Research · First look
The out-of-hours gap
What happens to UK solar and home-improvement leads when the office is shut, and where AI can help. A first look at 4,885 enquiries, 1,208 called leads and one retailer’s WhatsApp assistant.
55%
of solar and home-improvement enquiries came in outside weekday working hours (8am–6pm)
2×
solar leads called within the hour booked nearly twice as often as those called later the same day
37%
of homeowners’ WhatsApp messages to a retailer’s AI assistant arrived while the showroom was shut
01
Most homeowners enquire when the office is shut
We looked at 4,885 enquiries sold to UK installers: 4,331 for solar between 10 August and 22 October 2026, and 554 for windows and conservatories between 2 September and 22 October.¹ 55% came in outside weekday working hours, including 30% at the weekend and 18% on weekday evenings.
Solar and home improvement are different markets, with different buyers, but here they look almost the same: 55% of solar enquiries and 54% of home-improvement enquiries came in out of hours.
A homeowner who fills in a form at 8pm is on their phone, thinking about the job, and has often asked other firms too. By the time the office opens, the moment has usually passed.
When solar and home-improvement enquiries came in
Top: each market’s split. Below: both markets together, by day and hour, with weekday office hours (8am–6pm) outlined. UK time.
Solar · 4,331
55% out of hours
Home improvement · 554
54% out of hours
- Office hours
- Weekends, weekday evenings and before 8am
Source: POCKLA, 4,331 solar enquiries (10 Aug–22 Oct 2026) and 554 window and conservatory enquiries (2 Sep–22 Oct 2026).
Show the numbers
| Market | Enquiries | Out of hours | Share |
|---|---|---|---|
| Solar | 4331 | 2384 | 55% |
| Home improvement | 554 | 299 | 54% |
| Both | 4885 | 2683 | 55% |
| Hour (UK), both markets | Mon | Tue | Wed | Thu | Fri | Sat | Sun |
|---|---|---|---|---|---|---|---|
| 12am | 2 | 10 | 4 | 4 | 2 | 6 | 6 |
| 1am | 1 | 2 | 3 | 1 | 2 | 1 | 1 |
| 2am | 1 | 2 | 2 | 5 | 0 | 2 | 2 |
| 3am | 0 | 0 | 1 | 2 | 3 | 1 | 0 |
| 4am | 0 | 2 | 2 | 1 | 2 | 1 | 1 |
| 5am | 4 | 3 | 10 | 3 | 3 | 7 | 5 |
| 6am | 13 | 14 | 12 | 20 | 16 | 24 | 14 |
| 7am | 20 | 34 | 45 | 71 | 46 | 69 | 52 |
| 8am | 39 | 35 | 26 | 56 | 28 | 58 | 51 |
| 9am | 35 | 36 | 41 | 49 | 31 | 48 | 37 |
| 10am | 17 | 24 | 33 | 51 | 38 | 51 | 28 |
| 11am | 27 | 35 | 55 | 55 | 36 | 51 | 37 |
| 12pm | 35 | 37 | 47 | 37 | 36 | 54 | 31 |
| 1pm | 46 | 34 | 58 | 54 | 25 | 54 | 33 |
| 2pm | 36 | 32 | 44 | 57 | 44 | 63 | 35 |
| 3pm | 40 | 49 | 58 | 56 | 31 | 79 | 40 |
| 4pm | 45 | 52 | 72 | 80 | 63 | 72 | 35 |
| 5pm | 48 | 55 | 64 | 66 | 54 | 62 | 37 |
| 6pm | 43 | 35 | 53 | 49 | 32 | 23 | 18 |
| 7pm | 41 | 45 | 60 | 41 | 34 | 23 | 21 |
| 8pm | 31 | 29 | 33 | 28 | 16 | 27 | 11 |
| 9pm | 38 | 23 | 20 | 25 | 32 | 40 | 71 |
| 10pm | 29 | 13 | 28 | 17 | 16 | 25 | 18 |
| 11pm | 8 | 12 | 15 | 9 | 17 | 11 | 7 |
02
Then the typical solar lead waits almost two hours
We can only time calls for solar.² Of the 1,208 solar leads that installers called, half waited more than 109 minutes for the first call. 27% were called within the hour, and 15% waited a day or more.
Time from a solar lead arriving to its first call
Purple: called within the hour.
- Within the hour (27%)
- Later
Source: POCKLA, 1,208 called solar leads, 10 Aug–22 Oct 2026.
Show the numbers
| First call after the lead arrived | Leads | Share |
|---|---|---|
| Under 5 min | 73 | 6% |
| 5–15 min | 115 | 10% |
| 15–30 min | 88 | 7% |
| 30–60 min | 53 | 4% |
| 1–2 hours | 289 | 24% |
| 2–4 hours | 90 | 7% |
| 4–8 hours | 47 | 4% |
| 8–12 hours | 74 | 6% |
| 12–24 hours | 201 | 17% |
| 1–2 days | 68 | 6% |
| 2–3 days | 20 | 2% |
| 3+ days | 90 | 7% |
03
Every hour of waiting cost appointments
Some teams are simply faster than others, and they may also be better at selling. So we never compare one team with another: we compare fast and slow first calls on the same team’s leads.
In team A, leads first called later the same day were nearly as likely to be reached as those called within the hour (69% against 74%), but only about half as likely to book (15% against 29%). Leads first called the next day still answered more than half the time, yet almost none booked. In team B, both fell: fewer homeowners answered, and fewer booked.
Both teams’ differences are larger than chance would explain at these sample sizes. They show a link, not proof: a team may call its most promising leads first.
Share of solar leads that booked an appointment, by time to first call
Grey bars are shown but not part of the finding. Each team records appointments its own way, so the two are never combined.
