FS-DOC-03Operations
Process Runbook / SOP
Team Performance Dashboard
[YourCompany.com] · Management Department · Prepared by FullSpec · [Today's Date]
This runbook is your day-to-day operating guide for the automated Team Performance Dashboard process. It tells you exactly what the automation does each week, what your one human task is, how to handle anything that goes wrong, and who to contact for support. FullSpec built and configured everything on the automation side. Your team's responsibility is the single review-and-approve step before the report goes out, and the routine maintenance checks described at the end of this document.
01Process overview
Every Monday at 7:00 AM the automation wakes up and does the reporting work that previously took one person three to four hours to complete manually. It calls the HubSpot and Jira APIs, writes all retrieved metrics into a Google Sheets staging area, sends Slack reminders to any team leads who still need to submit manual inputs, validates the data, refreshes the Looker Studio dashboard, and drafts a plain-English weekly summary in Notion. The manager then reviews that draft, makes any edits, and clicks approve. Once approved, the automation posts the dashboard link and summary to the management Slack channel. The report is live and distributed before the team sits down for stand-up, with no spreadsheet work required from anyone on your team.
Process name
Team Performance Dashboard
Trigger
Scheduled: every Monday at 7:00 AM (automatic, no action required)
Final output
Looker Studio dashboard link and AI-drafted weekly summary posted to the management Slack channel
Agents running
Data Collection Agent; Validation and Commentary Agent
Tools involved
HubSpot, Jira, Google Sheets, Google Looker Studio, Notion, Slack
Weekly volume
1 report per week, covering 6 to 12 metrics across 3 to 6 data sources (~4 reports/month)
Human checkpoint
Manager reviews and approves the AI-drafted Notion commentary before Slack distribution
Process owner
[Your name] — Operations Manager
Process Runbook / SOPPage 1 of 4
FS-DOC-03Operations
02Step-by-step: what happens and who acts
What you actually need to do: your only required action in this process is to open the Notion page each Monday morning, read the AI-drafted weekly summary, make any edits you want, and click Approve. Everything before and after that step is handled automatically by the automation.
Step
What happens
Who acts
Type
1
A scheduled trigger fires at 7:00 AM every Monday and starts the full data collection run. No one needs to initiate this.
Automation
Automated
2
The Data Collection Agent calls the HubSpot API and retrieves the current week's pipeline figures, deal count, and revenue metrics, then writes them into the Google Sheets staging area.
Data Collection Agent
Automated
3
A second API call pulls sprint velocity, open issue counts, and cycle time from Jira and appends those values to the same staging sheet alongside the HubSpot data.
Data Collection Agent
Automated
4
For any metric that requires a manual team-lead submission, the automation sends a targeted Slack message to the relevant person with a direct link to their input cell in Google Sheets.
Data Collection Agent
Automated
5
Once all mandatory fields are populated, or after a 90-minute wait period has elapsed, the Validation and Commentary Agent reads the staging data, checks for missing values, and flags any metric that has moved more than 25% week-on-week.
Validation and Commentary Agent
Automated
6
The Looker Studio dashboard is refreshed automatically so it reflects the newly written Google Sheets data. No manual chart updates or data range adjustments are needed.
Automation
Automated
7
The Validation and Commentary Agent drafts a plain-English weekly summary, noting notable metric movements and flagged anomalies, and writes it to the linked Notion performance page ready for the manager to read.
Validation and Commentary Agent
Automated
8
The manager opens the Notion page, reads the draft summary, makes any edits to wording or context, and clicks Approve. This is the only step that requires human action. Target time is around 20 minutes.
Operations Manager
Human
9
Once approval is confirmed, the automation posts the Looker Studio dashboard link and the approved Notion commentary to the management Slack channel, completing distribution for the week.
Automation
Automated
Process Runbook / SOPPage 2 of 4
FS-DOC-03Operations
03Handling exceptions
Situation
What the system does
What you do
A team lead does not submit their manual metric input by the 90-minute wait deadline
The Validation and Commentary Agent proceeds with the data available and marks the missing field as outstanding in the Google Sheets staging row. The draft Notion summary flags the gap explicitly so the manager is aware before approving.
Review the flagged field in Notion before approving. If the missing figure is material, message the team lead directly and hold approval until they submit. If it is not material, note the absence in the commentary and approve.
A metric from HubSpot or Jira moves more than 25% week-on-week
The Validation and Commentary Agent flags the movement as an anomaly in the staging sheet and includes a specific callout in the Notion draft summary, labelled as requiring manager attention.
