running list. Don't judge anything yet.

#### 3. Run the checks <a href="#id-3-run-the-checks" id="id-3-run-the-checks"></a>

Run every check below on every goal in scope. For each issue, record the goal, the check it failed, the evidence, and a proposed fix.

**Key results are metrics, not tasks**

Flag a key result if it starts with a verb such as launch, create, implement, run, or complete, or if it fails any of these:

* It includes a number or measurable change.
* It could go up or down depending on how well the team executes.
* Hitting it would show the team actually succeeded.
* It can be updated weekly or fortnightly, not only at the end of the cycle.

To fix it, ask "so that what?" of the task, and rewrite the key result as the outcome that answers it, with a baseline (or "unknown"), target, and unit. Move the original task under it as an initiative. A reasonable proxy is fine. Ask the user when you can't tell what the task is meant to change.

**Milestone key results are set up to track themselves**

If a key result is genuinely about finishing a body of work and the user wants to keep it that way:

* Check that its milestones exist as initiatives under it.
* Check that its progress type is Milestone (total count or percentage), with auto-sync on if the user wants it.
* If the progress type can't be changed through the API, give the user the steps: open the key result, then ... > Edit, select Milestone, choose the milestone type, and save. Milestone tracking needs a Premium subscription. Without it, the key result has to be updated manually.

**Each goal has the right owner**

Flag any key result or initiative that:

* Has no owner.
* Is owned by a senior leader who owns several goals, or whose reports are doing the work.
* Has an owner who left the workspace or moved teams.
* Has stale check-ins (none in the last two weeks) or check-ins with no comment.

The right owner has first-hand visibility of the work or metric and could write this week's check-in without asking anyone else. For a cross-functional goal, prefer the person who touches the metric most, then the person coordinating the work, then the person most consistent about updates. Propose the senior leader and other people involved as contributors. You can't see who is closest to the work, so present each owner change as a question with your suggestion.

**Due dates match the plan**

Flag key results with a due date different from the plan's timeline. If more than one or two differ, recommend reassessing those key results rather than keeping the dates.

Recommend a single due date for the whole plan by default, with an early finish treated as beating the target. If the user confirms a key result truly must finish early, propose this setup:

1. A second plan whose timeline ends on that date.
2. The same objective as in the main plan.
3. The due date in the key result's title (for example, "Hire 15 new SDRs by February 15").
4. A relationship linking that key result to the objective in the main plan as its parent.

**Prioritize**

Rank the findings by how badly they block useful updates: missing owners and task key results first, then wrong owners and stale check-ins, then due dates and milestone setup.

#### 4. Share the findings <a href="#id-4-share-the-findings" id="id-4-share-the-findings"></a>

Save the audit as {plan-name}-okr-audit.md, marked Draft, using the outline at the end of this page. Share it with a short summary: how many goals you checked, how many passed, and the three most important fixes.

Then walk through the proposed fixes in priority order, one group at a time (rewrites, owner changes, due dates, milestone setup). For each group, ask the user to approve, edit, or skip each fix. Ask the open questions you couldn't answer from the data, such as who is closest to the work or what a task is meant to change.

#### 5. Apply approved fixes <a href="#id-5-apply-approved-fixes" id="id-5-apply-approved-fixes"></a>

Apply only the approved fixes, a group at a time:

* Rewritten key results: update the title, baseline, target, and unit. Create the original task as an initiative under it.
* Owner changes: set the new owner from real workspace members. List the contributors to add, and give the user the steps if contributors can't be set through the API.
* Due-date changes: update the dates, or create the second plan and relationship described in step 3.
* Milestone setup: create missing milestone initiatives, and give the user the steps for anything you can't change.

Keep existing check-in history. Read each changed record back, and mark each fix in the audit file as applied, skipped, or left for the user.

#### 6. Wrap up <a href="#id-6-wrap-up" id="id-6-wrap-up"></a>

Mark the audit file Final and share it with a short summary: what changed, with links to the changed goals in Tability; what the user still needs to do by hand; which owners should be told about their new goals; and which goals to watch at the next check-in. Offer a short prompt the user can run next cycle to repeat the audit.

#### Audit file outline <a href="#audit-file-outline" id="audit-file-outline"></a>

```markdown
# {Plan}: OKR audit

Status · Date · Workspace · Plan link · Scope

## Summary

## Findings by priority

| Goal | Check failed | Evidence | Proposed fix | Status (proposed, applied, skipped, user to do) |

## Key result rewrites (before and after)

## Owner changes and contributors

## Due dates and milestone setup

## Steps left for the user

## Open questions

## Prompt to repeat this audit
```

END OF PLAYBOOK