Team A
552 leads · Sep–Oct 2026
Team B
656 leads · Aug–Oct 2026
- Part of the finding
- Shown, not part of the finding
Source: POCKLA, 1,208 called solar leads in two calling teams, Aug–Oct 2026.
Show the numbers
| Team | First call after the lead arrived | Leads | Ever reached | Booked an appointment |
|---|---|---|---|---|
| Team A | Under 5 min (not yet explained: see below) | 66 | 61% | 11% |
| Team A | 5–60 min | 238 | 74% | 29% |
| Team A | 1–24 hours | 168 | 69% | 15% |
| Team A | Next day or later | 80 | 54% | 3% |
| Team B | Within the hour (only 25 leads) | 25 | 48% | 32% |
| Team B | 1–4 hours | 313 | 46% | 22% |
| Team B | 4–24 hours | 220 | 34% | 15% |
| Team B | Next day or later | 98 | 24% | 9% |
Open question
What we can’t explain yet
In team A, solar leads first called within five minutes booked less often (11%) than leads called 5–60 minutes after arriving (29%). Most of those very fast first calls went unanswered: 12% reached someone on the first dial, against 42% for calls made 5–60 minutes in.
One explanation is that the homeowner is still on the website, or isn’t expecting a call yet. It’s 66 leads from one team, so we’re not drawing a conclusion. We’ll look again in the next update.
04
Where AI can help: answering when you’re shut
A UK flooring retailer uses an AI assistant on WhatsApp to answer homeowners’ questions and book calls with its flooring team, with a person stepping in when needed.³ 216 homeowners replied to it, and 37% of their 576 messages arrived while the showroom was shut, most between 5pm and 9pm on a weekday.
The assistant can answer at any hour. It booked 17 calls, 7 of them while the showroom was shut, and took 36 quote requests. A person took over in about 1 in 10 conversations.
That fits what we saw in solar. The very fastest calls often went unanswered, while calls in the first hour booked most often. A reply that answers the question straight away and lets the homeowner pick a time to talk covers the gap without someone by the phone at 9pm.
Homeowners’ WhatsApp messages to a retailer’s AI assistant, by hour of day
Opening hours taken as weekdays 9–5 and Saturday 9–1. UK time.
- While shut (37%)
- While open
Source: POCKLA, 576 messages from 216 homeowners to one UK flooring retailer, 6 Aug–22 Oct 2026.
Show the numbers
| Hour (UK) | While shut | While open |
|---|---|---|
| 12am | 2 | 0 |
| 1am | 0 | 0 |
| 2am | 1 | 0 |
| 3am | 0 | 0 |
| 4am | 0 | 0 |
| 5am | 0 | 0 |
| 6am | 3 | 0 |
| 7am | 11 | 0 |
| 8am | 14 | 0 |
| 9am | 0 | 35 |
| 10am | 2 | 84 |
| 11am | 5 | 29 |
| 12pm | 1 | 30 |
| 1pm | 1 | 70 |
| 2pm | 0 | 49 |
| 3pm | 4 | 31 |
| 4pm | 1 | 35 |
| 5pm | 25 | 0 |
| 6pm | 40 | 0 |
| 7pm | 52 | 0 |
| 8pm | 21 | 0 |
| 9pm | 12 | 0 |
| 10pm | 11 | 0 |
| 11pm | 7 | 0 |
05
What this means for installers
01
Cover the first hour, evenings and weekends included
More than half of enquiries arrive when the office is shut, in solar and home improvement alike, and the first hour is where appointments are won.
02
Let AI take the first reply
An assistant can answer the question and offer a time to talk at 9pm. A person then calls at the time the homeowner picked.
03
Measure your own numbers
Time from lead to first dial, split by when the lead arrived, tells you more than any average.
POCKLA sends every lead it sells a WhatsApp or SMS in the installer’s name within seconds, and agrees a call time the homeowner chose. Making that call on time is still the installer’s job, and this study shows why it matters. How our WhatsApp follow-up works.
06
How we did it
- Terms
- Solar means solar panel enquiries. Home improvement means windows and conservatories and, for the WhatsApp figures, flooring. The industry treats them as separate markets, so we show each market’s figures beside the total.
- Enquiries
- 4,331 solar enquiries (10 August–22 October 2026) and 554 window and conservatory enquiries (2 September–22 October 2026), all sold to UK installers and timed by when the enquiry reached us, in UK time.
- Called leads
- The 1,208 solar leads with at least one recorded call after delivery. Time to first call is the first recorded dial minus the moment the lead was delivered. “Booked” means the installer’s call record shows an appointment booked.
- Two calling teams
- Team A logs its calls in our call workspace (September–October). Team B’s dialler sends us its call outcomes (August–October). They record appointments differently, so speed is compared only within each team.
- One UK flooring retailer’s WhatsApp line, 6 August to 22 October 2026, answered by an AI assistant with a person stepping in when needed. We counted messages, bookings, quote requests and hand-offs to a person. We did not read message content.
- Left out
- 277 of team B’s leads whose first dial was recorded before their delivery time (a clock mismatch between the two systems), 8 of team A’s leads delivered before it started logging calls, test and practice leads, and heat pump and insulation enquiries, which we’ll cover separately.
- Privacy
- Enquiry and call figures are counts produced inside the source database; WhatsApp figures come from timestamps and labels only. No names, phone numbers, call notes or message text were taken out, every business is anonymised, and our contracts allow this analysis.
- Limits
- UK only, about ten weeks. Call timings are for solar only, from two calling teams, and the WhatsApp figures come from one retailer. The figures show links, not causes.
Cite this study
POCKLA (2026). The out-of-hours gap: UK solar and home-improvement leads, August–October 2026. pockla.io/research/out-of-hours-gap
More research