A slow reply to an inbound enquiry is not a service problem. It is a line item on your P&L. The person who filled in your form is filling in two or three others right now. You paid to acquire them; they buy from whoever answered first. Below: the research numbers, a formula for calculating your own losses, and an honest breakdown of what is worth automating and what is money down the drain.
What the research says
The canonical work here is the MIT Sloan study led by James Oldroyd, carried out with InsideSales.com. Sample: over 15,000 enquiries and 100,000 call attempts.
The headline result: replying within 5 minutes rather than 30 makes you 100× more likely to reach the person and 21× more likely to qualify the lead. Not percentage points — multiples.
The study was published in 2007, and that matters: communication channels have changed since. But the findings are reproduced in later industry measurements, and the underlying logic has not aged — someone making a decision contacts several companies in a row, and attention goes to whoever answers first.
Some figures in this article come from vendor industry reports rather than academic work. We flag those as industry estimates and do not present them as rigorous data.
Calculate your own losses in five minutes
The formula is simple. Take four numbers from your own analytics.
| What to take | Where to find it |
|---|---|
| Enquiries per month | CRM, or the inbox where forms land |
| Enquiry-to-deal conversion | Deals in a period ÷ enquiries in the same period |
| Average deal size | Revenue ÷ number of deals |
| Actual time to first reply | The gap between the enquiry timestamp and the first reply. Your last ten enquiries is enough |
Now a worked example. A company receives 60 enquiries a month, converts 12% into deals, and has an average deal size of €2,000. That is €14,400 of revenue from inbound.
Now assume a third of those enquiries arrive out of hours or while the salesperson is busy, and the reply goes out the next day. On the data above, a meaningful share of those people are already talking to a competitor by then.
Even if the delay costs you only half of that third — that is 10 enquiries a month, or €1,200 of lost revenue every month. €14,400 a year. And the ad budget for those 10 enquiries has already been spent.
The example above illustrates the mechanics; it is not a promise of results. Your loss rate depends on the niche, the channel, and how urgent the customer's problem is. In long-cycle B2B, delay is forgiven more often; in urgent-need services it costs you almost the whole enquiry.
The only way to know your own figure is to look at actual response times on recent enquiries and check which of them became deals.
You pay for a lead twice: once in the ad account, and again when it goes to whoever replied faster.
Three savings lines, not one
Automating lead handling is usually pitched as convenience. In practice it is three separate sources of money, and they should be counted separately.
1. Enquiries that stop getting lost
The most obvious and the largest. The enquiry is received at any hour, the person gets an immediate reply and knows what happens next. Calculated with the formula above.
2. Time given back to people
A salesperson spends part of every day on repetitive work: re-asking for contact details, clarifying budget, copying into a spreadsheet, filtering out bad fits. Count the hours a week that go to this and multiply by an hourly cost. For a team of three that is usually tens of hours a month — money you already pay for work that does not have to be done by a person.
3. The hire that doesn't happen
When volume grows, the first instinct is to hire. If the routine part is automated, the hiring threshold moves: the same people handle one and a half to two times more enquiries. For a company growing 30–50% a year, that is a deferred hire plus the recruitment and onboarding costs that come with it.
| Source of saving | How to measure it | When you see it |
|---|---|---|
| Enquiries not lost | Conversion before and after, controlled for response time | First month |
| Time recovered | Hours on routine × hourly cost | First month |
| Deferred hire | Enquiries per salesperson before and after | 3–6 months |
Why 95% of deployments don't pay off
Now the uncomfortable part, which people selling automation tend not to write.
An MIT report built on 52 executive interviews, a survey of 153 managers and an analysis of 300 public deployments found that roughly 95% of pilot projects produced no measurable effect on profit. Around 5% created meaningful value.
The report attributes this not to model quality, but to technology being deployed into processes that were never measured, and without integration into how the company actually works.
What set apart the 5% that did pay off:
- One specific pain point rather than ten directions at once
- Internal processes rather than the shopfront: automating what happens inside returns more than experiments on the customer-facing side
- Working with an external partner: per the report, those projects succeeded roughly twice as often as ones built in-house
If you have not measured your first-response time and the share of enquiries left unanswered, there is nothing to deploy yet. Those two numbers first, then a solution. Otherwise you are buying technology without knowing what it fixes, and in six months you will not be able to say whether it paid for itself.
What to do this week
- Take your last ten enquiries and calculate the actual time to first reply for each. Not the average you feel — the one you measure
- Find the ones that were never answered at all. Most companies have them, and there are usually more than expected
- Calculate your cost of delay using the formula above, on your own numbers
- Check that every enquiry source lands in one place. If enquiries arrive by email, in a messenger and through the site form, and nothing brings them together, some are lost simply because nobody knew about them
These four steps require no budget and no agency. They give you a number you can make a decision on — and without it, any deployment is a blind bet.
We build lead handling so the customer gets an answer immediately and the salesperson gets context rather than a bare name and phone number. And we fix the path to the enquiry on the site when leads are lost before they even reach handling.