Read the anomaly note in Notion, decide whether the movement is genuine or a data error, and either edit the commentary to add context or investigate the source data in HubSpot or Jira before approving.
A required metric is missing entirely because the API returned no data for that field
The agent writes a blank or null value to the staging row and marks it as missing. The Notion draft flags it alongside any anomalies. The Looker Studio panel for that metric will show no data rather than a stale figure.
Check whether the metric exists in HubSpot or Jira directly. If the field has moved or been renamed, contact FullSpec at support@gofullspec.com so the field mapping can be updated.
The same report appears to have run twice (duplicate staging rows for the same week)
The automation does not automatically deduplicate staging rows. The duplicate row will be visible in Google Sheets and may cause the Looker Studio dashboard to show inflated figures.
Open Google Sheets, identify the duplicate row by checking the date stamp column, and delete the incorrect entry. Refresh Looker Studio manually if needed, then contact FullSpec to investigate the cause of the duplicate trigger.
The manager does not approve or reject the Notion summary within a set window (no action taken)
The automation waits indefinitely for the approval signal. No Slack post will be sent until approval is confirmed. The process does not time out or self-distribute without the manager's action.
If you are unavailable, delegate the approval to a nominated backup reviewer (set one up in advance and confirm this with FullSpec during onboarding). Log in to Notion, review the draft, and either approve or add a note explaining the delay before approving.
A connected tool (HubSpot, Jira, Google Sheets, or Slack) is temporarily unavailable or returns an error during the run
The automation retries the failed step up to three times with a five-minute gap between each attempt. If all three retries fail, the run halts and an error notification is sent to the process owner's Slack account directly.
Check the Slack error notification for detail on which tool failed. If it is a temporary outage, the FullSpec team will monitor and re-trigger the run once the tool is back online. For persistent failures, contact support@gofullspec.com with the error message from Slack.
04Who to contact and when
Fill in the rows below once your team is confirmed. Keep this table updated whenever a role changes hands so everyone knows exactly who to call in an exception scenario.
Role
Name
How to reach them
Process owner (Operations Manager)
[Your name]
[Your email or Slack handle]
Backup approver (covers manager absence)
[Backup name]
[Backup email or Slack handle]
HubSpot admin (API access and field changes)
[HubSpot admin name]
[Admin email]
Jira admin (API access and project configuration)
[Jira admin name]
[Admin email]
FullSpec builder (build changes, field mapping updates, agent configuration)
FullSpec team
support@gofullspec.com
FullSpec support (errors, outages, unexpected behaviour)
FullSpec support
support@gofullspec.com
Process Runbook / SOPPage 3 of 4
FS-DOC-03Operations
05Ongoing maintenance
The automation runs without routine intervention, but a small number of maintenance tasks keep it accurate and reliable over time. The table below sets out when each task should happen and who is responsible.
Whenever a tracked metric changes name, moves to a different field, or is retired
Notify FullSpec at support@gofullspec.com with the old and new field details. Do not edit the Google Sheets staging schema directly, as this will break the agent's write logic.
Process owner, then FullSpec
Whenever the Notion commentary template needs a structural change (new section headings, different format)
Update the template in Notion and then confirm with FullSpec that the agent's output instructions match the new structure. Minor wording edits within an existing section can be made directly.
Process owner; FullSpec if structural
Monthly spot-check (first Monday of each month)
Open the Google Sheets staging area and verify that the last four weekly rows contain no unexpected blanks or repeated values. Spot-check two metric figures against their source tool to confirm accuracy.
Process owner
Monthly error-log review
FullSpec provides a brief monthly summary of any retry events, failed runs, or anomaly flags from the previous four weeks. Review this and raise questions with FullSpec if any pattern is unclear.
Process owner, supported by FullSpec
Whenever a team lead who submits manual inputs changes or their Slack handle changes
Update the team lead list held in the automation's configuration. Contact FullSpec with the old and new Slack user details so the targeted reminder messages reach the right person.
Process owner, then FullSpec
Quarterly volume and threshold review (every three months)
Review whether the 25% week-on-week anomaly threshold still fits the team's typical metric movements. If the business has grown or the reporting scope has changed, agree new thresholds with FullSpec and have them updated in the agent configuration.
Process owner, then FullSpec
The most common maintenance issue for this process is a metric field name changing in HubSpot or Jira after a tool update or internal restructure, without the automation being notified. When this happens the affected metric returns a blank or null value in the staging sheet and the agent flags it as missing data. If you notice a metric that was previously reliable suddenly showing blank, check whether its field name has changed in the source tool first, then contact FullSpec to update the field mapping.
Process Runbook / SOPPage 4 of 4