Free Post-Service Survey Template
Use this post-service survey to capture feedback while the interaction is still fresh, then pinpoint what drove the score (resolution, effort, speed, communication). You can run it as a lightweight CSAT + CES program, with optional NPS when you want a separate loyalty pulse. Set clear recovery triggers so low scores automatically route to follow-up.
Trusted by 5000+ Brands
Pick the Right Metric: CSAT vs CES vs NPS (Post-Interaction)
| Goal: Pick the metric combo you will trend after each ticket/visit, without overloading the survey. Default: Lead with CSAT (quick satisfaction check) and add CES (find friction). Keep NPS optional and run it less often. Do this now: Choose one primary metric for every send, then add 2-4 driver questions. If you are also planning loyalty tracking, separate it into a periodic NPS vs CSAT vs CES program. | |||||
|---|---|---|---|---|---|
| Metric | Typical question | Use it to do this | Best after | Pros | Watch-outs + how often |
| CSAT | "Overall, how satisfied were you with the support you received?" (1-5 or 1-7) | Track whether the last interaction met expectations and spot dips by team/channel. | Every closed ticket, completed visit, or completed appointment. | Fast signal, easy to explain, works well with a short driver set (resolution, speed, communication). | Scores vary by channel and issue type, so set internal baselines. Send right after the interaction. |
| CES | "How easy was it to get your issue resolved?" (1-5 or 1-7; label endpoints) | Find friction you can fix (transfers, repeats, unclear steps, waiting). | Every closed ticket when reducing effort is a priority, or as a rotating add-on (e.g., 50% of sends). | More actionable than a single satisfaction rating when the process is the problem. | Keep wording consistent so trends mean something. Pair with a "Resolved/Not resolved" branch. |
| NPS (optional) | "How likely are you to recommend us to a friend or colleague?" (0-10) | Run a loyalty pulse and compare segments over time. | Periodic relationship tracking (monthly/quarterly), not every ticket. | Widely recognized and easy to benchmark internally across groups. | In practice, NPS is not a universal predictor in every context or time horizon. If you use it, keep it separate from your per-interaction monitoring and trend it over time (see Keiningham et al. in the Journal of Marketing study on Net Promoter and growth). |
Once you have your metric set, lock in send timing next so responses reflect the actual interaction (not fuzzy recall). If you want a standards-based reference for monitoring customer satisfaction programs, align your workflow to the ISO 10004 overview of customer satisfaction monitoring.
When to Send This Post-Service Survey (Timing + Throttling)
After a support ticket is closed (send within 15-60 minutes)
Goal: Capture accurate, right-after feedback and tie it back to the closed case.
Default: Treat 15-60 minutes as a starter timing target (adjust after you review response rates and customer feedback). Many teams start around 30 minutes after close, using the same channel where the customer interacted.
Do this now: Set automation to trigger on "Closed/Resolved" status, then pause sends for that customer after a response.
- Channels: Email for detailed responses; SMS for speed; in-app link if you support logged-in users. For phone-heavy teams, use the channel-specific post-call survey template.
- Throttling rule (starter target): Cap at 2 invites per customer per week, then tune based on your baseline invite volume and opt-out rates. Add a 7-day cooldown after any completed survey to reduce fatigue.
After an on-site/in-home visit (send 1-3 hours after the visit or same-day)
Goal: Measure the visit experience while details like arrival window, professionalism, and fix quality are still clear.
Default: Treat 1-3 hours as a starter timing target (adjust for travel time, customer availability, and your channel). Many teams start around 2 hours after the technician marks the job complete.
Do this now: Trigger from your field service "Job complete" event and tag responses with tech ID, region, and job type.
- Channels: SMS with a short link (best right after a visit); email if you want a longer open-text comment; QR card left behind for low-connectivity areas.
- Throttling rule (starter target): Cap at 1 invite per customer per month for field visits, then adjust after you see your baseline completion and complaint rates.
After delivery/installation/appointment completion (send same day; within 2-6 hours)
Goal: Confirm the outcome (done, on time, explained well) and catch failures early enough to recover.
Default: Treat same-day (within 2-6 hours) as a starter timing target. Many teams start around 4 hours after completion, before the next business day starts, and then adjust based on customer availability and response patterns.
