Template · 9 min read
Weekly report template: 15 minutes to write, two minutes to read
The weekly report is the most repeated document in client work and the least designed. Most people re-invent it every Friday — new structure, new length, new level of honesty depending on the week — which is why it takes an hour to write and why clients quietly stop reading it. The fix is a format so fixed it's almost boring: five sections, same order, same definitions, every single week. Below is that format with the reasoning, a prompt that turns your rough notes into the finished report, and the delivery pattern that makes fifty-two of these compound into something bigger than any one of them.
What a weekly report is actually for
On the surface it answers "what happened this week?" Underneath, it answers the question every retainer client asks themselves monthly: "do I still know what I'm paying for?" Clients rarely churn over a bad week. They churn over accumulated silence — weeks where work happened but nothing visible reached them, until the invoice is the only regular communication they get. The weekly report is the cheapest churn insurance that exists: fifteen minutes that keeps the value visible fifty-two times a year.
That purpose sets the bar for the format. It must be cheap to write, because a report that takes an hour gets skipped on busy weeks — and skipped weeks are exactly the silence it exists to prevent. It must be fast to read, because its reader gives it two minutes on a Friday. And it must be identical in shape every week, because the real payload isn't any single report — it's the trend. A client who has seen the same five sections for six months can spot drift, progress, and trouble at a glance. Change the format and you reset that instrument to zero.
The honest counterpoint: not every engagement needs one. A two-week sprint doesn't; a monthly advisory retainer might do better with a monthly note. The weekly report earns its cost on ongoing execution work — the kind where the client can't see the work happening and the relationship runs on trust plus evidence.
The weekly report template, section by section
Five sections. The order is load-bearing: verdict first, work second, problems unmissable, future last.
- The number. One metric — the one this engagement exists to move — shown against its target with a green/red delta and a single sentence of interpretation: "Qualified leads: 14 vs. target 12 — second week above target; the new landing page is holding." Opening with the same number every week is what turns fifty-two reports into a trend the client trusts. Resist the dashboard urge: one number, deliberately chosen, beats nine that dilute each other.
- Wins. Two to four bullets of what shipped or moved, each tied to a goal of the engagement rather than ordered by how much effort it cost you. The client doesn't renew because you were busy; they renew because busy translated into their goals.
- In progress. Two or three bullets with expected completion. This section manages expectations cheaply — it's the difference between "still nothing on X?" and "right on schedule."
- Blockers & asks. What's stuck and what you need from the client, stated plainly, visually impossible to skim past. This is the section that does real work: a blocker named on Friday is a project risk; the same blocker surfaced six weeks later is your fault. If there are none, say "none" — the empty section is itself information, and deleting it on good weeks teaches the client to stop looking for it on bad ones.
- Next week. Two to four bullets. Small print, big effect: it's a public commitment that next Friday's "wins" section will be graded against, and clients notice when the two consistently match.
Two rules that outrank everything above: never redefine the metric — a "lead" in week 30 must mean what it meant in week 1, or the trend line the whole format exists to build becomes fiction — and never skip a bad week. The report that lands on time carrying bad news builds more trust than a month of good-news reports, because it proves the channel is honest.
One link per client, updated every Friday
every friday replaces the last — one bookmark all year
- Fri wk 33
Opened within an hour, as always
- Fri wk 34
Opened Monday, briefly
- Fri wk 35
Not opened at all
- Wk 36
Your move: call before the silence hardens
Deliver the report at one permanent URL per client, replaced every Friday. The client bookmarks it once — or just keeps the calendar reminder — and there's never a "which email was last week's?" hunt. Fifty-two attachments become one standing address, and the address itself starts carrying meaning: it's where the engagement lives.
The tracked link adds the signal the weekly cadence makes uniquely readable. One unopened report means a busy week. Three in a row is the most reliable churn tell in client services — disengagement you can see weeks before it shows up as a stiff email about "reviewing our vendor spend." That's not a reason to panic; it's a reason to pick up the phone while the relationship is still warm. No other document gives you this early-warning system, because no other document establishes a weekly baseline to deviate from.
The honest limits: opens aren't reads, a forwarded link muddies who's looking, and some mail clients prefetch links. Judge the trend against the client's own baseline, not any absolute standard — the signal is the change.
Build it with AI from your rough notes (prompt included)
The fifteen-minute budget is realistic because the hard part — remembering what happened — is solved by keeping rough notes as the week goes, and the tedious part — formatting — is solved by the prompt below. Friday afternoon: paste your notes into the bracketed slots, generate, spend two minutes reviewing, publish. The prompt enforces the fixed structure so the format holds still even when you're rushed.
