Respond first, win the quote: automating request triage and the first estimate
A prospect sends the same request to three companies. The one that answers quickly and clearly gets a head start. How to get there without spending your evenings on it.
A prospect who needs a caterer, a tradesperson, a seminar venue or an IT provider rarely sends a single request. They send three or four, on the same day, and compare what comes back.
In that game, the quality of the quote matters. But response time matters almost as much, and it is the one you control least when the owner writes quotes in the evening, between two emergencies.
What the numbers say
The most cited study on the subject was published by Harvard Business Review in 2011 (“The Short Life of Online Sales Leads”, Oldroyd, McElheran and Elkington). The researchers sent a test request to 2,241 US companies through their own websites and timed the response.
- Only 37% responded within an hour. 23% never responded.
- The average response time was 42 hours.
- Companies that got back within an hour were nearly 7 times as likely to qualify the lead as those that waited even one more hour, and more than 60 times as likely as those that waited 24 hours or longer.
The study is old, American, and about first contact rather than the quote itself. But the mechanism hasn’t changed: a request cools down fast, and whoever answers clearly first sets the terms of the conversation.
Why it’s slow in an SME
When you look at how a request turns into a quote in a small company, you almost always find the same friction:
- Requests arrive through every channel: email, form, phone, social media messages. There is no single place to see them.
- They are incomplete. The date, quantity, address or a constraint is missing. You have to follow up, and every round trip costs a day.
- Prices live in one person’s head. Often the owner’s. Until they have time, nothing goes out.
- Every quote starts from scratch, in a Word or Excel file copied from the last one, with the errors that come with it.
None of these problems needs artificial intelligence to be solved. Some of them benefit from it.
The four building blocks
1. A request form that asks the right questions
To price a job, you need three to six pieces of information: what, how many, when, where, and one or two constraints. A form that asks for them at first contact removes most of the back-and-forth. It is the cheapest building block, and often the most profitable.
2. An acknowledgement that commits to a deadline
An immediate automatic reply that says what happens next: “We will get back to you with an estimate before 6 pm tomorrow.” The prospect knows they have been heard, and you have set yourself a deadline.
3. Triage and extraction of free-text requests
Not every request will come through the form. For emails written in free text, an AI assistant can read the message, extract the useful information (dates, quantities, place, budget), flag what is missing and draft a reply asking for it. A person reviews and sends.
This is where AI really adds something: reading a poorly structured text and sorting it into boxes, without writing rules for every possible wording.
4. A quote generator based on a written price list
From the extracted information and your price list, a tool calculates the quote and produces the PDF. The owner or a colleague checks it, adjusts if needed, and sends it.
A quote never goes out without human validation. A wrong quote sent fast does more damage than a right quote sent late.
Building block zero: writing down the price list
This is the prerequisite, and it is often where everything is decided. If your prices only exist in one person’s head, no tool can price in their place.
Writing the price list down, with the special cases, discounts, minimums and options, takes a few hours of interviews. It is also what lets someone other than the owner write a quote, which is valuable in itself, before automating anything at all.
What stays human
- Validating every quote before it is sent.
- Unusual requests, which the tool should flag rather than price at random.
- The phone call that follows the quote, the negotiation, the relationship.
The goal is not to remove people. It is to let them spend their time on what gets contracts signed, not on copying and pasting.
Working out what it’s worth to you
Two gains, not to be added together because they are of a different nature.
Time saved: (current time per quote − time with the tool) × number of quotes per year.
Purely illustrative example: if a quote takes you 45 minutes today and 10 minutes tomorrow, and you write 200 a year, that frees up about 117 hours a year. Replace with your own numbers. To convert those hours into euros, or to find out they don’t produce any, our calculator Hours saved, euros earned (in French) is built for that.
Extra deals won: this is the biggest gain, and the hardest to predict. The only honest way to measure it is to compare your conversion rate before and after, over a few months.
Pitfalls to avoid
- Sending generated quotes without review. See above.
- Building an overly complex system. A form, a price list and a quote template often cover most of the need. Add AI when the volume of free-text requests justifies it.
- Forgetting GDPR. Requests contain personal data. Only send them to tools covered by a data processing agreement, keep only what the quote needs, and set a retention period.
- Choosing the tool before the process. n8n, Make, your CRM or a small custom build can all do the job. The right choice depends on your current tools, not on fashion.
Where to start this week
Take your last twenty requests and note, for each one, the date it came in and the date the quote went out. That gives you your real response time, often longer than you think. Note how many signed as well.
It is the starting point of any project on the subject, and it only takes an hour.
