How to Send Quotes on WhatsApp Without Losing the Deal

"How much does it cost?" is the most expensive message in your inbox

Quotes are where service businesses lose deals, and the loss almost always happens in the same place: the customer asks for a price without giving you what you need to calculate one, and the honest answer — "it depends" — starts a conversation that runs for hours. Whoever gets back first with a concrete, defensible number usually wins the job. Everyone else ends up competing against a quote that is already on the table.

This guide covers the whole path: how to send quotes on WhatsApp fast, what to ask and in what order, when an AI agent can price a job and when it must not, how to close in-thread, and how to follow up without paying for every reminder.

The channel is on your side. Kantar for Meta (n=11,056, 22 markets, April–September 2025) found that 72.4% of people are more likely to buy from a brand they can message. What the channel does not fix is the arithmetic of time: every hour between the question and the number is an hour that same customer spends messaging someone else.

The five variables behind every price

Every service quote needs the same five things: scope (what exactly gets done), size (how many units, how many square metres, how many people), timing (when it is needed), location (where, and what access there is) and terms (warranty, payment, who supplies materials).

The mistake is not failing to know them. The mistake is asking for all five in one message. A paragraph with five questions comes back with two answers, and the three missing ones are usually the ones that move the price.

WhatsApp gives you a better structure, and it has hard limits worth knowing before you design the intake. Meta documents them for the Cloud API:

• Reply buttons: three per message, maximum, with titles up to 20 characters.
• Lists: ten rows in total, spread across a maximum of ten sections, with row titles up to 24 characters and descriptions up to 72.

Treat those as a design rule rather than a constraint. If a question needs more than ten options, it is almost always the wrong question — split it across two turns or turn it into a range. Three buttons force you to offer the three variants you actually sell, which is a commercial discipline before it is a technical one. The agent's capabilities are configured on top of that same message structure.

How to send quotes on WhatsApp, step by step

1. Reframe in the first minute. "Hi, thanks for reaching out" advances nothing. Acknowledge the job and ask the first question in the same message: "Yes, we do office moves. Is this within the city or intercity?"

2. Run the intake over three or four turns, not one. One question per message, with closed options wherever possible. The customer answers with a tap instead of typing, and you get clean data instead of a paragraph somebody has to interpret.

3. Give a number, even if it is a range. A range with its condition stated — "between X and Y depending on the floor; I'll confirm once I see the photos" — works far better than "let me work it out and get back to you". A range keeps the conversation alive. Silence ends it.

4. Send the quote as a message, not only as a PDF. The PDF is the formal record. The three lines inside the thread — what's included, what it costs, how long the price holds — are what the person actually reads on a phone and what they can reply to without leaving the conversation.

5. Close with a deposit, not with "let me know". Meta's selling stack stops at the order: its Cloud API guide for selling products and services covers catalog, cart and order, and includes no payment step at all. The money moves through your own processor, in a link that travels down the thread. How that close is built is covered in the WhatsApp sales playbook.

What the AI can price, and where it has to stop

An agent can quote anything that is documented, and nothing else. That is the line, and it is worth writing down before you switch anything on.

On the documented side: unit rates, zone surcharges, volume discounts you have already approved, standard lead times. If your price list lives in the knowledge base, the agent retrieves it by meaning rather than by keyword match, and builds the quote from the rule you wrote.

On the other side: any variable it cannot observe from the chat. A site visit, custom fabrication, a third-party cost that hasn't landed yet, a pricing exception. The agent's job there is not to guess — it is to finish the intake and hand the conversation to a person with the full context, so nobody has to re-ask what the customer already answered.

That split also matches how the customer feels about it. The same Kantar study found that 67.7% consider an AI reply helpful, while only 42.9% believe AI would improve their experience — people accept a fast, correct answer and distrust the general promise. An agent that quotes what is tabulated and escalates everything else lives exactly in that gap. It is the same division of labour described in building an AI sales team on WhatsApp.

Follow-up: a sent quote is not a sale

The second loss happens after the price goes out, and its mechanics are predictable enough to plan around.

WhatsApp keeps a 24-hour customer service window open from the customer's last message. Inside that window you write freely. Outside it you need a Meta-approved template, and since 1 July 2025 Meta bills per delivered template, by template category.

That produces a concrete follow-up plan:

• Follow-up 1, same day, inside the window. Confirm the quote arrived and offer the only decision that comes next: book it, adjust the scope, or pass.
• Follow-up 2, outside the window, carrying new information. A utility template that says something — the price expiring, a slot opening, availability changing — not "did you get a chance to look at it?". If you are paying for the message, give it a reason to exist.
• After that, stop. Chasing with nothing new is the fastest route to a spam report, and number quality is an asset you don't easily get back.

Categories, windows and costs are laid out in the WhatsApp Business API guide.

The five numbers, from question to payment

Five numbers are enough, and all five read week to week:

1. Time to first reply. Minutes, not hours.
2. Time to quote sent. It decides more deals than the price does.
3. Share of conversations that receive a quote. If it is low, what's breaking is the intake, not the pricing.
4. Close rate on quotes sent. If it falls while quote volume rises, you are quoting work you don't want.
5. Average ticket by channel. Compare WhatsApp against the rest before moving budget.

For those five to exist, every conversation has to leave a record: contact, tag, status and quoted amount. A CRM that fills itself from the thread turns an inbox into a list of opportunities sorted by expiry date — which is the difference between following up and remembering to.

What to do this week

Don't start with software. Start with the document.

1. List the twenty jobs you quote most and write, next to each one, the actual pricing rule you use today. That document is the knowledge base; without it, any agent ends up improvising.
2. Define three intake questions and their closed options. Three buttons, twenty characters each — if it doesn't fit, the question is too broad.
3. Write the escalation rule in one sentence: which case the AI never prices.
4. Submit one follow-up template for approval, carrying real information rather than an empty nudge.

Only then connect the agent. In xcale that path is the ordinary setup: the price list goes into the knowledge base, the intake is built from buttons and lists, the agent quotes what is documented and escalates the rest with the full conversation attached, and the record lands in the CRM without anyone typing it. There is a 7-day free trial, which is long enough to measure what happens to this week's quotes; plans and pricing are published.