Do this now: Add one "Resolved/Completed as expected" question first, then branch to recovery if they say no.
- Channels: SMS for completion-day feedback; email if you need photos or longer comments; QR at the point of service if the customer is on-site.
- Throttling rule (starter target): Cap at 1 invite per week per customer across all appointment types, and suppress invites for any active open case. Tune these caps after you establish your baseline volume and opt-out rate.
Customize the Template for Your Service Workflow (Without Adding Drop-Off)
- Keep the core short (and consistent)
Goal: Customize wording for your workflow while keeping trendable results.
Default: CSAT + "Resolved/Not resolved" + 2-4 drivers + 1 open-text prompt.
Do this now: Lock the primary scale wording (1-5 or 0-10) and reuse it across channels for at least a quarter.
- Keep: One primary metric (CSAT or CES) and one outcome question ("Was your issue resolved?").
- Keep: A short driver set: effort, speed, communication, professionalism.
- Limit: 1-2 open-text questions max (mobile completion drops fast when text boxes multiply).
- Edit the wording to match the channel
Phone/chat example: Replace "technician" with "agent" and ask about "wait time" and "clarity of next steps".
In-person example: Ask about "arrival window," "explained the work," and "left the area clean" (only if those are actual standards you manage).
- Do this: Put the interaction type in the first question stem: "About your chat today..." or "About your visit today..."
- Do this: Keep answer labels symmetric (e.g., "Very dissatisfied" to "Very satisfied").
- Add branching that protects the customer (and your ops team)
Branching example A (resolution): If they choose "Not resolved," then ask:
- "What is still missing or incorrect?" (short open text)
- "Can we contact you to fix this?" (Yes/No)
- If Yes: "Preferred contact method" (phone/email/SMS) and "Best time"
Branching example B (effort): If CES is 1-2 on a 1-5 ease scale, then ask "Where was it hard?" with picklist options like "I was transferred," "I had to repeat information," "Steps were unclear," "Wait time was too long," plus "Other".
Do this now: Route "Not resolved" plus permission-to-contact to a recovery queue with a clear SLA.
- Run a quick bias check before you launch
Default: Neutral wording + one idea per question.
- Check for leading language: Remove words like "quickly" or "friendly" from the question stem unless you want to bias responses (use Kantar's unbiased survey question guidance as your checklist).
- Check for double-barreled items: Split "knowledgeable and friendly" into two questions or drop one.
- Check scale balance: Use the same number of positive and negative points and label endpoints.
- Keep it mobile-first (so you get completes, not drop-offs)
Default: 5-9 questions total, single-column layout, no long grids.
Do this now: Test the survey on a phone and aim for a short completion time (rule of thumb: about 1-2 minutes), then adjust after you measure your baseline completion rate (AAPOR's best practices for survey research include practical data-quality checks you can apply before sending).
- Keep scales short: 5-point satisfaction and 0-10 NPS are easy to tap.
- Put the open text last: You want scores even from busy customers.
Once your question set is locked, decide how you will handle small samples so you do not overreact to one bad day.
Sample Size and Small-N Rules (So You Do Not Overreact)
Goal: Report post-service results without false alarms from small samples.
Default: Use minimum-n rules as internal starter targets, then adjust after you establish your baseline volume and variability.
Do this now: Set minimum-n rules in your dashboard and combine low-volume categories (e.g., merge two small regions) until they clear the threshold.
- Minimums you can use today (starter targets):
- Overall program health: Trend weekly or monthly as volume allows.
- Team/tech cuts: Show results only when n>=30 in the selected window (starter target); prefer n>=50 before you rank teams (starter target). If your volume is lower, use a longer rolling window instead of publishing unstable cuts.
- Low volume categories: Combine categories or switch to a 90-day rolling view.
- Sampling distribution: Sample across channels, regions, and issue types instead of blasting one queue.
- Send invites proportionally (e.g., if chat is 30% of volume, aim for ~30% of completes from chat).
- Track contactability differences (SMS deliverability, email bounces) so you do not mistake coverage gaps for performance gaps.
Nonresponse handling: Track how many you sent vs how many completed, and watch for skew (for example, only angry customers responding after a bad outage). Your reporting will get more honest if you monitor nonresponse and response bias alongside CSAT/CES trends.
