A customer asked for a quote. The rep spent two days emailing to collect details, checked three files to work out the price, copied and edited an old quote and forgot to update one figure. The quote went out on day four. By then the customer had said yes to someone else.

The quote stage is one of the biggest time sinks in sales and one of the least measured. It is also one of the best suited to automation, because its steps are clear and repetitive.

The end-to-end flow, step by step

1. Enquiry and information. The request arrives by web form, WhatsApp or email. An AI agent fills gaps with short questions: product, quantity, delivery location, target date. Everything is logged in the CRM.

2. Pricing. Price lists, discount rules and cost items live in one source, and the system calculates the price from the inputs. AI is not needed here; clear rules are enough and more reliable.

3. Draft quote. The document is generated from a template. AI can tailor the scope description to the needs the customer stated in their first message.

4. Human approval. The draft goes to the rep or a manager. Quotes above a set value or with non-standard discounts always pass through approval. This step is not automation's weakness; it is its safeguard.

5. Sending and signature. The approved quote goes out by email, and WhatsApp if wanted, and can be accepted electronically.

6. Follow-up. Was it opened, and when? If there is no reply after a set time, a reminder goes out and a task opens for the rep.

7. After the win. An accepted quote becomes an order, project record or invoice automatically. Nobody retypes anything.

Automation is not about removing people from the process. It is about making sure people are present only where a decision is needed.

Where AI genuinely helps

• Pulling structured details out of free-text requests.
• Spotting missing information and asking the right question.
• Writing scope and description text tailored to the customer.
• Reading replies and classifying them: accepted, negotiating, question, declined.

What not to leave to AI

• The price calculation itself. It should run on rules and always give the same result.
• Non-standard discounts and payment terms.
• Legally binding contract clauses.

What you gain

Measure three numbers: time from enquiry to quote, win rate of quotes sent and quotes per rep per week. When quotes get faster, win rates often rise too; our piece on response time explains why.

Frequently asked questions

Does quote automation work in every sector?

It pays off most where quotes follow clear rules and repeat often: manufacturing, wholesale, service packages, installation work. For highly bespoke projects designed from scratch, automation is limited to gathering details, templates and follow-up, which still saves real time.

Will it work with our accounting and ERP systems?

Most modern accounting and ERP systems offer an API. Price lists, stock and customer data can be read from them, and accepted quotes written back as orders or invoices. Without an integration, regular data exports are a start; what matters is a single source of truth.

Will customers notice the quote was automated?

Noticing is not a problem. What matters is that the quote is accurate, fast and specific to them. A quote prepared automatically but checked and approved by a person, addressing what the customer asked for, is usually received better than a hand-made one arriving days later.

The flow rests on a sound CRM; start with what a CRM is. For the basics of agents, read what an AI agent is. To build end-to-end automation with us, see AI agents and automation.