Every sales team I have worked with keeps some version of the same number on the wall. Five minutes. Sixty seconds. Respond before the competitor does.
It is a good rule, and the data behind it is not soft. Harvard Business Review’s audit of 2,241 US companies found that 24% took longer than a day to respond to a web-generated lead and 23% never responded at all. Firms that replied within an hour were close to seven times more likely to qualify the lead.
Somewhere along the way, though, “respond fast” quietly became “quote fast.” Those are not the same instruction, and treating them as one is expensive.
Here is the argument, and it will sound slightly heretical on a sales blog: a deal you lose to a faster competitor is usually recoverable. A deal you win with a wrong number often is not.
Speed Belongs To The Reply, Not To The Price
Response time and quote time get measured with the same stopwatch in most CRMs, which is the root of the problem. They are different commitments.
A reply costs you nothing to make. It is an acknowledgment, a qualifying question, a booked site visit. Getting that out in minutes is close to free money, and the research on lead response time and revenue is unambiguous about the compounding effect.
A price is different. A price is a promise with your margin attached to it. Once a buyer has seen a figure, that figure becomes the anchor for every conversation that follows, including the awkward one where you explain the number went up.
What A Wrong Number Actually Costs
Most teams never calculate this, because the damage shows up in three separate places and none of them are labelled “bad estimate.”
The Revision Tax
Every corrected quote is a second sales cycle you did not budget for. New approvals, another round of internal sign-off, sometimes a fresh procurement review. A revision is rarely a five-minute email. In field service and construction sales, a re-quote frequently costs more hours than the original bid did, because now you are also managing an unhappy buyer.
Forecast Contamination
Fast quotes inflate pipeline value with figures nobody has stress-tested, and those figures flow straight into the forecast. Your coverage ratio looks healthy. Your win rate looks stable. Then Q3 closes 18% under plan, and nobody can point to the cause, because the cause was distributed across forty deals that were each priced slightly optimistically.
If your pipeline has started telling you comfortable lies, premature quoting is worth checking before you blame lead quality.
The Trust Discount
Buyers remember the correction more clearly than they remember the speed. A quote that moves from $14,200 to $19,600 does not read as “we refined our scope.” It reads as either incompetence or a bait price, and the buyer now discounts everything else you say, including your timeline.
Two Clocks, Not One
The fix is structural rather than cultural. Stop running one timer and start running two.
Clock One: The Reply
Starts when the lead arrives. Stops when a human has responded with something useful. Target measured in minutes. Automate aggressively here, because nothing in this window requires judgment.
Clock Two: The Commitment
Starts when the reply goes out. Stops when a number you will honour is in writing. Target measured against information completeness, not against the calendar.
The second clock has an exit condition rather than a deadline. It closes when you have the inputs that make the number defensible. If you can get those inputs in twenty minutes, close it in twenty minutes. If the job needs a site visit, the clock stays open, and you say so out loud.
A Three-Tier Quote Ladder
Buyers are not actually demanding a final price in the first hour. They are demanding to know they have been heard, and roughly what league this sits in. You can give them both without committing to a figure you have not earned.
| Tier | What you send | When | What it requires |
| 1. Acknowledgment | Confirmation, two qualifying questions, next step booked | Under 15 minutes | Nothing |
| 2. Banded range | “Jobs like this usually land between X and Y – here is what moves it” | Same day | Comparable past jobs, stated assumptions |
| 3. Committed price | Fixed figure, scope, validity window | After verification | Measured or calculated inputs |
Two rules make the ladder work.
State the tier explicitly. Write “this is a range, not a quote” on the range. Buyers are fine with that when you say it plainly and furious about it when they discover it later.
Never skip a tier to look responsive. Jumping from tier one to tier three because a buyer sounded impatient is how the revision tax gets paid.
What This Looks Like In A Technical Quote
Take HVAC replacement, which is a useful example because the wrong-number failure mode is so well documented. The fast path is to size the system off square footage. It takes ninety seconds, and it produces a wrong number often enough that the industry has a standing warning about it.
ENERGY STAR tells homeowners outright that oversized equipment cycles too frequently and shortens its own lifespan, and advises them to ask whether the contractor used the actual characteristics of the building rather than a rule of thumb.
So the rep who quotes fast off square footage is not just risking margin. They are quoting equipment that will generate callbacks, comfort complaints, and a warranty conversation eighteen months out.
The slower path is a load calculation, and the reason it used to lose deals is that it took hours. ACCA-approved Manual J software, browser tools like Cool Calc, and platform calculators such as the one Dalton Mills publishes for estimating how much heating and cooling a building needs all compress what an afternoon of work into something a rep can run between appointments.
Whether a rep runs Wrightsoft, Cool Calc, Dalton Mills, or a paper worksheet matters far less than whether the calculation happens at all before a figure goes out.
The pattern generalises beyond HVAC. Roofing squares, panel capacity, duct static pressure, crawl space access – in every trade there is one input that determines whether the price holds, and it is rarely the one the buyer volunteers on the phone.
Find that input for your category. Then build the tier-two range so it is honest about what happens when that input comes back badly.
What Not To Automate
The obvious next move is to push all of this into software, and most of it should be. Instant acknowledgments, assumption checklists, reminder sequences: automate the lot. Automated follow-ups solve a real problem, given how many reps stop after a single touch.
Three things should stay human.
- Scope interpretation. When a buyer says “standard install,” someone has to decide what that covers. A calculator will not ask about the 1920s knob-and-tube wiring in the basement.
- Risk pricing. Contingency for an unknown is a judgment about the customer and the site, not a formula.
- The escalation call. When tier two comes back at double what the buyer expected, that conversation is worth having live, and having early.
This is the boring reality underneath a lot of the enthusiasm about AI in smaller operators. The useful applications tend to be narrow and specific, closer to faster forecasting and fewer manual steps than to end-to-end decision-making. The Dalton Mills load calculator is a fair illustration of the type: it removes the arithmetic – it does not remove the site visit.
Instrumenting This In Your Crm
Four changes, none of them heavy.
Keep the estimating output inside the record. If the calculation happens in an external tool – a takeoff app, a spreadsheet, a Dalton Mills style calculator – paste the result and its assumptions into a CRM field rather than leaving it as an attachment nobody opens. Sales managers cannot audit what they cannot query.
Split the timestamps. Add a “first response sent” field and a “committed price sent” field. Measure them separately, or you will keep optimising the wrong one.
Add a quote confidence flag. Three values: range, verified, fixed. Weight forecast contribution accordingly. A tier-two range at 60% confidence should not sit in the pipeline at the same weight as a verified price.
Track revision rate per rep. Quotes revised after sending, as a percentage of quotes sent. Anyone consistently above 15% is quoting too early, and they usually know it.
Set a stage exit rule. A deal cannot advance to “proposal sent” until the confidence flag reads verified or fixed. This single gate does more for forecast accuracy than any weighting formula.
Run those four for a quarter, and you will have the data to settle the speed-versus-accuracy argument internally, which beats arguing about it in a pipeline review.
Conclusion
The speed-to-lead rule is correct, and it has been widely misapplied. Answering in five minutes is close to costless. Pricing in five minutes is a bet you are making with margin that has not been counted yet.
Split the clocks. Give buyers a fast acknowledgment and an honest range, then commit to a figure once you have the one input that decides whether the job is profitable. Most of the tooling needed to make that second step quick already exists.
Your first number does not have to be your fastest. It has to be the one you can still stand behind at invoicing.
