Have the weekly report write itself from the systems you already use

Connect your dashboards to an agent that pulls the numbers and writes your weekly update in your voice.

For anyone who owes someone an update every week · 8 steps · 7 min

How it works today.

Every Friday afternoon, you open your CRM dashboard, your analytics tool, your project management system, and maybe a spreadsheet or two. You copy numbers into a document template, calculate percentage changes from last week, and write the same commentary structure with updated figures. The actual writing takes 20 minutes, but gathering the data takes another 30.

You know exactly what your stakeholders want to see (revenue pipeline, key metrics, project status, blockers) but pulling it together feels like busy work. You have the same three-paragraph structure every time, just with different numbers and a rotating cast of wins and challenges.

Before you start.

All of it has to be true, or step one fails in a way that is annoying to debug.

  • API access or export capabilities for your main data sources (CRM, analytics, project tools)
  • A document template or previous weekly reports to establish your format and voice
  • List of specific metrics and data points your stakeholders expect each week
  • Access to a builder tool that can connect to your data sources and generate documents

The steps.

  1. Map your data sources and required metrics

    List every system you currently check for your weekly report and the specific metrics you pull from each. Note whether each source has an API, export function, or requires manual copying. Document the exact dashboard URLs, report names, or database queries you use.

  2. Set up data connections in your builder tool

    Connect your builder to each data source using available APIs or automated exports. Test each connection by pulling a sample of your required metrics. If a source lacks API access, set up scheduled exports to a shared location your builder can access.

  3. Create the data collection workflow

    Build a workflow that pulls all your required metrics on a schedule. Include error handling for when APIs are down or data is missing. Test the workflow manually to ensure it captures everything you normally gather on Friday afternoons.

  4. Write the report generation prompt

    Create a prompt that takes your raw data and formats it into your standard report structure. Include your typical commentary style, how you present wins and challenges, and any specific language your stakeholders expect.

    Paste this
    Using this weekly data: [data from workflow], write a weekly update report following this structure: [paste your template]. Use this voice and tone: [paste examples from 2-3 previous reports]. Highlight percentage changes from last week, call out any metrics that moved significantly, and include a brief commentary on what drove the changes. Keep the executive summary to 3 sentences.

  5. Set up the automated report generation

    Connect your data collection workflow to your report generation prompt. Configure it to run every Friday morning and output the report to your preferred format: a document, an email draft, or a shared folder. Test the full end-to-end process with real data.

  6. Create review and approval gates

    Set up the system to generate a draft report and send it to you for review before distribution. Include a simple approval mechanism: reply with 'approved' or edit the draft directly. The system should not send reports to stakeholders without your explicit approval.

  7. Configure stakeholder distribution

    Set up automated distribution to your stakeholder list once you approve the report. This might be email, posting to a shared channel, or updating a dashboard. Include your standard subject line format and any required attachments or links.

    Paste this
    Send this approved weekly report to [stakeholder email list] with subject line 'Weekly Update - [current date]'. Include standard footer: 'Questions? Reply to this email or find me on [your preferred channel].'

  8. Test the complete workflow

    Run the entire process manually for two weeks to catch edge cases. Check how it handles missing data, API timeouts, or unusual metric spikes. Adjust your prompts based on the quality of generated commentary and stakeholder feedback.

What you keep.

Automating the typing does not move the accountability. These stay with a person.

  • Final approval before any report goes to stakeholders
  • Interpreting unusual data patterns or explaining context the system cannot know
  • Adjusting the report format or adding ad-hoc sections for special situations
  • Managing stakeholder feedback and updating the system based on changing requirements

Once it works.

The first run is the demo. These are where the time actually comes back.

Fire it from an event

Set the system to generate emergency reports when key metrics hit predefined thresholds.

Put it on a schedule

Expand to monthly or quarterly reports using the same data sources with different aggregation periods.

Scale it wider

Create customized versions of the same report for different stakeholder groups with relevant metrics for each.

Run it on a loop

Use stakeholder replies and questions to automatically update next week's report focus areas.

Run this next.

One workflow is a tip. Chained, they are how a week actually changes shape.

Operations

Ask a question of a spreadsheet instead of building another pivot table

Get the numbers analysed before they go in the report.

7 stepsCorrespondence

Turn a meeting recording into notes, actions and the follow-up email

Feed the week of meetings into the same update.

7 steps

The work this replaces.

These are the O*NET work activities this workflow covers, and the categories of tool that address them.

Addressed by builders, tools and auth, generalist agents
n8nZapierComposioArcade.devChatGPT (agent mode)ClaudeWhole landscape