Nick QuirkMost agencies have a client update problem they don't know they have. The update goes out every week. Nobody complains. Then, four months in, the client asks "so what have we actually done on technical SEO?" — and you realize they never read a single one.
A client-ready project overview isn't a status report you export. It's a deliberately short, written artifact designed for one specific reader: a busy person who cares about outcomes, not your task board. Getting this right does more for retention than almost any other operational habit, because it's the only part of your process most clients actually experience.
Why most client overviews go unread
Three failure patterns show up over and over when I audit agency reporting:
1. It's a task dump. Someone exports 47 completed tickets — "Updated meta description /pricing," "Reviewed GSC coverage report" — and pastes them in. The client sees noise and concludes you're busy, which is not the same as concluding you're effective. Worse, the volume invites scrutiny of individual line items instead of the outcome.
2. It answers questions nobody asked. Internal process detail — which sprint something landed in, who picked up the reassigned ticket — reads as filler to a marketing director with 200 unread emails.
3. It buries the one thing they wanted. The client opens the update looking for "did the migration date move?" and has to scroll past 600 words of content-calendar detail to find out. Once that happens twice, they stop opening.
The fix isn't better formatting. It's deciding, on purpose, what the overview is for.
The four questions every overview must answer
Before you write anything, know that a client reads an update to answer four questions. In this order:
- Are we on track? One sentence, unambiguous. "On track for the June 12 launch." Not "progressing well."
- What did I get for my money since last time? Two to four outcomes, not tasks.
- What's coming next, and when? Dates, not "soon."
- What do you need from me? Explicit, named, with a deadline.
If your overview answers those four things in under 250 words, it will get read. If it answers three of them and drifts, it won't. That's genuinely most of the discipline.
The one-sentence status line
Start every overview with a status line the client could forward to their CEO without editing. Examples that work:
- "On track. Audit delivered, 31 of 44 fixes shipped, remaining 13 blocked on dev capacity until Feb 3."
- "At risk. The staging environment has been down 6 days, which pushes the content migration one week to March 14."
- "Ahead of plan. Organic sessions up 18% MoM; we're pulling Phase 3 keyword expansion forward."
Notice each one contains a number and a date. "At risk" with no consequence attached is a mood, not a status. Say what it costs.
Write outcomes, not activity
This is the single biggest upgrade available to most agencies. Compare:
Activity: "Completed 14 on-page optimization tasks, updated internal linking on 9 pages, fixed 6 broken redirects."
Outcome: "Cleaned up the redirect chains and internal links on the /services cluster — those 9 pages now have a clear path from the homepage, and three of them moved from page 3 to page 2 for their target terms."
Same work. The second version tells a client why they paid you. It also survives being read aloud in a meeting where you're not present, which is where retainer decisions actually happen.
Practical rule: for every activity you're tempted to list, write "…which means…" and finish the sentence. If you can't finish it, the item probably belongs in the internal log, not the client overview. This is the same instinct behind good phased to-dos rather than endless task lists — grouping work by the outcome it produces rather than the order someone happened to do it in.
Structure that survives a 45-second read
Here's a template that's held up across dozens of agency retainers. Total length: 200–300 words.
1. Status line (1 sentence)
As above. Bold it. It's the only thing some readers will see.
2. Shipped (3–4 bullets, outcome-framed)
Cap it at four. If you did more, pick the four that matter most to the client's stated goal. Volume is not the argument you want to be making.
3. Next (2–3 bullets with dates)
Each item gets a date or a date range. "Technical audit of the new subdomain — draft to you Feb 14."
4. Needs from you (0–3 items, named person, deadline)
This section is the highest-ROI part of the document and the most commonly fumbled. Don't write "awaiting client feedback." Write "Sarah — approval on the 6 title tag revisions in the shared doc by Thursday, or the Feb 3 push slips a week." Naming the person and the consequence is the difference between a 60% and a 95% response rate. It's also your paper trail when a delay wasn't yours, which is quietly one of the better defenses against scope creep and timeline drift.
5. One number (optional, but strong)
A single metric tied to the engagement's goal. Not a dashboard screenshot — one number, with context. "Non-brand clicks: 4,210 last 28 days, up from 3,560. Baseline at kickoff: 2,890."
Match the overview to the engagement type
A one-off audit and a 14-month retainer need different overviews.
Project-based work (migration, audit, site build) benefits from phase framing. "Phase 2 of 4 complete" gives clients a sense of progress that raw task counts never will. If you're structuring the work in phases anyway — as you should for something like a properly structured SEO audit project — the reporting layer comes almost free.
Retainers have the opposite problem: nothing ever "completes," so updates blur into sameness. Fix this with a rolling 90-day view. Show the theme for this quarter ("Q1: fixing indexation and consolidating the blog"), what's done against it, and what's left. Clients renew retainers when they can see a narrative arc, not a treadmill. If your pipeline mixes both models, the reporting cadence is part of what makes that retainer/project balance sustainable rather than chaotic.
Make it cost you 10 minutes, not 90
An overview format nobody can sustain is worse than no format. If your PM spends 90 minutes assembling updates for eight clients every Thursday, the quality will decay by client six.
Two things make this cheap:
Write the outcome when you close the work, not on report day. When a task moves to done, add one sentence of client-facing language right then. On Thursday you're editing, not reconstructing. This is a 20-second habit that saves an hour a week.
Let the tool assemble the skeleton, and you write the top. Pulling completed items, phase status, and hours against retainer should be automatic. The status line and the "needs from you" section should be human. Any decent agency PM setup can handle the assembly — PeakKR was built around this split specifically because agency reporting is the part that gets skipped when it's manual. If you're evaluating options, the practical question isn't feature count but whether client-facing output is a first-class thing or an export; that's the lens I'd use across any PM tool comparison.
Two habits that separate good from great
Send bad news first and early. Clients forgive slipped dates. They do not forgive finding out about a slipped date three weeks after you knew. A status line that says "at risk" in week two buys enormous credibility for the "on track" lines that follow.
Ask, once, what they actually want to see. On the kickoff call: "Who reads these updates, and what do they need to know?" One agency I worked with discovered their overviews were being forwarded verbatim to a CFO who only cared about spend-to-date and pipeline impact. They added two lines. Renewal conversations got dramatically easier.
Checklist: before you hit send
- Status line is one sentence, contains a number or a date, and is honest
- Under 300 words total, readable in 60 seconds
- Every "shipped" item survives the "…which means…" test
- Every "next" item has a date, not "soon" or "in progress"
- Every client ask names a person, a deliverable, and a deadline
- Zero internal jargon: no ticket IDs, sprint numbers, or tool names
- One metric, with a baseline for context
- Bad news appears in the first three lines, not the last
- It reads fine if forwarded to someone who's never spoken to you
- It took under 15 minutes to produce
