Key takeaways
- Pick one simple measure, ask every customer the same way after every job, and track the trend rather than chasing a benchmark.
- Problems are common in home repair: 11% of people who bought home maintenance and repair services in a year had a consumer problem[1].
- NPS is promoters (9-10) minus detractors (0-6); with small samples, read the comments before the number.
- Call back every unhappy customer within two working days; that is where the value of a survey lies.
- Keep surveys free of offers, and never reward customers for positive reviews[2].
Why measure satisfaction at all
Most plumbing and heating firms only hear from customers when something goes badly wrong, or when someone leaves a review. Everything in between is invisible: the customer who was annoyed that the engineer arrived late but did not complain, or the one who loved the tidy job but never thought to say so. A short survey after each job closes that gap and gives you early warning before a pattern turns into complaints.
The stakes are real. In the government's consumer protection study, 25% of UK adults had bought home and garden maintenance and repair services, which include plumbing and central heating, in the previous 12 months, and 11% of those buyers had a consumer problem[1]. The median net monetised detriment was £109 per problem, among the ten highest of any sector in the study[1]. Problems you hear about early are cheaper to fix.
Three ways to measure, and when each fits
All of these measures work if you ask consistently, and none of them works if you ask only after the jobs that went well. The choice comes down to what you want to learn. If referrals and repeat work drive your diary, the recommend question in NPS fits. If you want to coach engineers visit by visit, a satisfaction score per job is more useful. If the problem is the paperwork and the phone calls around the job, ask how easy it was to get things sorted.
| Measure | The question | Score | Best for | Watch out for |
|---|---|---|---|---|
| Net Promoter Score (NPS) | How likely are you to recommend us to a friend or neighbour? | 0-10, reported as promoters minus detractors | Tracking loyalty over months; firms that rely on referrals | Needs a steady flow of responses to be stable |
| Customer satisfaction (CSAT) | How satisfied were you with today's visit? | 1-5, reported as % scoring 4 or 5 | Judging individual visits and engineers | Scores cluster high; read the comments |
| Would you use us again? | Would you call us again next time? | Yes / not sure / no | Sole traders who want one simple signal | Gives little detail without a follow-up question |
| Customer effort | How easy was it to get this sorted? | 1-5 | Testing booking, quoting and invoicing | Less useful for judging workmanship |
Whichever you pick, always add one open question: 'What is the main reason for your score?' The number tells you whether something is wrong; the reason tells you what to fix. Most small firms find that the comments are where nearly all the useful information sits.
How NPS is calculated
NPS sorts each answer to the 0-10 recommend question into three groups. Customers scoring 9 or 10 are promoters, 7 or 8 are passives and 0 to 6 are detractors. The score is the percentage of promoters minus the percentage of detractors, so it runs from minus 100 to plus 100. Passives count towards the total but not towards either group.
- Count the responses for the period, for example a calendar month.
- Count promoters and detractors separately.
- Work out each as a share of all responses.
- Subtract the detractor share from the promoter share.
Example: in a month, a firm gets 40 responses: 28 promoters, 8 passives and 4 detractors. Promoters are 70% and detractors 10%, so the NPS is 60. If two of those detractors had been passives instead, the score would rise to 65. With samples this size, a handful of customers moves the number, so look at the trend over several months.
Simpler options for small firms
A sole trader doing a few jobs a day does not need a survey platform. A single text the morning after the job, asking for a score and a reason, works well. What matters is asking every customer in the same way, so the results are comparable. Pick one measure and stick with it for at least six months before changing anything.
- Text version: 'Thanks for having us yesterday. On a scale of 0 to 10, how likely are you to recommend us? Reply with a number and anything we could have done better.'
- Email version: one question with clickable scores, then an optional comment box. No login, no second page.
- Phone version: for larger installs, the office calls a week later, asks the same question and notes the answer word for word.
- Engineer-level CSAT: add 'How was your engineer today?' with a 1-5 score if you want to coach individuals.
When and how to ask
Ask soon after the job, while the visit is fresh: the same evening or the next morning for repairs, and a week or two after an install, once the customer has lived with the new system. Do not ask while the engineer is still in the house; customers feel obliged to be polite. Survey every customer, including the awkward jobs. Surveying only the happy ones tells you nothing and makes the score meaningless.
Keep surveys separate from marketing. Direct-marketing emails and texts to individuals need prior consent unless the soft opt-in for existing customers applies[3]. A message that asks for feedback and then promotes a boiler service offer starts to look like marketing, so leave offers out.
