They split the job. An instant-price booking form handles standard domestic cleans from bedrooms, bathrooms and extras, while AI reads the messy enquiries by text, email or photo, asks what's missing and books a slot. Regular cleans can go from enquiry to booked in under five minutes; move-out and commercial jobs still need a look first.
Most of the "instant" part isn't AI at all. A booking-form price is a rules calculation: so much per bedroom, so much per bathroom, plus extras. AI earns its keep on the enquiries that never fit the form, like "can u do my 3 bed fri, got 2 dogs", which in many cleaning firms are most of them. The mistake that costs money is letting the AI price freehand instead of from your formula.
Three kinds of cleaning enquiry, three different speeds
| Enquiry | How it's priced | Can it be instant? | Realistic time to a booked slot |
|---|---|---|---|
| Regular domestic (weekly, fortnightly) | Formula: rooms, bathrooms, extras, frequency | Yes, with a first-clean uplift | 2-5 minutes via the form; 10-20 minutes via AI-drafted reply |
| One-off deep clean or move-out (end-of-tenancy) | Formula as a starting point, adjusted from photos | A range, yes; a firm price, only with photos | Same day, once photos arrive |
| Office and commercial | Hours from a walk-round or detailed floor plan | No | Days: a visit, then a written proposal |
Designing for the first row and pretending the other two are the same is how fast quoting loses money. The questions that make each type quotable are in the intake questions that make AI cleaning quotes accurate.
What the instant price is actually calculating
Every booking form with instant pricing runs a formula you set. Here's an illustrative one for regular domestic cleans, with the time estimate alongside so the calendar knows how long to block:
Price per visit = $60 base + $18 per bedroom + $22 per bathroom
Time per visit = 1.5 h + 0.4 h per bedroom + 0.6 h per bathroom
First clean = 1.5 x regular price, 1.5 x regular time
Extras = oven $45 (+1 h), inside fridge $25 (+0.5 h),
interior windows $30 (+0.75 h)
Pets that shed heavily = +$15 per visit (+0.3 h)A three-bedroom, two-bathroom house comes out at $158 a visit and 3.9 hours, with a first clean of $237. The numbers are yours to change, but the structure matters more than the figures: a price and a duration for every line, and a first-clean multiplier, because the first visit in a house nobody has cleaned properly takes longer than the tenth.
Booking software handles this without any AI. BookingKoala's plans start at $27 a month (Growing is $57, Premium $197) with customisable booking forms that show customers your prices and availability. ZenMaid starts at $19 a month, and its Pro plan at $39 adds website booking forms, with custom branding on Pro Max at $49. Check per-user charges for your team size on each pricing page.
AI's real job: the enquiry nobody filled in properly
Customers message, email and leave voicemails. A chat assistant is good at turning those into the fields your formula needs, and at spotting what's missing. A prompt you can adapt, with your price formula pasted in:
You work for a domestic cleaning company. Read the customer
message below and return:
1. Service type (regular / one-off deep / move-out / other)
2. Bedrooms, bathrooms, extras, pets, frequency, preferred days
3. Price and time using ONLY the formula below. Show the working.
4. Anything missing that we need before booking, as short
friendly questions (max 3).
5. A draft reply to the customer, under 90 words.
If it's commercial, or you can't tell the size, don't price it:
say it needs a quick call or photos.
FORMULA: [paste]
MESSAGE: [paste]Given this message, "hiya can someone do my house fri or sat, 3 bed semi 2 bathrooms one ensuite, got 2 dogs (shed a lot lol). need oven doing too. how much weekly after that", an illustrative output:
Service: first clean, then weekly
Bedrooms 3, bathrooms 2, extras: oven, pets: 2 dogs (heavy shedding)
Preferred: Friday or Saturday
First clean: $237 + oven $45 + pets $15 = $297 (about 7.2 h)
Weekly after: $158 + pets $15 = $173 (about 4.2 h)
Missing: street and area for travel planning; will someone be
home or do we need a key; any parking on the street?
Draft: Hi, thanks for getting in touch! For your 3-bed with two
bathrooms and the dogs, the first clean including the oven is
$297, then $173 a week. Could you let me know your street and
area, whether you'll be home or we'll need a key, and if there's
parking nearby? Then I can confirm Friday or Saturday for you.Check the working before trusting it. The first-clean time should be 3.9 hours × 1.5 (5.85), plus 1 hour for the oven and 0.3 for the pets, which is 7.15, so "about 7.2 h" is right. That's a full day for one cleaner or a morning for two, which the draft doesn't mention and the customer will want to know. Add that to the reply and it's ready to send.
Deep cleans and move-outs: pricing from photos
For one-off cleans, the formula gives a starting point and photos decide where in the range the job sits. Ask for four photos: the oven open, the inside of the fridge, the worst bathroom, and the kitchen floor. Then:
This is a one-off move-out clean. Using the formula as the base,
look at the photos and place the job at base, base + 25% or
base + 50%, based on grease, limescale, floor condition and
clutter. Say what you saw that justifies it. If the property
isn't empty or there's mould, say "needs a call" and don't price.
