Speed to Lead: Measure Your Own Number First

The most reliable way to lose a job you have already paid to win is to take four hours to call back.

That is not a slogan. It is the most consistent finding in the research on inbound lead response, and it is also the single thing I most often find broken when I take over marketing for a local service business. The advertising works. The phone rings. Nobody answers it, or somebody answers it tomorrow.

What the research actually says, and what it does not

The study everyone cites is real, and it is worth naming and dating rather than waving at.

James Oldroyd's lead-response research, summarized in the Harvard Business Review piece The Short Life of Online Sales Leads (2011), examined a large volume of inbound web leads and found that firms responding within an hour were far more likely to have a meaningful conversation than those responding even an hour later — with the odds falling sharply across the first hour rather than gradually across the first day.

Now the limitations, because they matter and are almost never mentioned:

• It is from 2011. Customer expectations have moved further toward immediacy since, not back.
• It studied business-to-business web leads, not homeowners with a leaking water heater. The direction of the finding is very likely the same; the exact magnitudes should not be quoted at a plumbing business as though they were measured there.
• It measured contact and qualification, not revenue.

So the honest version is: responding fast matters a great deal, the effect is concentrated in the first minutes rather than spread across the first day, and the precise multiplier for your trade in your market is not known to anyone — including me — until you measure it yourself.

Which you can. That is the point of the rest of this.

Why the first few minutes are different

The mechanism is not mysterious. A homeowner with an urgent problem does not submit one form. They submit three or four, or call down a list, and then they get on with their day. The first company to reach a live human is talking to somebody who has not yet spoken to anyone else. The fourth is arguing with a decision already half-made.

There is a second effect that is easy to miss: intent decays. Somebody who filled in a form at 9pm about a slow drain has, by the following afternoon, either fixed it, decided to live with it, or booked whoever called first.

Measure your own number before you change anything

Do this before buying software, because most businesses are wrong about their own response time and the direction of the error is always the same.

For two weeks, log every inbound lead:

1. Timestamp of arrival — from the form notification, call log or platform.
2. Timestamp of first genuine human contact — not an automated acknowledgment. A conversation.
3. Channel it arrived on.
4. Time of day and day of week.
5. Whether it became a booked job.

Then calculate the median, not the average. One lead you called back four days later drags an average into fiction. The median tells you what usually happens.

Look separately at:

• Business hours versus evenings and weekends. For most trades this is where the losses concentrate, and it is usually much worse than the owner believes.
• By channel. Form submissions typically fare worst, because a call demands attention and a form waits politely in an inbox.
• Median versus worst case. The worst case is the one leaving reviews.

Now you have a baseline that is yours rather than borrowed from a 2011 study of a different industry.

What to fix, in order of cost

Answer the phone. Unglamorous, and still the largest available gain in most cases. Work out what proportion of calls go unanswered — most phone systems will tell you — before spending anything on automation.

Route after-hours calls to a human. An answering service that books appointments is not expensive relative to one recovered job a month. Do the arithmetic with your own average job value; for a trade where a job is worth several hundred dollars, the threshold is low.

Acknowledge form submissions instantly, automatically. An immediate text saying a person will call within fifteen minutes buys you those fifteen minutes. It does not replace the call. A business that automates the acknowledgment and not the follow-through has automated its way to a faster disappointment.

Make the follow-up sequence somebody's job. Most leads that go cold do so because no one owned the second attempt, not because the first attempt failed.

Text, if your customers text. Response rates to a text are typically better than to a voicemail, and for a homeowner at work it is often the only channel that gets through.

The objection worth taking seriously

"I am on a roof. I cannot answer the phone."

That is true, and it is why the answer is a system rather than more effort from an already-busy owner. But the system should be sized to the problem you measured, not to what a software vendor wants to sell you. If your two-week log shows a median of eleven minutes during the day and three hours in the evening, you have an after-hours problem, and an expensive all-channel automation platform is the wrong purchase.

Measure first. The fix is usually smaller and cheaper than the pitch.

What to track afterwards

Re-run the same two-week log a month after changing anything, and compare medians. Then look at the number that actually matters: booked jobs per hundred leads, and cost per acquired customer by channel.

Response time is a proxy. It is a good proxy, and it is still a proxy — if the median falls and booked jobs do not move, something else is wrong, most likely in what happens on the call rather than how fast it starts.

Where I come out

I build and run websites and advertising for local service companies, so I have a direct interest in what happens after somebody clicks. My honest view: for most of the businesses I have taken over, improving lead response was worth more than anything I could have done to the advertising, and it cost less.

It is also the least discussed part of local marketing, because there is nothing to sell in "answer your phone faster." That does not make it less true.