Ask customers what makes support feel bad and the same answer keeps coming up: not slow resolutions, but silence. The ticket that vanished. The “we’ll get back to you” that never did. Most support teams answer incoming messages reasonably well; where they leak trust is the follow-up email, the message you owe them with nobody prompting you to send it.
One clarification before anything else, because most advice under “follow-up email” is about something different: this is not a sales sequence. Sales follow-ups chase a reply from a prospect who owes you nothing: “just bumping this to the top of your inbox,” send number four. A customer service follow-up is the reverse: an existing customer is waiting, and the obligation runs from you to them. The timing, tone, and templates that work for prospecting are actively wrong here, so if you landed on this page looking for cold-outreach cadences, this is the other kind.
The mechanics aren’t hard. There are exactly three moments where a follow-up in customer service earns its keep, each with a simple timing rule, and the messages themselves are short. The hard part is operational: remembering, and having the message cost so little to write that it actually gets sent. We’ll cover both, plus five templates you can copy today.
The three follow-up moments that matter
1. The customer went quiet
You asked for an order number or a screenshot, and the thread went silent. Maybe they’re on vacation; maybe your email hit spam; maybe they solved it and moved on. You don’t know, and an open ticket in limbo helps nobody.
Timing rule of thumb: first nudge after 2–3 business days. Second and final nudge 3–4 business days later, with a stated closing date. Then a courtesy close that makes reopening effortless. Chasing harder than that reads as pestering; closing silently after one attempt reads as not caring.
2. A promised update is due
You said “I’ll update you by Thursday.” It is now Thursday and the fix isn’t done. This is the moment that separates trusted teams from the rest, and the follow-up must go out on the day you promised, news or no news.
Timing rule of thumb: the date you said. Not the day after. If your team can’t reliably hit self-imposed dates, promise looser ones (“by end of week”) and beat them.
3. Post-resolution check-in
The issue is “resolved,” but some resolutions deserve verification: a workaround holding the line until a real fix, a replacement in transit, a tricky configuration change, or any case that followed an escalation or an apology. A short check-in confirms it actually worked and catches the quiet failures before they become angry reopens.
Timing rule of thumb: 3–7 days after resolution, calibrated to the case: 3 days for a workaround, 5–7 for a shipped replacement. Reserve check-ins for cases with genuine failure risk or relationship damage; checking in on every password reset is noise.
Why silence after “we’ll get back to you” burns trust
A promised update is a small contract. When Thursday passes with nothing, the customer doesn’t think “they must be busy”; they think “they forgot me,” and every hour of silence after the deadline compounds it. By the time they write “any update???” on Monday, you’re no longer handling a technical issue; you’re handling a technical issue plus a credibility problem, and their patience for the remaining wait has collapsed.
The perverse part: teams go silent precisely because there’s no news, reasoning that an empty update wastes the customer’s time. It’s exactly backwards. “Still working on it, here’s what we’ve ruled out, I’ll write again Friday” costs the customer ten seconds and tells them the one thing they actually need to know — someone is holding this. No-news updates are cheap insurance on the most expensive asset you have.
Five copyable templates
Placeholders in brackets; cut what doesn’t apply.
Template 1: First nudge after no reply
Subject: Still happy to help with [issue], one thing needed from you
Hi [Name],
Following up on your ticket about [issue]. To move it forward I still need [the specific thing: order number / screenshot / confirmation], and I haven’t heard back since [date].
No rush if things are busy; reply whenever suits, and I’ll pick this straight back up.
[Name]
Template 2: Final nudge with closing date
Subject: Closing your ticket about [issue] on [date], unless you still need it
Hi [Name],
I’ve checked in a couple of times about [issue] and haven’t heard back, so I’ll close the ticket on [date] to keep things tidy.
Nothing is lost: reply any time, before or after that date, and it reopens with the full history. If you sorted it out yourself, no reply needed at all.
[Name]
Template 3: Promised update, no news yet
Subject: [Issue]: the update I promised for today
Hi [Name],
I told you I’d have an update by today, so here it is honestly: not resolved yet. So far we’ve [what’s been done/ruled out], and the next step is [next step].
New checkpoint: I’ll write again by [date] whether or not it’s fixed by then. Nothing needed from you.
[Name]
Template 4: Promised update, with progress
Subject: [Issue]: good progress, one step left
Hi [Name],
Update as promised: we’ve confirmed the cause was [plain-language cause], and [what’s been done]. In practice that means [what the customer should already notice].
The remaining step is [step], expected by [date]. I’ll confirm here as soon as it lands.
[Name]
Template 5: Post-resolution check-in
Subject: Checking in: is [issue] still behaving?
Hi [Name],
It’s been [X days] since we [fixed / shipped the replacement / set up the workaround] for [issue], and I wanted to make sure it’s actually holding up on your side.
If everything’s fine, no reply needed. If anything is off, even slightly, reply here and it comes straight back to me rather than the general queue.
[Name]
Note what all five share: a specific subject line, a reason for the email in the first sentence, an explicit “no reply needed” where that’s true, and a next step with a date where it isn’t. If a check-in surfaces a still-broken issue after you’ve already apologized once, don’t send template 5 again — escalate to a proper apology; the structure is in our guide to writing an apology email to a customer.
For high-frequency teams, templates 1 and 2 are strong candidates for your snippet library; they change little between cases. More on building that library in 15 canned responses that don’t sound canned.
Automate the draft, not the send
The failure mode of follow-ups is rarely bad writing; it’s that nobody sends them. So the instinct is full automation: fire template 3 automatically when a due date passes. Resist it. Fully automated follow-ups develop a recognizable dead-eyed cadence, and worse, they fire in contexts a human would catch: the customer who already replied on another channel, the ticket attached to an active incident bridge, the account mid-cancellation.
The durable setup splits the job:
- Automate the trigger. Your helpdesk reminds the agent: promised update due today, customer quiet for 3 days, resolution 5 days old. Machines are perfect at dates.
- Generate the draft. This is where an AI drafting step removes the real friction. In Replydesk, paste the ticket thread, pick the reply workflow, and get a follow-up draft that already reflects what’s actually in the thread (what was promised, what’s been tried, what’s still open) rather than a template you now have to fill from memory. The free tier’s 20 drafts a day covers a full queue’s worth of follow-ups, no card required; how the drafting flow works in general is covered in our free AI email draft generator guide.
- A human sends. Ten seconds of review per message: does the date hold, is the tone right for this customer, is anything overtaken by events. Then send.
When the draft costs ten seconds instead of ten minutes, the no-news update stops being the task everyone defers and becomes the reflex it should be. That’s the whole game: follow-ups are cheap individually and enormously valuable in aggregate, and the teams that send them reliably are the ones customers describe, accurately, as the ones that never drop anything.