FORMULA AND BASE PRICE: [paste] PHOTOS: [attach]Illustrative output: "Base + 25%. The oven has heavy baked-on grease on the door glass and racks; the shower screen shows thick limescale; the kitchen floor looks sound; the flat appears empty. Suggested price: $403 (base $322), about 9 hours for one cleaner, or a morning for a team of two." What you'd fix: the base is wrong. On this grid a move-out for a three-bed, two-bath home is the first-clean price plus oven and fridge, $237 + $45 + $25 = $307, so base + 25% is $384, not $403. The assistant had quietly added the $15 pet surcharge from an earlier message in the same chat, which is a good reason to start each quote in a fresh conversation. And "appears empty" needs confirming, because a flat with the tenant's belongings still in it is a slower job. The photos turn a guess into a reason, which is also what you show the customer if they question the price.
Booking the slot without wrecking the schedule
A fast quote followed by a slot you can't staff is worse than a slow quote. Whatever books the job, whether it's the customer through a form, an AI receptionist or your coordinator, it needs to book from a calendar that tells the truth:
- Durations from the formula, so a 7-hour first clean doesn't land in a 3-hour gap.
- Travel buffers between jobs, set by area. Twenty minutes between two houses on the same estate is not the same as forty across town.
- Cleaner holidays and sickness blocked out, or the system will cheerfully book someone who's away.
- Skills and restrictions: who does ovens, who's fine with dogs, who has a car for the outlying jobs.
- Recurring slots held. Regular customers expect the same day and, ideally, the same cleaner.
Some field-service platforms now answer and book by phone and text. Jobber's Receptionist, for example, can answer calls and texts around the clock and book jobs into your Jobber calendar, as a $29-a-month add-on for 30 conversations (then $0.79 each), or included on its Plus plan. Whether a phone agent or a website chatbot suits a cleaning firm better is compared in AI receptionist or chatbot for a cleaning company.
Commercial enquiries: collect the details, don't price
An office enquiry such as "can you quote for our office, about 450 square metres, cleaning five evenings a week?" has no honest instant price, because the hours depend on layout, kitchens, toilets, floor types and what "clean" means to the client. What AI can do in minutes is send the right questions and book the walk-round. An illustrative draft:
Thanks for getting in touch. To quote accurately we'll do a 20-minute walk-round; before that, it helps to know: how many toilets and kitchen areas are there; is it mainly carpet or hard floors; how many desks; do you need bins, dishwashers or meeting rooms reset each evening; and are there any areas we shouldn't enter? We can visit on Tuesday at 17:30 or Wednesday at 8:00.
Replying within ten minutes with sensible questions often wins the walk-round before competitors have replied at all. After the visit, the same assistant can turn your walk-round notes into a proposal skeleton: scope by area, frequency, hours per visit and a monthly price you calculate yourself from your hourly rate.
One domestic cleaning firm, before and after
Consider a hypothetical firm with eight cleaners, the owner and a part-time coordinator; its figures are invented but realistic in scale. It gets about 45 enquiries a week: 25 for regular domestic cleans, 14 for deep or move-out cleans and 6 commercial or odd jobs.
Before. The owner answered enquiries in the evening. Median time to a first reply was five hours, the quote took another round of messages, and 32% of enquiries booked, about 14 jobs a week. The coordinator spent around nine hours a week on enquiries and scheduling back-and-forth.
The change, over four weeks.
- Week 1: the price formula written down, with durations, and loaded into a booking form on the website.
- Week 2: calendar cleaned up with holidays, buffers and skills.
- Week 3: the triage prompt above set up in a shared chat assistant project, with the formula attached; the coordinator pastes in every message that isn't a form booking and sends the checked draft.
- Week 4: an automatic acknowledgement on every channel within a minute ("thanks, you'll have a price within 20 minutes during office hours").
After. Twelve of the 25 regular-clean enquiries now book themselves through the form. The rest get a priced, checked reply in about 20 minutes during office hours. The share of enquiries booking rose to about 40%, roughly 18 jobs a week, and the coordinator's enquiry time fell to about four hours. The deep cleans still wait for photos, but now the photo request goes out in minutes rather than the next evening.
The gain came mostly from speed. People asking for a cleaner often message several firms at once, and the first sensible reply with a price has the advantage.
Where fast cleaning quotes lose money
- "Three-bed" means different things. A three-bed flat and a three-bed house with a loft conversion and a utility room are different jobs. Ask for photos or a floor area on anything above your typical size.
- The first clean is under-priced. Without a multiplier, the first visit runs two hours over and the cleaner's next job starts late.
