You can automate client status updates only halfway. What happened and what is next comes straight out of the system the work already sits in. Whether the agreement still holds is a judgement, and a person gives it. So the workable form is a prepared update, not an automatic one.
This matters most if you run an agency on retained clients. The question arrives from every client in the same week, the answer sits in a different project each time, and retainer margin disappears into the time someone spends collecting it.
What goes into a client status update?
A client status update answers what moved, what is next, and whether the agreement still holds. The first two are facts that already exist somewhere in your stack. The last one is a sentence a person adds. Everything beyond those three is padding that trains the client to stop reading.
- What was delivered since the previous update.
- What is in progress now, and who is working on it.
- What is next, with the date it is planned for.
- What the project is waiting on from the client.
- Whether the agreed date still holds, in one line.
- The name of the person who sent it.
Fact or judgement: the split that makes the update work
A status update carries two kinds of information, and only one can be automated. Facts are the state of the work: delivered, in progress, planned for a date. They sit in a system and a workflow can fetch them. A judgement is the answer to what the client is actually asking when they ask how things are going: is the promise still good?
That judgement cannot be derived from the data, however complete the data is. A task marked done says nothing about a client who has stopped trusting the date, and a project that lost its only reviewer still reads as progress to a workflow. The judgement needs someone who knows what was promised and how the last call went.
Most advice on status updates lands on the message instead: a template, a reporting view, a summary generated inside a project tool. Automating the message is the easy half, and it is the half that cannot go out unread.
The same split runs through a report to your own management, but the receiver changes the shape. A board report answers in figures and sits in a dashboard, which is why WeAdapt treats automated management reports as a separate build. A client update answers one question, in words, to someone who replies.
Where does the content of the update come from?
From the system the work already sits in, pulled on a fixed day. In n8n a workflow is a collection of nodes connected to automate a process[1], the nodes fetch, send and process data and hold the flow logic[2], and a production workflow needs at least one trigger to decide when it runs[2]. A Schedule Trigger supplies the fixed day: it runs workflows at fixed intervals and times[3].
| What the client wants to know | Where the answer lives | Who delivers it | What you do when it is missing |
|---|---|---|---|
| What moved since last time | Task rows in the sheet the work is tracked in[4] | System | Nothing to fetch is not nothing to send |
| What is next, and when | The planned date field on those rows[4] | System | A person fills in the date, or the line is dropped |
| Whether anything changed in the agreement | Deal and contact fields in the CRM[5] | System | Leave the line out rather than guess |
| What the project needs from the client | An open question in someone’s head or a thread | Person | A person writes it; never generated |
| Whether the agreed date still holds | Nowhere in any system | Person | The update waits for an answer |
Delivery is the last node. A Slack node posts the message to a channel or a person[6]. A Gmail node sends it, or stages it, because the node handles drafts as well as messages[7]. That draft is the whole trick: the facts arrive assembled, a person adds the judgement line and presses send.
A prepared update fails quietly, because nobody misses a message they were not expecting. An Error Trigger node builds an error workflow that gets the details of the failed run and the errors[8], and the workflow settings are where you point at it[9]. WeAdapt wires that in on the first build, because an automation which stops running is worse than one that was never built.
How often do you send an update with no news?
On the same day as always. A week in which nothing was delivered is still a week the client spent waiting, and silence is what makes someone ask for a call. One line about where the work stands plus the date of the next step is an update. An empty list of completed items is not.
The fixed day is what the system is good at: being on time and remembering, including the weeks that are dull. A person forgets those weeks first, and a client notices.
There is a stop rule. When the update has nothing to report three weeks in a row, that is a signal about the project and not about the message. At that point a person calls the client and the scheduled update stays off until the work moves again. One more empty update after that teaches the client to ignore all of them.
When does a person write the update instead?
The moment the judgement changes. A date that will not hold, a decision only the client can make, something that went wrong: a person writes that update themselves and the system sends nothing that week. The workflow still assembles the facts, which help the person writing, but the send step stays closed while the flag is set[2].
This is the boundary WeAdapt builds into every client-facing flow, from client onboarding onward: the system collects and times, people keep the promises. It is also why a build starts on the systems you already run instead of a new tool, which is the premise of workflow automation as a service.
Who signs the update
One person, by name, not “the team”. A status update is a promise about what happens next, and a promise needs an owner the client can reply to. A shared inbox as sender turns the judgement line into an anonymous claim. In the multi-client setup WeAdapt built for SEM Agency, one system serves every client, and every update still carries one name.
Where WeAdapt would start
With one client, one fixed day, and a draft instead of a send. The workflow collects the facts from the system the work already sits in, stages the message, and leaves the judgement line manual. Once that runs without anyone chasing it, the same build extends across the book. Which updates are worth systemising, and which client conversations should stay conversations, is a consultancy question.
Want the weekly question answered before your clients ask it? Book a call.