Capability checklist

What Outcome does, item by item.

The capabilities buyers ask about, each marked with where Outcome stands today.

87 capabilities are live, 46 are in preview, and 18 more are answered by a deliberate choice.

  • LiveIn the product today. 87
  • PreviewWorking, refined with customers. 46
  • By designA deliberate choice, with what Outcome does in its place. 18
Goals and OKRs16 in the product
  • Goal tree with owners and targetsLive

    Every goal sits in one tree with one accountable owner, a target, a date and a current figure.

  • Objectives and key resultsLive

    A goal can be judged by its key results, by its own number, or by both.

  • Progress figures read from connected systemsLive

    Current values come from the tools you connect, each with its source and date, and refresh on a schedule.

  • Check-ins with a status and commentLive

    Where no system holds the figure, the goal owner is reminded, files a dated entry, and adds their own view and a comment.

  • Back-dated entriesLive

    An entry can be filed for a past date. Future dates are refused.

  • Computed status: on track, at risk, off trackLive

    Status is computed from the figures and judged against quarterly targets where set. The owner's own view sits beside it.

  • Quarterly targets and a scorecardLive

    Goals carry quarterly targets, and a scorecard shows target against actual for each period.

  • Planning periods and fiscal calendarLive

    Goals are planned and reported in your own periods, while deadlines keep their real dates.

  • Tree, table and map viewsLive

    Goals can be viewed as a tree, a table with one row per goal, or a map with status on each card.

  • Your own names for goal levelsLive

    Each company names its own goal levels and the words used for alignment.

  • Health trend over the cycleLive

    The goals dashboard shows how the split between on track, at risk and off track has moved over time.

  • Meeting view with one goal per screenLive

    A walk view shows one goal per screen, with previous and next, for review meetings.

  • Starter templatesPreview

    Starter goal trees exist for delivery, support and sales.

  • Quick-create a goal, including from JiraPreview

    Goals are created quickly from one box in Outcome.

  • Restoring a goal to an earlier statePreview

    Earlier targets and wording stay on the record. Bringing one back means entering it again as a new recorded change.

  • Retrospectives when a cycle closesPreview

    A retrospective can be drafted in your own AI assistant from checked figures.

  • Typing a progress number over the current valueBy design

    Figures come from the connected system, or from a dated entry by a named person where no system holds the figure.

  • Weighted average roll-up of child goalsBy design

    Each child goal takes a share of the parent's gap in the parent's unit, and the uncovered part stays visible.

  • Rolling goals or plans into the next periodBy design

    When a date passes, the goal owner records whether it landed, stops it, or sets a new date with a reason.

  • UndoBy design

    A change is reversed by making a new recorded change, so the history stays complete.

  • A goal under more than one parentBy design

    A goal has one parent, and other alignments are shown as views that take no share of progress.

Portfolio and planning13 in the product
Customer feedback10 in the product
  • Feedback from customer callsLive

    Customer calls are read from Attio into masked questions and replies, placed on the product area they concern.

  • Product usage beside feedbackLive

    Product usage is read from PostHog. Showing changes in usage beside what customers said is in preview.

  • Survey answersLive

    Hotjar and PostHog survey answers are read.

  • Routing feedback to an ownerLive

    Rules send feedback from each source to the person accountable for that product area.

  • A named decision on each topicLive

    Each topic ends with a named person's decision: make it a goal, attach it to one, or decline with a reason.

  • Feedback sent to delivery with quotesLive

    A topic becomes a goal whose work goes to Jira, GitHub or a coding agent with the customer's words attached.

  • Telling customers when it shipsLive

    Outcome lists the customers owed news, drafts the sentence, and records when a person says they told them.

  • Support desk connectionPreview

    Zendesk tickets arrive as feedback, and CSAT, ticket volume and reply times as numbers for a goal.

  • Application errors as feedbackPreview

    Errors and user feedback from Sentry join the same queue and the same owners as other feedback.

  • Automatic themes and sentimentPreview

    Feedback is collected by product area with a computed urgency, which is live. Grouping similar items into topics is in preview.

  • Resolving individual support casesBy design

    Outcome works on topics across customers, and single cases stay in your support tool.

Engineering and AI adoption12 in the product
  • Pull request cycle time and throughputLive

    Merged pull requests, cycle time, size and reverts are read from GitHub as weekly measures.

  • Deployment frequency, lead time, change failure rateLive

    Deployment frequency and lead time are read from GitHub. Change failure rate appears where GitHub reports failed deployments.

  • AI tool usage and spend by teamLive

    Claude Code and Cursor usage is read daily by team. Copilot, Claude Enterprise and OpenAI usage and spend are read from each provider.

  • Effect of AI on deliveryLive

    Agent and human pull requests are compared on cycle time. Checking work against the measure it was meant to move is in preview.

  • Change work against run workLive

    The ratio of change work to run and protect work is shown, with unclassified work listed separately.

  • Agent share of merged pull requestsLive

    GitHub reports agent-authored merges beside all merges, so the agent share is a measure a goal can be judged by.

