Pressure Washing Leads: Reply to Quote Requests Faster With AI

Reply within minutes with a message that confirms the request, gives a price range and asks for two photos, then follow up on a set schedule. Replying fast usually costs less than buying more leads, but prove it with your own sum: leads times close rate times profit per job, before and after the change.

Most pages on this topic are written by people who sell leads, so the sums here use your numbers, not a vendor's. They also change the question. It stops being where to find more leads and becomes how much of each lead's value leaks away in the first hour.

Follow me on Instagram@sagnikteaches

What one lead is worth, and what speed adds

Take an illustrative firm that gets 40 quote requests a month and books 20 per cent of them, which is eight jobs. The average job is $320 and the margin is 55 per cent after fuel, chemicals and crew pay, so each job leaves $176. Eight jobs leave $1,408 a month. Each lead is therefore worth 0.20 times $176, or $35.20. Pay more than that for a lead and, at that close rate, it loses money.

Connect on LinkedInSagnik Bhattacharya

Now compare three ways to add profit, using the same firm.

Subscribe on YouTube@sagnikteaches

Two more sums sharpen it. The cost per booked job from a bought lead is the lead price divided by the close rate: $30 divided by 0.20 is $150, but at 12 per cent it is $250, which is more than the $176 a job leaves. And the lift a reply tool needs to pay for itself is tiny. At $29 a month against 40 leads and $176 a job, the close rate only has to rise by about 0.4 of a percentage point, which is roughly one extra booked job every six months. The 25 per cent figure in the table is an assumption, not a promise. The next section shows how to find your own.

Measure your own reply-time effect with a 30-day sheet

Nobody can give you a trustworthy lift from reply speed, so measure it. Log every request for 30 days, or 40 leads if that takes longer, with the columns below. Booked means a yes or a no, nothing in between.

Columns: A received (date and time) | B source | C first reply (date and time) | D minutes to reply | E photos received | F quoted | G booked | H job value

D2: =IF(C2="","",ROUND((C2-A2)*1440,0))

Close rate, replies within 5 minutes:
=IFERROR(COUNTIFS(D:D,"<=5",G:G,"yes")/COUNTIFS(D:D,"<=5"),"n/a")

Close rate, replies from 6 to 60 minutes:
=IFERROR(COUNTIFS(D:D,">5",D:D,"<=60",G:G,"yes")/COUNTIFS(D:D,">5",D:D,"<=60"),"n/a")

Close rate, replies after an hour:
=IFERROR(COUNTIFS(D:D,">60",G:G,"yes")/COUNTIFS(D:D,">60"),"n/a")

Treat a bucket with fewer than about 20 leads as noise; that is my own rule of thumb, not a statistical law. If every reply currently takes a day, there is nothing to compare yet. Run the fast reply for a fortnight and compare it with the fortnight before, remembering that a wet week and a sunny weekend change demand by themselves. Add the source column from day one, because speed matters differently by source. A form on your own site and a shared marketplace lead can behave very differently, and the sheet will show it.

Here is what a finished sheet might look like, with illustrative numbers. Of 42 requests, 14 got a reply within five minutes and five of those booked, a close rate of about 36 per cent. Sixteen were answered between six minutes and an hour and four booked, which is 25 per cent. Twelve waited more than an hour and one booked, about 8 per cent. A spread like that looks like proof, but read it with suspicion. Each bucket holds fewer than 20 leads, so by the rule above it is still a hint. The quick replies may also have gone to the easy leads, the ones that arrived with photos on a Tuesday morning, while the slow ones landed late on a Saturday night. Split the same table by source and by whether photos came with the request. If the gap survives inside each group, speed is doing real work. If it shrinks, part of the gain was the type of lead, and a reply tool will pay back less than the raw table suggests.

The instant reply: what it says and what it must not

A good first message does five things in under 60 words: confirms what you understood, gives a range, asks for specific photos, says when the quote will arrive, and offers an opt-out. The range comes from your rate card, which the rate card tutorial helps you build, and the photo request depends on the surface.

The instant reply needs no AI to be fast. AI earns its place when the request text is messy, such as "drive and the patio round the back and maybe the front steps", and has to be turned into services and size bands. Keep the price out of its hands. Give it the table and let it choose the row, never compute.

You draft the first text reply to a pressure washing quote request. Use only the request and the price table below.

Rules
1. 60 words or fewer, plain words, no emoji.
2. Name the services the customer asked for.
3. Quote only the range from the table for that service and size band. If the size is unclear, give the lowest and highest range for the service and ask.
4. Ask for the photos listed for that service.
5. Say this is an automatic reply and that a person will send the quote.
6. End with "Reply STOP to opt out."
7. Never state a fixed price, an arrival time or a discount.

PRICE TABLE
Driveway up to 2 cars: $179 to $259
Driveway 3 or more cars: $249 to $349
Patio up to 20 sq m: $119 to $189
Photos for driveways: wide shot from the street, close-up of the worst patch
Photos for patios: the whole patio, close-up of the worst area

REQUEST
"Hi, drive and the patio round the back need doing, drive is a two car one with a bad oil patch. Free this weekend?"