- Extras get forgotten. Ovens, inside cupboards, blinds and limescale are where time goes. Make them explicit options, never "we'll see on the day".
- Pets, parking and access. A 20-minute hunt for parking or a key safe that doesn't work comes out of your margin.
A realistic mistake: one firm's AI-drafted reply quoted a regular weekly price for a move-out clean because the customer wrote "clean before I leave on Friday". The cleaner arrived to an empty flat with a dirty oven and marks on every wall, and a four-hour booking took nine. The fix was one line in the prompt: any mention of moving, leaving, keys being handed back or a landlord means "move-out, photos required, no price yet".
A good habit that protects regular pricing: tell every new regular customer that the price is reviewed after the first two visits. After the second clean, compare actual time with the formula and adjust, with a short, polite message if the price needs to change.
When a regular customer changes frequency
Customers switching from weekly to fortnightly expect the same price per visit, but a fortnightly clean takes longer because two weeks of dust and limescale build up. Put the rule in the formula, for example fortnightly at 1.15 × the weekly price and time, and let the assistant draft the reply. For the $158 house, an illustrative draft: "Happy to move you to fortnightly. Because there's more build-up between visits, fortnightly cleans take a little longer, so the price would be $182 a visit. Your first fortnightly visit would be on the 21st." The number came from the rule, not from the AI.
The messages that make booking feel instant
Speed is also about what the customer hears at each step. Three short templates cover most of it:
Booked: Thanks, you're confirmed for Friday 14th, 9:00-13:00, with two of our team. We'll text the morning before with their names. Reply here if anything changes.
Day before: Hi, just a reminder that we're with you tomorrow at 9:00. Is access still by the key safe? Anything you'd like us to focus on?
After the second visit: Thanks for having us twice now. The house is taking about 3.5 hours a visit, which is a little under our estimate, so your weekly price stays at $158. We'll keep an eye on it.
An assistant can draft these from the booking details, but they work best as fixed templates with the details merged in, so every customer gets the same accurate message. Once customers are booked, the next revenue question is keeping them, which helping a cleaning company rebook more regular clients covers.
Numbers to watch once it's running
- Median time to first reply, by channel. Under five minutes for an acknowledgement, under 30 for a price in office hours, is a reasonable target.
- Enquiry-to-booking rate, split by service type.
- First-clean overrun: actual hours minus formula hours. If it's regularly positive, raise the multiplier.
- Price reviews after visit two: how often the price had to change. More than one in five means the formula or the questions need work.
When the form price and the AI's price disagree
If a customer gets $158 from the website form and $173 in a message from you, you have a trust problem. It usually comes from one of three places: the formula in the booking software and the one pasted into the assistant have drifted apart after a price change; the assistant applied a modifier, such as pets, that the form doesn't ask about; or the assistant rounded differently. The fix is one source of truth. Change prices in the booking software first, paste the same formula into the assistant's project the same day, and add any modifiers you use in messages to the form as questions. Then test both with the same three sample houses after every change.
If replies are still slow because the enquiries arrive while everyone is on a job, the approach in replying to trade enquiries in 60 seconds with AI applies to cleaning firms just as well.
Cleaning owners' questions about fast quoting
Should the booking form show the price before the customer gives contact details?
For regular domestic cleans, showing the price first usually builds trust and cuts back-and-forth, and you can still collect details at the booking step. For deep cleans and move-out cleans, ask for photos and contact details first and send the price afterwards, because the form price is only a starting point for those jobs.
Can AI answer enquiries on WhatsApp for a cleaning business?
Yes, within limits. Meta's WhatsApp platform terms allow bots tied to a specific business task such as bookings, but not general-purpose assistants. Meta's own Business Agent is charged per token, which Meta puts at roughly 4-5 cents a message. Many small firms use the free WhatsApp Business app with quick replies instead and paste messages into an assistant for drafts.
What if the AI books a slot we cannot actually staff?
That happens when the calendar shows availability that does not exist, such as a cleaner on holiday who was never blocked out. Fix the calendar first: holidays, travel time and cleaner skills must be in the system the AI books from. Until you trust it, let the AI offer two slots and have a person confirm.
Further reads
- Photo Checklists and Quality Control for Cleaning Teams Using AI — Photo checklists that catch problems before customers do.
- AI Job Scheduling and Dispatch for Trades and Field Service Teams — Scheduling and dispatch once bookings arrive faster.
- Jobber, Housecall Pro, and ServiceTitan AI Features Compared — What the big field-service platforms' AI features do.
- How to Add an AI Chatbot That Captures Leads on Your Website — Add a website chatbot that captures enquiries properly.
- Getting More Reviews After Every Job With AI Follow-Ups — Ask for a review after the first clean goes well.
- AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.
Sources: BookingKoala pricing page; ZenMaid pricing page; Jobber help article on Receptionist powered by Jobber AI; Meta Business Agent and WhatsApp platform terms. Checked September 2026.