Key takeaways
- Six numbers tell you most of what you need: answer rate, missed calls, abandoned calls, time to answer, call-back time and the share of new-customer calls that become booked jobs.
- Check them weekly, in ten minutes, and compare each week with the same week last year rather than with last week.
- Heating demand is seasonal: household gas use was more than five times higher in the first quarter of 2026 than in the third quarter of 2025[1].
- Every funnel leaks. Even the Boiler Upgrade Scheme went from 160,874 applications to 96,611 paid installations by September 2026[2][3]. Measure where your calls drop out.
- Chaser calls ('where is my engineer?') are a hidden cost; track them separately and fix the cause.
Why a weekly check beats a monthly report
Most phone systems and call tracking tools produce a monthly report that nobody reads. By the time it arrives, the bad week that cost you three boiler installations is long gone. A heating firm needs a short weekly check instead: a handful of numbers, looked at every Monday, with one question attached to each. Is it getting better or worse, and what are we doing about it? Around 95% of UK plumbing and heating businesses are micro firms with fewer than ten employees[4], so the check has to be quick enough for an owner to do between jobs.
The six numbers
| Metric | How to work it out | What it tells you | If it slips |
|---|---|---|---|
| Answer rate | Calls answered by a person or service ÷ all inbound calls | How much demand you are actually catching | Look at when calls are missed before adding staff |
| Missed calls | Calls that rang out or went to voicemail | Lost opportunities, by hour of day | Add overflow cover for the worst hours |
| Abandoned calls | Callers who hung up while ringing or on hold | Patience running out | Shorten ring time before voicemail; offer call-backs |
| Time to answer | Average seconds before pick-up, by day | How the caller experiences you | Check ring groups and who is expected to answer |
| Call-back time | Time from voicemail or form to first reply | Whether messages are worth leaving | Name an owner for every message |
| Calls to booked jobs | New-customer calls that ended in a booking ÷ all new-customer calls | Whether answered calls turn into work | Listen to unbooked calls; check script and availability |
You do not need industry benchmarks to use these well. Your own trend is the benchmark. A firm whose answer rate drifts down three weeks running has a problem, whatever the absolute number.
Answer rate and missed calls
Answer rate is the headline, but on its own it hides the detail. Break missed calls down by hour and day. Many small firms find their misses cluster in a few predictable windows: first thing in the morning when everyone is loading vans, the lunch hour, and late afternoon when engineers are driving home. Those clusters are cheap to fix with a ring group change, a staggered lunch or overflow to an answering service, and fixing them often recovers more calls than adding a whole extra person.
Abandoned calls and time to answer
An abandoned call is someone who chose to give up. If your system plays a long greeting or a menu before ringing, callers may abandon before anyone has had a chance to answer. Check how long your phones ring before diverting to voicemail and how long any recorded message is. For urgent heating calls, a long menu is a quick way to send people to the next firm on the search results page.
Time to answer is best looked at by day rather than as a weekly average. Monday mornings after a cold weekend look very different from a quiet Thursday in June, and averaging them together hides both.
Call-back time
Voicemails and web forms only work if someone calls back quickly. Measure the time from message to first reply and look at the slowest few each week, not just the average. Where no time was agreed, consumer law expects services to be carried out within a reasonable time[5], and customers' sense of reasonable is shaped by what your greeting promised. If your voicemail says 'we'll call you back today', measure against that promise.
From calls to booked jobs
The last number matters most: what share of new-customer calls became a booked visit or a quote appointment? Ask whoever answers to tag each call with an outcome, then count. Every funnel loses people. Even a government grant with strong demand shows it: by the end of September 2026 there had been 160,874 Boiler Upgrade Scheme voucher applications[2], 131,619 vouchers issued[6] and 96,611 installations paid out[3]. If you install heat pumps, you will recognise the stages; if you do boiler work, the lesson is the same. Find the stage where people drop out and work on that one.
Source: Department for Energy Security and Net Zero, 2026[2]
Show the data
| Item | Value |
|---|---|
| Voucher applications | 160,874 |
| Vouchers issued | 131,619 |
| Installations paid | 96,611 |
Unbooked calls fall into a few groups: outside your area, not your type of work, no suitable slot, price concerns and undecided. The first two are marketing problems, the third is capacity, and the last two are often about how the call was handled. Listening to three unbooked calls a week tells you which.
Chaser calls: the hidden statistic
Tag calls from existing customers asking 'when is the engineer coming?', 'did you get my message?' or 'where is my quote?'. These chaser calls are not demand; they are the cost of poor communication, and they crowd out new enquiries at busy times. If chasers rise, look at confirmations, 'on my way' calls and quote turnaround, not at phone staffing.
Compare like with like
Heating work follows the weather. UK household gas demand was 109,126 GWh in the first quarter of 2026 against 19,636 GWh in the third quarter of 2025, more than five times higher[1], and demand in the second quarter of 2026 was up 10% on the same quarter a year earlier[7]. Call volumes for heating firms tend to follow the same shape. Comparing a November week with an October week will always look alarming; compare it with the same week last year, and note unusual weather alongside the figures.
