Notes for owners · CRM, automation and RevOps
How to find out where enquiries are leaking: trace them from form to quote
Send three test enquiries through each source, then follow every real one from form to first reply to quote. A leak map that shows which step swallows your leads.
The GullySales team · Updated 17 Oct 2026 · 5 min read

On this page
"We get enquiries every day and sell almost nothing" is a sentence owners say with real confidence, and it is often wrong in a specific way. Enquiries tend not to shrink evenly. They can vanish at one handoff: the form to a person, the person to a first reply, the reply to a quote. To find the handoff, count the same people at each step. To find the break your records cannot show, the enquiry that was never recorded, send some yourself.
Why the headline numbers hide the leak
Between "enquiries received" and "orders won" sit four or five handoffs, and each can lose people without anyone noticing. A lost lead leaves nothing behind. No complaint arrives, no file is open, and the buyer simply buys elsewhere.
Comparing the total from one report with the total from another does not help, because the two reports count different things. The website tool counts form fills. The phone counts answered calls. The sales sheet counts what somebody remembered to type. The gap between them is not the leak, only the disagreement between three rulers.
So choose one source and one fixed set of dates, and follow every person in that set to the end. The same names must appear at every step.
Send tracer enquiries before you read a report
Do this yourself, not through a staff member who knows the trick. Fill in your website form with a name you will recognise. Call your published number from a phone nobody has saved. Send a WhatsApp message to the business number. Leave one missed call if you use a missed-call line. Then write the answers on a card like this, one card per test:
Source: ______ Sent at: ____ (day, time)
Who was told? ______ When did a person see it? ______
First reply: channel ______ at ______
Did it use my name or my words? yes / no
Did anyone ask me a question back? yes / no
Did a record appear anywhere? where: ______
Run each source three times: once in office hours, once in the evening, once on a Sunday. Some breaks are timing breaks. The form that works at noon but mails a shared inbox nobody opens at night shows up only in the evening run. The point is not an average; one test that never appears is already a finding.
Rebuild one fortnight as a single list
Now take a fixed slice of real enquiries, say every enquiry from one fortnight, and put them on one sheet, one row per person. Add four date columns: arrived, first reply, quote sent, decision. Leave blank what is blank. If your records are on paper or inside WhatsApp chats, rebuild by hand; it is slow once and worth it.
Count how many rows have each date. Do not look for a rate. Look for the column where the number falls furthest below the one before it. That is where you work. Fixing the third column first when the second is the broken one wastes effort, since a leak early on starves everything after it and makes later steps look worse than they are.
A water-purifier dealer in Tirunelveli
Imagine a dealer selling water purifiers and softeners to homes and small hostels around Tirunelveli. This case is invented to show the tracers at work. Enquiries come from a website form, a Facebook lead form, calls from a hoarding, and walk-ins. The dealer believes the problem is price, because buyers say "we will tell you".
The tracers show the website form arriving at the shop. The Facebook form did not: it mailed an address belonging to a nephew who had left, and the notifications sat unopened. Phone calls were answered but written nowhere, so nobody could say how many people had called or what they wanted.
The dealer's first reaction is to hire a telecaller, which would add a person to a pipe with a hole in it. Instead the dealer points the Facebook form at a shared mailbox that two people check, and places a register by the phone with name, number, need and time. After a fortnight a fixed slice can be counted. In the next fortnight 31 Facebook leads arrive and 5 become quotes, where there had been none to count, and the price complaint turns out to be a smaller worry than the leads nobody saw. Some of those earlier Facebook enquiries are by now too old to call, perhaps two months of them, and that loss stays.
What the pattern tells you
If the tracers never arrive, the fault is mechanical and cheap: an address, a number, a missing owner. If they arrive but replies are slow, the fault is about who is responsible for which source. If replies are quick but quotes are few, look at qualifying, because people may be asking for prices before anyone knows what they need.
One more pattern deserves a pause before you act. A source can look poor because only its worst enquiries reached a person. Check that before you stop paying for a channel. The note on tracking lead source explains how to keep the source field from being overwritten later.
When tracing is more than you need
With only a handful of enquiries a week, skip the sheet and ring each person yourself. An afternoon of calls teaches more than a table. And if there are no records at all, there is nothing to trace: begin a one-line-per-person register, and run the tracers on the plumbing alone. The post on records that are too thin to analyse shows how to begin.
For the step after the first reply, read why businesses lose leads after the first conversation.
Close the widest gap, then trace again
Choose the one handoff where the count dropped furthest, repair only that, wait for a fresh slice and trace it the same way. If the break stays hidden after your own tracers, lead leakage analysis is where we would look, part of how you sell, and tell us what you need starts it.