Benchmarks: useful context, poor targets
There is no reliable public NPS benchmark for UK plumbing and heating firms. The nearest national measure is the UK Customer Satisfaction Index, which covers organisations across the economy rather than trades. In July 2026 it stood at 78.3 out of 100, one point higher than a year earlier, with banks and building societies the highest-rated sector at 82.0[4].
Source: The Institute of Customer Service, 2026[4]
Show the data
| Item | Value |
|---|---|
| UKCSI, July 2025 | 77.3 points out of 100 |
| UKCSI, July 2026 | 78.3 points out of 100 |
| Highest sector, July 2026 (banks and building societies) | 82 points out of 100 |
The same release noted that good value, competitive pricing, quality and being easy to deal with had become more important to many customers[4]. Those are the themes to watch in your own comments. Your own trend, month on month, tells you far more than a cross-sector index.
Doing something with the results
Read every comment weekly
Someone senior reads every response, not only the low scores. Praise for a named engineer gets passed on the same week.
Call back every detractor
Phone anyone who scores 0 to 6 within two working days. Listen, apologise for the experience where it is fair to, and agree what you will do.
Tag the theme
Label each low score or critical comment with one theme: arrival time, communication, mess, price clarity, workmanship or invoicing.
Fix one thing a month
Pick the theme with the most tags and change a process. Our guide to turning reviews and complaints into improvements sets out a monthly routine.
Report the trend
Share the monthly score and the top three themes with the whole team. People improve what they can see.
Rules to follow when surveys become reviews
If you publish survey comments on your website as reviews, consumer protection rules on reviews apply. Every trader that publishes consumer reviews is responsible for preventing and removing fake reviews, concealed incentivised reviews and false or misleading review information[5]. Under the Digital Markets, Competition and Consumers Act 2024, commissioning a fake review, or one that hides an incentive, is treated as unfair in all circumstances[2]. Do not offer discounts or prize draws for good scores, publish the critical comments alongside the good ones, and keep the original responses so you can show where each published review came from.
Mistakes that skew the score
- Asking only after installs, which tend to score higher than emergency repairs where the customer was already stressed.
- Letting engineers choose which customers receive the survey.
- Changing the question wording halfway through the year, so this month cannot be compared with last.
- Reporting a single month's score from a dozen responses as if it were a trend.
- Treating a passive score of 7 or 8 as a success; passives are often one bad visit away from trying another firm.
- Collecting scores but never calling anyone back, so customers learn that the survey changes nothing.
Example: six months of scores
Example: a fictional four-engineer firm in Glasgow starts texting the NPS question after every job. Month one gives an NPS of 41, with most detractor comments about not knowing when the engineer would arrive. The firm starts sending a 'leaving now' message from the van and narrows its arrival windows. By month four the arrival comments have almost gone and the score sits in the mid-50s. The next theme is invoices that arrive a week late, so the office moves to same-day invoicing. The number matters less than the list of fixes it produced.
Frequently asked questions
What is a good NPS for a plumbing firm?
There is no reliable public NPS benchmark for UK plumbing and heating firms. Treat your own score over time as the benchmark, and focus on the comments from customers who score 0 to 6.
How many responses do I need for NPS to mean anything?
With fewer than about 30 responses in a period, a few customers can swing the score a long way. Combine months, or report a rolling three-month score, and always read the comments.
Can I offer a discount for completing a survey?
An incentive to complete a private survey is different from paying for reviews, but keep it clear of any review you publish. Commissioning a review that hides an incentive is banned under the DMCC Act 2024[2].
Is a feedback text classed as marketing?
A plain request for feedback about a job you have done is generally a service message, but adding offers or promotions can turn it into direct marketing, which needs consent unless the soft opt-in applies[3].
References
Department for Business, Energy and Industrial Strategy (GOV.UK) (2022). Consumer protection study 2022: understanding the impacts and resolution of consumer problems (opens in a new tab) ↑
legislation.gov.uk (The National Archives) (2024). Digital Markets, Competition and Consumers Act 2024, Schedule 20 (opens in a new tab) ↑
legislation.gov.uk (The National Archives) (2003). The Privacy and Electronic Communications (EC Directive) Regulations 2003, regulation 22 (opens in a new tab) ↑
The Institute of Customer Service (2026). The UK Customer Satisfaction Index (UKCSI): July 2026 key findings (opens in a new tab) ↑
Competition and Markets Authority (GOV.UK) (2025). Short guide for businesses: publishing consumer reviews and complying with consumer protection law (opens in a new tab) ↑