Most articles on this topic will tell you to send your survey "within 24 hours" and call it a day. That advice is wrong twice. It's wrong about the window — 24 hours is already past the point where most responses actually convert — and it's wrong about the assumption that there's one right answer for every service. There isn't.
Here's what we've learned watching service businesses send post-job surveys: the window changes by job type, and the difference between "optimal" and "close enough" is a 2-3× response rate.
The general rule: 2 hours after job-close
For most service work — a plumbing repair, an HVAC tune-up, a completed cleaning visit — the optimal moment is roughly 2 hours after the job is marked complete. Late enough that the customer has had time to walk around and confirm everything works. Early enough that the experience is still vivid.
Why 2 hours and not 30 minutes? At 30 minutes the customer is often still doing the thing they were doing before you arrived — picking the kids up, finishing dinner, getting back to work. The survey gets dismissed as "not now."
Why 2 hours and not 24? At 24 hours the customer's memory of the specific tech has faded into a generic impression of your company. Detail-level feedback drops; you get vague scores without actionable verbatims.
Adjust by job type
Emergency calls — fire at 6 hours, not 2
For burst pipes, no-heat, no-cool, sewer backups — anything where the customer was under acute stress when you arrived — wait longer. The first 2 hours after relief, the customer is decompressing. A survey at that moment competes with the emotional letdown.
Wait until the next morning if the emergency happened overnight, or 6 hours after if mid-day. The customer has had time to feel the relief settle in and is actively willing to thank you.
Multi-day jobs — fire after the inspection passes, not the install
Electrical work, plumbing remodels, restoration — anything that involves permits or follow-up inspections — should survey on the inspection-pass event, not the work-complete event. The customer's real experience of the job isn't complete until they know it's done right.
Most field-service software treats "job complete" and "inspection passed" as the same status. Fix this in your CRM or trigger config — if it's an option — to fire on the later event for permit-dependent work.
Recurring service — fire SAME-DAY, every visit, but make it 1 question
For residential cleaning, lawn care, pool service, pest control — anything where the same customer rates you 12-20 times a year — fire the same evening but cut the survey to a single question:
"On a scale of 0-10, how was today's visit?"
The full 3-question survey is too much to ask every two weeks; the response rate collapses to near-zero. A 1-question SMS lets you track per-visit NPS across the year, which catches quality drift you'd otherwise miss. Send the full survey once a quarter as a separate "how are we doing overall" check-in.
Routine maintenance — fire 4 hours after, batch detractors
Quarterly HVAC tune-up, annual gutter cleaning, fire-safety inspection — non-urgent recurring work — fire ~4 hours after the visit. The customer isn't stressed; they're doing whatever they normally do; a low-key SMS lands well.
Day-of-week and time-of-day
Two patterns that consistently outperform across trades:
- Survey response rates peak Tuesday-Thursday between 5pm and 8pm. People are home, dinner is starting or done, phones are in hand. Friday evenings drop because weekend-mode kicks in; Sunday evenings drop because weekend-recovery-mode kicks in.
- Saturday morning is a sleeper hitfor home-services work done Friday. Customers wake up, the repaired thing is still working, they're in a good mood. Don't over-think it for residential trades — a 9am Saturday SMS works.
Don't fire surveys between 9pm and 8am local. Even SMS with sound off generates resentment. The customer reads it at 7:30am, deletes it without thinking, and the response opportunity is gone.
What about the second touch?
Most playbooks recommend a follow-up survey 48 hours later if the first one wasn't answered. We don't. Two reasons:
- A second touch on an unhappy customer is a free invitation to escalate. Silence after one survey isn't neutral — it's often a soft "don't bother me." Pinging again can trigger the bad review you were trying to avoid.
- The response rate on a second touch is typically under 5%. The signal-to-noise is bad enough that it's easier to just learn from the first-touch data and move on.
The exception: if the customer paid late or disputed an invoice, that's a signal you should reach out manually — not via another survey. Phone call from a manager. (We cover this in how to recover unhappy customers before they leave a bad review.)
What about the channel?
Short version: SMS first, email as fallback. The data is consistently 4-5× higher open rates and 3× higher completion rates on SMS for service-business audiences. Email survey is what you fall back to when you don't have a verified phone or the customer explicitly opted out of SMS.
Don't do both. Sending email AND SMS for the same job triggers spam-receipt instincts; many customers will just block you.
How to actually implement this
Manually tracking which job type fires which survey on which cadence is unmanageable past 10 jobs/day. The whole point of a tool like Canopy is automating the per-job-type, per-trigger timing rules above so you don't think about it.
We integrate with the field-service tool you already use (HubSpot, Pipedrive, Salesforce, HousecallPro, ServiceTitan) so the survey trigger fires on the right job-close event, with the right delay, for the right job type, via the right channel — all automatically.
30-day trial, one click. If you want the per-trade overview first, try HVAC, plumbing, or residential cleaning.