  • Time to restore servicePreview

    Time to restore is read by service from PagerDuty or ServiceNow, with time to acknowledge from PagerDuty. Both connectors are in preview.

  • Seats held against seats in usePreview

    Seats held and weekly active users are read from each provider and shown by team.

  • AI use outside engineeringPreview

    Claude Enterprise usage and Microsoft 365 Copilot licence counts are read. Gemini and ChatGPT Enterprise are not.

  • Production errors and reliabilityPreview

    Errors from Sentry are grouped into themes with an owner and a brief for a fix.

  • AI adoption rate over headcountPreview

    A ratio measure can divide active AI users by headcount from your HR system. Seat and active-user figures are in preview.

  • Telemetry pushed from any agent (OTLP)Preview

    Pushed telemetry is accepted from Claude Code and Cursor, and from Kiro in preview.

  • AI usage per personBy design

    Usage is shown by team with a minimum of five people, and Outcome does not rank individuals.

  • Hours-saved estimatesBy design

    Outcome reports what the business measure did, taken from its source.

  • AI code reviewBy design

    Outcome does not review pull requests for bugs or security.

  • Composite productivity frameworkBy design

    Each figure stays its own measure, judged against a goal, and there is no composite index.

  • AI maturity scoreBy design

    Outcome shows the goal, the work that supports it, and whether one moved the other.

Cost and spend13 in the product
  • AI spend by tool, model and teamLive

    One page shows each AI tool's spend by week, tool, model family and team, against the period before.

  • AI spend traced to work and goalsLive

    AI sessions are tied to their repository and branch, so spend is placed on the work and goal it served.

  • Cost shown on each goalLive

    Each goal and pull request shows an estimate of the people's time and the AI spend behind it.

  • People cost without timesheetsLive

    People cost is estimated from the work people did, using rates you set.

  • AI budget and month projectionLive

    Set a monthly AI budget and see where the current month is projected to land. There are no per-tool budgets.

  • AI cost per pull requestLive

    A ratio measure gives AI spend per pull request, and each piece of work shows the AI cost behind it.

  • Spend that reached no goalLive

    The share of AI spend that reached no goal is shown as its own amount, so it can be questioned.

  • Cost per unit of progressPreview

    Each goal shows what a unit of movement on its measure has cost, beside comparable goals.

  • Spend alertsPreview

    Sudden rises in spend and budget overruns are sent to the owner who can act, in Slack or by email.

  • Idle seatsPreview

    Team owners are told which AI seats are idle.

  • Technology spend by towerPreview

    Your technology ledger is imported as a file and shown by tower and value stream.

  • Cloud billing connectionPreview

    AWS and Azure cost is read from their FOCUS cost exports into the ledger, by service, account and tag. Google Cloud cost is read through BigQuery.

  • Cost per pull request by teamPreview

    Each room shows its spend and its pull requests.

  • Credits converted into currencyBy design

    Credits and currency stay as separate measures shown side by side, without an assumed exchange rate.

  • Spend control such as model routing or throttlingBy design

    Outcome reports whether spend was worth it, and does not route models or limit tokens.

Decisions and governance10 in the product
  • Decision registerLive

    Each decision records who made it, why, and what it changed, in a register anyone in the company can browse.

  • Change history on every goalLive

    A change to a goal's target, date, owner or parent is recorded with who made it and their reason.

  • Questions routed to one ownerLive

    Each question has one owner, and regulated decisions can require two signatures.

  • Earlier decisions shown before you decideLive

    Before deciding, you see earlier answers to similar requests, their reasoning, and the effect your decision will have.

  • Decisions reviewed laterLive

    Past decisions come back for review when they fall due.

  • Result compared with what work promisedLive

    After work ships, Outcome shows whether the measure moved as much as the work was expected to deliver.

  • AI suggestions a person acceptsLive

    Agents propose goals, links and work from your own data, and each proposal takes effect only when a person accepts it.

  • Putting an item off for a set timeLive

    An item can be put off for a week, a month or a quarter, and it then returns.

  • One inbox of proposals and findingsLive

    Each person has one inbox of proposals and findings to accept, change or decline in place.

  • Early warnings that raise a decisionPreview

    Customer feedback raises a case with its owner. Changes in product usage and jumps in spend do the same in preview.

Assistants and agents15 in the product
  • Assistant inside the productLive

    Ask a question in plain words and get a short answer with its sources. Colleagues can join the same conversation.

  • Assistant in SlackLive

    Ask @Outcome in a Slack thread where a goal stands, and get an answer from current figures with links.

  • Connector for your own AI assistantLive

    Connect Claude, and other assistants that support the same connector standard, with access limited to what each person approved.

  • Answers with sourcesLive

    Answers come from Outcome's record of goals, work and feedback, with links.

  • Ready-made agentsLive

    A set of named agents comes with the product.

  • Handing work to a coding agentLive

    Hand a piece of work to Cursor or Claude from the goal it serves, with a named approver. Devin is in preview.

  • Multi-step research reportsLive

    A researcher agent and evidence packs are available.

  • Agent asks a question mid-taskLive

    An agent that needs an answer appears under Needs you, and you answer it in place.

