Chief of staff weekly report: a template and the numbers that belong in it

It's Friday at 4 pm. You asked 6 team leads for their numbers: 2 sent screenshots, 1 sent a link to a sheet that has changed since Tuesday, and finance says revenue will be ready Monday at 10 am, an hour after the leadership meeting.

Writing the weekly report is not the hard part. Getting the same numbers, defined the same way, from the same people every week is. This guide gives you a template of about 10 numbers, a rule of one owner and one source per number, a way to collect them without chasing, and a list of what to leave out. The template is also a CSV you can download.

What is a chief of staff weekly report for?

The Monday pack has one job: tell the leadership team whether the company is on plan, and if not, where it is off and what needs deciding. It is not a status essay from every team. A useful test is that a leader could act on every number in it that week.

That sets the format. One page, numbers first, about 10 rows, each next to last week and the plan. Words go in 2 places only: the top risks and the decisions needed. You can see a dashboard built from 4 of these numbers, with its own sample data, in the chief of staff weekly pack example.

Which numbers belong in the weekly report?

The template uses 6 sections. Write a one-line definition for each number before the first report goes out.

• Revenue or pipeline. Invoiced revenue month to date against plan to date, so a slow first week doesn't look like a bad month. Add weighted pipeline: each open deal's amount times its deal probability, which is how HubSpot defines its weighted amount property. Decide which stages count before you start.
• Hiring. Hires against the quarter's plan, plus offers out. Count a hire when the offer is accepted or when the person starts, and write down which.
• Key results. The count of key results on track. Google's re:Work guide grades OKRs from 0.0 to 1.0, puts the sweet spot at 60 to 70%, and says that someone who always scores 1.0 isn't setting ambitious enough goals. Define on track in one line, for example a projected score of 0.7 or higher at the latest check-in.
• Cash and runway. Cash on hand from the bank, and runway: cash divided by average monthly net burn over the last 3 months. Note anything left out, such as a one-off tax refund.
• Customers. Paying customers at the end of the week and customers lost this week. Pick one meaning of paying, for example an active paid plan with free trials excluded.
• Top risks and decisions needed. At most 3 risks, each with an owner and a date, and the decisions leadership must make this week. These are words, but they still need an owner.

Runway deserves one more line. Paul Graham's essay "Default Alive or Default Dead?" asks a sharper question than months of cash: assuming expenses stay constant and revenue keeps growing at its recent rate, does the company reach profitability on the money it has? If your CEO thinks in those terms, add it as a note on the runway row.

What does the template look like?

Here is the template with hypothetical numbers for a company of about 150 people. The pack goes out Monday, October 12, 2026, with data through Friday, October 9.

Every row has exactly one owner and one source. The chief of staff owns the pack and the definitions, and only the rows nobody else can own.

The downloadable CSV has 9 columns: Section, Metric, Owner, This week, Last week, Change, Source, Data through, and Note. Data through is the easiest column to skip and the one readers need most. In this example, key results were scored at Thursday's check-in, so that row runs through October 8, and the runway row divides Friday's cash by burn averaged over July to September. A reader who sees both dates won't mistake one for the other. Use ISO dates such as 2026-10-09 in the sheet. They sort correctly and can't be read as day-month or month-day.

How do you collect the numbers without chasing people?

Chasing happens when the request is new every week. Make it the same request, in the same place, with the same deadline.

1. One shared sheet, one tab per owner. Each lead updates their own tab, and the summary tab reads from them. In Google Sheets you can limit who can edit a sheet or range to specific people. Google notes that this is not a security measure, since people can still copy the data.
2. A fixed deadline. For example, Friday at 5 pm for data through that day, which leaves Monday morning for questions. Same day and time every week.
3. Let the sheet tell you who updated. Google Sheets notification rules can email you when anyone changes the spreadsheet, either right away or as a daily digest. In Excel for Microsoft 365, Show Changes lists who changed a cell, when, and the previous value, for up to 365 days back.
4. Owners explain their own changes. If pipeline jumps $90k, the sales lead writes why in the Note column before the deadline, not in a Monday thread.
5. No update shows as no update. If a lead misses the deadline, the pack says "no update, last data October 2" instead of last week's number. A visible gap is easier to discuss than a quietly copied number.

What should you leave out?

A weekly pack gets worse as it grows, because each added row dilutes the ones that matter. Leave out:

• Numbers nobody decides with. If website visits never change what leadership does this week, report them monthly elsewhere.
• Activity counts. Meetings held or tickets closed belong in team reports unless they are a goal.
• Percentages without a base. "Conversion up 12%" means little without "from 50 to 56 out of 380."
• Second versions of a number. If finance owns revenue, sales does not bring its own revenue figure.
• Long commentary. One line per note. Anything longer becomes an agenda item.

How do you keep the numbers comparable week to week?

Keeping numbers comparable takes 3 habits. Write each definition once, in a metrics dictionary, and link it from the pack. Use the same cutoff every week, such as data through Friday. And date every definition change with its effect, for example: "From October 1, free trials no longer count as paying customers. September restated: 23 fewer."

Keep a dated copy of each week's pack. Then "Last week" is copied from a file, not retyped from memory, and anyone can check what was reported.

Where Every Number fits

Every Number builds dashboards where every number has a receipt: its definition, the filters applied, the rows used, the rows left out and why, the source file, and the date the data runs through. Today that is a $99 concierge pilot. You send one spreadsheet, a person builds 5 to 7 metrics in 5 business days, and you approve each definition before anything is built. See the chief of staff use case or the pilot.

The pilot works from one spreadsheet. If each lead keeps a separate file, the first step is the shared workbook above, which is worth doing on its own. If a BI team already builds your pack and leadership trusts it, you probably don't need us.