An illustrative output:

Hi [first name], thanks for your request for a driveway and patio clean. Driveways up to two cars usually run $179 to $259 and patios up to 20 sq m $119 to $189, depending on size and stains. For a firm price, please reply with a wide photo of the drive from the street, a close-up of the oil patch and one of the whole patio. This is an automatic reply; a person will send your quote today. Reply STOP to opt out.

Check what it did. It used only the table's ranges, asked for the right photos, said it was automatic and promised no arrival time, so "free this weekend?" gets answered by a human with the quote. It also said "today", which the rules did not authorise. Either delete that word or make sure you really do send quotes the same day. A reply that promises more than you deliver is worse than a slower, honest one.

Ways to build it, from free to a small monthly fee

• Saved text snippets, free. Three or four messages saved on your phone, one per service, pasted in when a request arrives. It works if you are near your phone and fails when you are on the roof of a conservatory. Good for the first fortnight of your sheet.
• Form to automation to text, from about $20 a month. The quote form feeds an automation tool, an AI step picks the services and fills the template, and a texting or email step sends it. On Zapier's pricing page, the Free plan has 100 tasks a month, and the entry paid tier gives 750 tasks for $19.99 a month on annual billing or $29.99 monthly. Triggers and filters do not count as tasks, but AI steps do, at a rate that depends on the model tier. If an AI step counts as up to five tasks, a lead might use seven in all, with one task each for the text and a sheet row. At 40 leads a month that is 280 tasks, comfortably inside 750. Have the automation send the draft to you for approval for the first two weeks.
• A field-service app. Jobber's pricing page lists request forms you can share on a website or social media, quotes drafted automatically from new requests, automatic quote follow-ups by email or text, and a Receptionist add-on that answers calls and texts. It says two-way texting is available in only some countries. Automatic quote follow-ups, which send up to two reminders, and auto-drafted quotes start on the Connect plan rather than Core, and the Receptionist is a $29-a-month add-on unless you are on Plus. Check the plan table before counting on any of them, because inclusions change.
• Phone calls too. A missed call is a lead like any other, and missed call text-back for a service business covers the cheapest way to answer it.

Photo triage: what AI can read from two photos

Once photos arrive, an AI can describe what it sees and suggest what to ask next. It cannot measure the area reliably from a phone photo, because perspective hides depth, and it cannot tell whether a surface will survive pressure. Use it to prepare your quote, not to produce it. For measuring from the air, see the satellite-image estimating tutorial.

Look at these two photos from a pressure washing quote request. Say what surface it appears to be, what dirt or stains are visible, and anything that could change the method or the price (loose paint, a cracked or flaking surface, rust, a sealed finish, steep or high access). Do not estimate the area and do not give a price. If you cannot tell something, say so. End with up to three questions to ask the customer.

Surface: a concrete driveway with a brushed finish, roughly two cars wide. Visible: green-grey algae across most of it, dark oil staining near the garage door, weeds in the expansion joint. Could change the method: the oil may need a degreaser step; the edge by the lawn looks crumbly, so pressure may need to be lower there. Cannot tell: whether it has been sealed. Questions: 1. Is the oil staining old or recent? 2. Has the driveway ever been sealed? 3. Is there an outside tap within hose reach?

That is an illustrative output, and it is useful for a reason: it ends in questions, not a number. A person still reads the photos before quoting. Customers' photos can show house numbers, vehicles and children, so use them only to quote and do not repost them without asking.

Follow-up cadence: five touches in two weeks, then stop

Most quote requests that go quiet were not refusals. The customer got busy, compared a second quote, or lost the thread. A short, fixed schedule picks up the ones that were still open without turning you into a nuisance.

After touch five, stop. A reply from the customer pauses every automatic step so a person takes over, and a STOP ends texting at once. AI can draft variants of touches two, four and five so they do not read like a stamp, but a person should send every quote. If your app can chase unanswered quotes itself, as Jobber's quote follow-ups do, map its reminders onto touches four and five, then review a week's messages before you trust it. The quote follow-up email sequence tutorial has wording for the email touches.

Consent for texts: what to collect and what to keep

Texting is more tightly regulated than email in many places, and the rules differ from one place to the next. What follows is practical guidance, not legal advice, so ask an adviser who knows the rules where you work.

• Ask where the number is collected. Put a line beside the phone field: "By sending this you agree we may text you about this quote. Reply STOP to opt out."
• Keep a record. Store the form entry with its date and time against the lead, or keep a screenshot. If a customer later says they never agreed, the record answers it.
• Identify yourself and include the opt-out line in the first message. Texting providers usually handle STOP for you. Twilio's help article, for example, says it handles standard opt-out replies such as STOP, UNSUBSCRIBE and CANCEL by default, blocks further texts to that number and lets the person opt back in by replying START. After a STOP, send nothing more by text.
• Stay inside the purpose. A quote request is permission to text about that quote. Seasonal promotions and reminders need their own permission.
• Do not text bought lists. People on a seller's list did not ask you for anything, which also makes a cold text a poor first impression.
• Respect sleeping hours. A request at 11 pm can get an email straight away and a text in the morning.
• Ask your texting provider about registration. In some markets, business numbers must be registered before messages are delivered reliably, and the process can take days.