Copy, fill the brackets with this week's real notes and numbers, paste into Claude or ChatGPT:
You are producing a weekly client report as a single, self-contained HTML file. CONTEXT — replace everything in brackets: - Client: [client name] - Week: [week of DATE] - Prepared by: [your name / company] - The one number that matters, with its target: [metric — actual — target] - Wins this week: [2-4 bullets, each tied to a goal of the engagement] - In progress: [2-3 bullets with expected completion] - Blockers and asks: [what is stuck and what you need from the client, or "none"] - Next week: [2-4 bullets] - Brand: [primary hex color] + [logo URL, or "text wordmark"] REQUIREMENTS: - Sections in this exact order: The number · Wins · In progress · Blockers & asks · Next week. Never reorder, never add sections. - Lead with the metric against its target, with a green/red delta and a one-sentence read on the direction. - Keep every number exactly as I gave it. If something is missing, leave a clearly marked [TBD] — never invent data. - If the blockers section has content, make it visually impossible to miss — that is the section the client most needs to see. - Total length: readable in two minutes. Cut before you pad. - Self-contained HTML: inline CSS, no external scripts, readable on a phone. Stamp "Week of [date]" under the title. - Tone: direct, factual, zero filler. A weekly report is a habit, not an essay.
- 1 Paste the prompt (with your details filled in) into Claude or ChatGPT. You get a finished, self-contained HTML document — review it, correct anything, iterate until it's right.
- 2 Publish it as a permanent tracked link: with the Pagelive connector in Claude, say "publish this page" and the link comes back in chat — next cycle, "update it" replaces the content at the same URL. Or paste the HTML at app.pagelive.io.
One review habit worth keeping: check the "wins" section against last week's "next week" before publishing. That's the accountability loop the format creates, and it's the first thing a sharp client checks too. If they don't match, say why in a sentence — the miss explained beats the miss hidden, every week.
The mistakes that kill weekly reports
- The activity dump. Fourteen bullets of everything touched, none tied to a goal. It reads as busy-ness, takes longer to write, and buries the two things that mattered.
- Skipping the bad weeks. The most tempting mistake and the most expensive. A gap in the cadence tells the client something went wrong and that your reporting can't be trusted to say so.
- Format drift. New sections, reordered sections, a different metric up top "just this week." Every deviation resets the trend and makes next week's report harder to write.
- Burying the blocker. The thing you need from the client, mentioned in passing at the bottom of bullet six. When it stalls the project, the paper trail shows you technically mentioned it — which convinces nobody, including you.
- Writing it Monday for last week. Late reports read as afterthoughts. The value is the rhythm; a report that drifts around the calendar trains the client out of the habit of reading it.
- Letting it grow. Reports have a ratchet problem — sections get added, none get removed. The moment it stops being a two-minute read, opens drop, and the whole instrument fails quietly.
What to build it in
Plain email is the honest default: zero friction, lands where the client already looks. Its weaknesses are the archive — fifty-two emails don't stack into anything findable — and the total absence of a read signal beyond hoping for a reply.
Docs and workspace tools (Google Docs, Notion) keep history well, especially a running doc with the newest week on top. They fit clients you already share a workspace with; outside one, permissions and logins add friction to a document whose entire job is being effortless to open.
Project-management tools can auto-generate status digests from tasks. Tempting, but an auto-digest is an activity dump by construction — it reports what was ticked, not what mattered, and clients can tell nobody wrote it. Fine as your source notes; weak as the report itself.
An HTML report at a permanent tracked link is the pattern above, and it's what Pagelive is built for: publish Friday's report to the same URL each week, let the client keep one bookmark, and watch the open pattern that makes the churn tell visible. The honest limits: nothing refreshes itself — you (or your Friday prompt routine) publish each week — and a client who genuinely prefers reading in their inbox should get the email too, with the link as the standing archive. This weekly rhythm also slots under the bigger cadence: the QBR reviews the quarter the weeklies documented, the project status report covers the milestone view, and the always-live client report covers the metrics-dashboard variant of the same idea.
Frequently asked
Five sections, in a fixed order: the one number that matters (against its target), wins, in progress, blockers and asks, and next week's plan. The discipline is keeping the order and the metric definitions identical every single week — the report's value is the trend, and the trend only exists if the format holds still.
Readable in two minutes, writable in fifteen. One screen is the target. If writing it takes an hour, the format is wrong — you're re-deciding structure every week instead of filling in a fixed one. Detail that doesn't fit belongs in a linked appendix or the next call, not in the body.
A status report answers 'where are things?' and can be purely internal. A weekly client report also answers 'was this week worth what you're paying me?' — it ties the week's work to the goals of the engagement. Same skeleton, higher stakes: it's the retainer's weekly proof of value.
AI can draft it well if you give it the real inputs — the numbers, what shipped, what's stuck — and a fixed structure to pour them into. What it can't do is know your week or own the judgment calls. The workflow that works: keep rough notes as the week goes, hand them to a prompt that enforces your format, review the draft in two minutes, publish.
Pick one day and time and never move it — Friday afternoon or Monday morning are the usual choices, and the consistency matters more than which you pick. A report that lands at the same hour every week becomes a habit on the client's side; one that drifts trains them to stop expecting it, and then to stop reading it.
Related: the always-live client report · QBR template · project status report template
Fifty-two reports. One link. You'll know the week they stop reading.
Draft it with AI from your notes, publish to the same tracked link every Friday, and keep the engagement visible all year. Free for 5 pages, no card.