  • Agent activity visible step by stepLive

    Each agent run shows its status, steps and cost.

  • Goals drafted from a documentLive

    Paste or upload a plan, deck or spreadsheet and Outcome proposes goals, each citing its line, for a person to accept.

  • Work knowledge graphPreview

    People, teams, goals and work form one graph behind each answer.

  • Network firewall for agentsPreview

    Connector requests pass through one guarded route and previews build in a sandbox.

  • Admin governance of agentsPreview

    Each AI assistant gets only the access a person granted.

  • Agents that run on a schedule or triggerPreview

    A team can switch on agents that run on new work or feedback under a named signer.

  • Scheduled leadership reportsPreview

    A weekly Slack digest is live and a goal summary can be sent on demand.

  • Meeting notes and transcriptionBy design

    Recorded customer calls are read for feedback, and meetings are not captured or summarised.

Teams and workforce9 in the product
  • Teams and org chart from your HR systemLive

    Teams, reporting lines and photographs are read from BambooHR or TeamForm. Microsoft Entra ID, Okta and Workday are in preview.

  • A workspace for each teamLive

    Each team can have a room showing its own goals, work and feedback, tied to the company goals above.

  • What a team ownsLive

    A room names the repositories and Jira projects its team owns, with one accountable person.

  • A page for each personLive

    A person's page shows what they own and what their work serves.

  • Headcount as a measureLive

    BambooHR and TeamForm publish headcount as a measure a goal can be judged by.

  • Allocations by teamLive

    Dated allocations of people to teams are read from TeamForm, and leavers stop counting.

  • Several views of the organisationLive

    The company can be viewed by reporting line, team, product and value stream, each area with one accountable owner.

  • One workforce baseline across HR and financePreview

    HR, allocation and cost ledger data are read in.

  • Ownership suggestionsPreview

    Agents suggest which goal a piece of work serves and which person a login belongs to, but not a component's owning team.

  • Writing changes back to the HR systemBy design

    Outcome reads your HR system and does not write to it.

  • Compensation planningBy design

    Outcome does not plan compensation and does not hold an individual's pay.

Security and administration11 in the product
Integrations and data24 in the product
  • JiraLive

    Jira work is read and linked to goals.

  • GitHubLive

    Pull requests, checks and deployments are read from GitHub.

  • SlackLive

    Slack is installed in one step for questions, reminders, digests and notices.

  • Spreadsheets as a sourceLive

    A goal can take its figure from a Google Sheet you already keep, with forecast rows kept apart from results.

  • CRMLive

    Pipeline, bookings and cases are read from Salesforce (in preview); calls and pipeline from Attio.

  • Product analyticsLive

    Product usage is read from PostHog and can judge a goal.

  • HR systemLive

    Teams, reporting lines and headcount are read from BambooHR or TeamForm. Microsoft Entra ID, Okta, Workday and SAP SuccessFactors are in preview.

  • Any system with an HTTP APILive

    A custom source feeds a goal from any HTTP API without code. Requesting a new integration is in preview.

  • Read APILive

    A documented read-only API, with keys an admin issues, lets other systems show current goals.

  • CSV exportLive

    The goal tree exports as CSV.

  • PDF exportLive

    Your connected assistant writes a report or board paper from Outcome's figures and saves it as a PDF.

  • Slides or image exportLive

    Your connected assistant builds slides or a page from Outcome's figures. The goal tree also exports as CSV.

  • Azure DevOpsPreview

    Boards work items, pull requests and pipeline deployments are read, with weekly delivery numbers; a decided gap can be filed as a backlog item.

  • LinearPreview

    Issues, projects and initiatives are read, customer requests arrive as signals, and a decided gap can be filed as an issue.

  • GitLabPreview

    Issues, epics, merge requests and deployments are read, with weekly delivery numbers and DORA on Ultimate; a decided gap can be filed as an issue.

  • Power BI as a sourcePreview

    A goal can take its figure from a measure in a Power BI semantic model, read each day.

  • Tableau as a sourcePreview

    A goal can take its figure from a Tableau data source field or Pulse metric, read each day.

  • Data warehousePreview

    A goal can take its figure from a saved SQL query in Snowflake, Databricks or BigQuery, read on a schedule you set. It is in preview.

  • ServiceNowPreview

    Services, work and incidents are read from ServiceNow, with restore time and change failure by service, and a decided gap can be raised there.

  • PagerDutyPreview

    Incidents, services and teams are read from PagerDuty, with time to acknowledge and time to restore by service. Never schedules or responders.

  • WorkdayPreview

    Workers, managers and nested supervisory organisations are read from Workday reports, with headcount and open positions by organisation. It is in preview.

  • Sentry, Vanta and KiroPreview

    Connections to Sentry for errors, Vanta for compliance progress and Kiro for usage are in preview.

  • Goals shown in Jira and ConfluencePreview

    An Atlassian app that shows goal cards in Jira and Confluence is in preview.

  • BI export such as Power BIPreview

    A BI tool can read goals through the read-only API, in preview, or from CSV.

If a requirement is not on this list, ask us.