Lead sources: what to buy, what to earn, what to test

The same sheet that measures reply time also prices each source. For every source, work out the cost per booked job, which is the cost per lead divided by the close rate, and compare it with the $176 a job leaves in the example.

The Local Services Ads row comes from Google's own help page on the move, which also confirms you still pay for valid leads such as calls and messages, not for clicks. Prices for marketplace leads vary by job and market, so use your own invoices rather than anyone's published average.

Four ways an automatic reply goes wrong

Each of these is easy to prevent once you know it exists, and expensive when you find out from a customer.

• The wrong size band. The customer writes "big drive", the AI picks the two-car row, and the range is $90 too low. The rule that prevents it: if the size is unclear, give the full range for the service and ask.
• The bad hour. A form arrives at 2 am and the automation texts straight away. Set a send window, queue anything outside it for the morning, and let email go at any hour.
• The ignored "no thanks". A customer writes "found someone else" and the day-three follow-up arrives anyway. Make any inbound message pause the sequence, and test the pause by replying to your own test request.
• A quote that contradicts the range. The range said $179 to $259 and the quote says $340 because the photos show a three-car drive. Explain the change in one sentence in the quote. If it happens often, the rate card needs fixing, not the wording of the reply.

A four-week rollout, with what to check each week

1. Week one: baseline. Start the sheet and change nothing. Note the median reply time and the close rate you really have, because most owners are surprised by at least one of them.
2. Week two: saved snippets and the photo request. Reply from the snippets within minutes whenever you can. Check how many requests now arrive with photos, and how many quotes you can send the same day.
3. Week three: the automation, with approval. Turn on the instant reply and have drafts come to you first. Read every one for invented prices, promised times and missing opt-out lines.
4. Week four: follow-ups and the first comparison. Switch on touches two, four and five, then compare close rates across the reply-time buckets. If speed shows a gain, decide how much budget moves from buying leads to keeping the automation running. If it shows none, check whether replies were really faster, whether the range put people off, and whether the leads were the sort that never book.

Whatever the sheet shows, the decision about buying more leads is now yours to make with a number in front of you. Buy when the cost per booked job is well under the profit per job and your crew has idle days. Fix the reply first when it is not, because a lead you pay for and answer late is a cost you pay twice.

Questions about answering pressure washing leads fast

What if I cannot reply because I am up a ladder?

That is what the instant reply is for. It sends the acknowledgement and the photo request automatically, and you send the quote at lunch or when the job ends. Missed-call text-back and AI receptionist add-ons extend the same idea to phone calls. The point is that nobody waits hours just to hear that you got their request.

Should the reply say it is automatic?

Yes, in a short phrase such as automatic reply from your business name, and add that a person will send the quote. If an AI answers questions back and forth, say so at the start. Some places require telling people when they are talking to a chatbot, so check the rules that apply where you work.

Is buying leads ever worth it?

Yes, when you have idle crew days, you reply within minutes, and your measured cost per booked job sits well below the profit on a job. Buy a small batch, tag the source in your sheet and compare after 30 days. If your replies are slow, fix that first or you will pay for the same lead twice.

How fast is fast enough?

There is no verified universal number, and your own sheet is the best judge. A first sensible target is an automatic acknowledgement within a minute or two and a quote within a few hours of the photos arriving. Then compare close rates across reply-time buckets and tighten whichever bucket is losing jobs.

Further reads

• Pool Service Leads: Compare Paid Leads, Local Services Ads and Fast Replies — Paid leads, Local Services Ads and fast replies for pool companies
• Carpet Cleaning Leads: Buy Them or Reply Faster With AI? — The same buy-or-reply question for carpet cleaners
• Deck Builder Leads: What They Cost and How to Reply Fast With AI — What deck leads cost and how to answer them quickly
• Tree Service Leads: Paid Leads vs Answering Every Enquiry Fast — Paid leads against answering every tree enquiry fast
• Handyman Leads: Thumbtack, Angi and Bark Compared on Cost per Booked Job — Marketplace lead sources compared on cost per booked job
• How to Reply to Every Enquiry in Under Five Minutes With AI — A general routine for answering every enquiry in five minutes
• Pressure Washing Proposal Template: Bid HOAs and Commercial Sites — Filled-in bid wording for HOAs and commercial sites, a prompt that turns site-walk notes into a draft, and a price check before you send it.
• Can AI Help a Roof Cleaning Business? Photo Quotes and Safety Notes — Which roof cleaning jobs AI can take (replies, photo requests, captions, rebooks) and which safety checks must stay human.
• Can AI Write a Pressure Washing Contract? What to Check — AI can draft the first version of a pressure washing contract; the access, damage and payment clauses are the ones to check before a lawyer sees it.
• AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.

Sources: Jobber's pricing page; Zapier's pricing page; Google's help page on Local Services Ads moving to Performance Max campaigns with pay-per-lead goals; Twilio's help article on opt-out keywords. Close rates, job values and margins in the examples are illustrations.