Outcome tracking: Log outcomes consistently (sent, delivered, opened, started, completed) and use standard definitions so your completion rates are comparable month to month (AAPOR's Standard Definitions for survey outcome rates is the clean reference).
- Small-n rule of thumb: Use trends over time, not single-period winners. Add simple confidence bands if your dashboard supports it, and avoid "best/worst team" callouts when n is low.
- Do this next: If you need a deeper walkthrough, use this site's sample size guidance to set thresholds that fit your volume.
Results Guide: Scoring, Internal Benchmarks, and Close-the-Loop Triggers
- Score the three metrics the same way every time: Goal: Turn raw answers into a weekly view your ops team can act on. Default: CSAT% + CES average, with NPS optional. Do this now: Write your scoring rules into the dashboard label (example: "CSAT (Top-2 box on 1-5)").
- CSAT scoring (pick one and stick to it): Use Top-box ("5" only) or Top-2 box ("4-5") divided by total responses, then multiply by 100. Example label: "CSAT Top-2% (4-5 on 1-5)."
- CES scoring (make friction visible): Report the average and a simple friction flag like % low-ease (example: % of 1-2 on a 1-5 ease scale). If you change the scale direction, rename it ("Ease" vs "Effort") so nobody misreads trends.
- NPS scoring (only if you run it): Promoters (9-10) minus detractors (0-6). Put the scale cutoffs in the chart subtitle so teams do not argue about categories.
- Benchmark internally, then segment: Set a baseline from your last 4-8 weeks (starter window), then trend by team, region, channel, and issue type. Tag each response with context (case type, time to close, transfer count) so your "why" analysis is fast; ACSI's customer satisfaction measurement overview is a good reminder to treat scores as a system signal, not a single-agent report card.
- Set close-the-loop triggers before you look at averages: Define triggers as internal starter thresholds, then tune them after you see your baseline. Example triggers: CSAT is 1-2 on a 1-5 scale, CES is low-ease, or they choose "Not resolved." Start with a simple outreach SLA target (starter target: contact within 1 business day), then adjust based on staffing and impact; AHRQ's service recovery program guidance maps well to a queue + SLA setup.
- Route with clear rules (example you can copy): If "Not resolved" AND permission-to-contact = Yes, assign to "Recovery" and notify the case owner. If CSAT is low AND no permission, log as "Recovery needed - cannot contact" and review weekly for process fixes (transfers, training gaps, scheduling errors).
Frequently Asked Questions
Should I use CSAT, CES, or NPS for a post-service survey?
Use CSAT to check how the last interaction felt, right after the ticket/visit. Add CES when you want to find friction you can fix (transfers, repeats, unclear steps). Use NPS only if you are running a separate loyalty pulse on a monthly/quarterly rhythm, not on every ticket.
How soon after the interaction should I send the survey?
Send it soon after the interaction so customers remember details. As starter timing targets, many teams send 15-60 minutes after ticket close, 1-3 hours after an on-site visit, or within 2-6 hours after delivery/installation/appointment completion, then adjust based on response patterns. To prevent over-surveying, start with simple caps (for example, 2 invites per customer per week for support and 1 per month for field visits) and add a cooldown after any response.
How many questions should a post-service survey include?
Keep most post-service surveys to 5-9 questions. Ask resolution first, then 2-4 driver questions (effort, speed, communication), then 1 open-text prompt. Avoid long grids so the survey stays easy on mobile.
How do I handle customers who say the issue was not resolved?
Branch immediately: ask what is still missing, then ask permission to contact and the preferred method/time. Treat this as a recovery workflow rather than a new support ticket; keep consent explicit and route the case to a queue with a clear outreach SLA.
What is a good CSAT/CES/NPS score for post-service?
Do not chase a universal benchmark. Set an internal baseline from a starter window like the last 4-8 weeks, then improve trends by channel, team, and issue type. Use rolling averages and minimum-n rules before ranking teams so you do not crown winners off small samples.
Should the survey be anonymous if I want to follow up?
Anonymous surveys can increase candor, but follow-up requires contact details. Use a hybrid setup: keep feedback anonymous by default, then ask permission to contact (Yes/No) and the preferred channel if they want a response. Do not collect extra personal data you will not use.
Related Survey Templates
FREE TO START -- NO CREDIT CARD REQUIRED