Where the numbers come from
You probably already have most of this data, spread across three or four places. The trick is pulling it together in the same format every week.
- Your phone system or hosted telephony app: inbound calls, answered, missed, ring time and abandoned calls, usually as a CSV export or a built-in report.
- Call tracking, if you use it: calls by marketing source and first-time versus repeat callers.
- Voicemail and web forms: a count of messages and the time each was first answered, which may mean a quick look through the shared inbox.
- Your job-management or diary software: bookings made, and which ones came from new customers.
- Outcome tags: the short label whoever answered gave each new-customer call, such as booked, quote visit, outside area or spam.
If you cannot get a number easily, start without it. Four reliable numbers checked every week beat six numbers that need an hour of spreadsheet work and get abandoned by February.
Clean the data before you judge it
Raw call logs include calls that should not count. Sales calls, recorded marketing messages and wrong numbers inflate your inbound total and can make your answer rate look worse than it is, especially if staff rightly ignore numbers they recognise as nuisance callers. Internal calls between engineers and the office also slip in. Filter out known nuisance numbers and your own team's mobiles before calculating the six numbers, and count spam separately so you notice if it suddenly rises. A sudden jump in very short calls from unknown numbers is usually a nuisance campaign, not a surge in customers.
Share the numbers with the team
Call statistics change behaviour only when the people answering see them. Put the six numbers on one page and share it at a short Monday huddle or in the team chat. Keep the tone practical rather than critical: the point is to find patterns, not to blame whoever was on the phone. Engineers in particular often have no idea how many calls go unanswered while they are on site; seeing that a cluster of misses happens between 8am and 9am, when they are all at the merchant, makes the case for changing the routine better than any lecture. Celebrate improvements too. When the answer rate climbs back after a change, say so.
A ten-minute Monday routine
Export last week
Pull the call log from your phone system or tracking tool, plus form and voicemail counts.
Fill in six numbers
Answer rate, missed, abandoned, time to answer, call-back time and calls to booked jobs.
Compare
Put them next to the same week last year and the last four weeks.
Pick one problem
Choose the single worst change and agree one action, such as moving lunch breaks or changing the ring group.
Check last week's action
Did the change you made last Monday help? Keep it, adjust it or undo it.
Example: one week on one page
| Metric | This week | Same week last year | Note |
|---|---|---|---|
| Inbound calls | 212 | 184 | Cold snap from Wednesday |
| Answer rate | 81% | 88% | Misses between 8am and 9am |
| Abandoned calls | 14 | 9 | Long greeting added in September |
| Average time to answer | 19 seconds | 14 seconds | Monday and Thursday worst |
| Median call-back time | 3 hours | 2 hours | Two voicemails left overnight |
| New-customer calls booked | 58% | 61% | Several outside area |
Example: reading this fictional page, the owner sees that demand is up because of the weather, but early-morning misses and a longer greeting are costing calls. The one action for the week is to shorten the greeting and have the office answer from 7.45am. Next Monday will show whether it worked. For the bigger picture of when calls arrive across the year, see when plumbing and heating calls peak.
Frequently asked questions
What is a good answer rate for a heating firm?
Use your own trend as the benchmark. Track it weekly, break misses down by hour, and fix the worst windows first.
What is the difference between a missed and an abandoned call?
A missed call rang out or went to voicemail without anyone answering. An abandoned call is one where the caller hung up while waiting, often during a greeting, menu or hold.
Should I compare call numbers month to month?
Not for heating work, which is strongly seasonal. Compare each week with the same week last year and note the weather.
Where do I get these figures?
Most business phone systems and call tracking tools export a call log. Add web form and voicemail counts, and outcome tags from whoever answers.
References
Department for Energy Security and Net Zero (2026). Energy Trends: UK gas, ET 4.1 natural gas supply and consumption (quarterly), September 2026 (opens in a new tab) ↑
Department for Energy Security and Net Zero (2026). Boiler Upgrade Scheme (BUS) monthly statistics: England and Wales, September 2026 (opens in a new tab) ↑
Department for Energy Security and Net Zero (2026). Boiler Upgrade Scheme (BUS) monthly statistics: England and Wales, September 2026 (opens in a new tab) ↑
Office for National Statistics (Nomis) (2026). UK Business Counts - enterprises by industry and employment size band (SIC 2007 class 43.22), via Nomis (opens in a new tab) ↑
legislation.gov.uk (The National Archives) (2015). Consumer Rights Act 2015, section 52 (opens in a new tab) ↑
Department for Energy Security and Net Zero (2026). Boiler Upgrade Scheme (BUS) monthly statistics: England and Wales, September 2026 (opens in a new tab) ↑
Department for Energy Security and Net Zero (2026). Energy Trends: UK gas, ET 4.1 natural gas supply and consumption (quarterly), September 2026 (opens in a new tab) ↑