I lost a buyer to four hours of silence, then built the fix

The pattern

An enquiry stops being urgent without anyone deciding that.

It arrives. It is seen. A reply is intended. Every one of those is true and none of them is the failure point.

Then something else takes priority. A call, a delivery, a meeting, a drive. The enquiry does not get rejected or deprioritised. It simply stops being the thing in front of anyone.

No alarm sounds, because nothing was ever set to count. When it surfaces again, the buyer has made another arrangement.

Every business already intends to answer quickly. Intent was never the part that broke.

The gap is unmeasured, not unintended

The instance

One case, recorded in full.

Sneaker retail. Enquiries handled by me directly.

Sequence

Day oneMet a man at tennis. Pitched him. Numbers exchanged.
Day twoProspect sends a message requesting specs and recommendations. A budget is stated.
On arrivalI read it while driving. I replied that detail would follow once I got home.
Four hoursNo further contact. Nothing counts the elapsed time. Nothing raises it.
On returnThe conversation had been deleted. No follow-up was possible.

The pitch was not the problem. The price was not the problem. The product was not the problem. A buyer with a stated budget asked to be sold to and received a holding reply for four hours.

That one got noticed because I watched it happen. Nothing was counting the others, and no record existed to check them against.

What was built

A clock on every enquiry. A record every time it runs out.

Two workflows in n8n. Intake runs in seconds and never blocks. Governance runs on its own separate loop, so a fault in monitoring can never drop a live enquiry.

The path an enquiry now takes

Captured on arrivalThe submission posts straight into n8n and the fields are validated before anything else happens.
Written to the CRMHubSpot is searched for an existing contact, then the contact is created or updated and a deal is opened. The same customer never lands twice.
Customer acknowledgedA confirmation goes out in the same run, so nobody is left wondering whether the message arrived.
Owner named and alertedAn alert carrying the customer, the service, the amount and the deal reference. The clock starts here.
Checked every 10 minutesThe governance workflow reads the deal stage from HubSpot and asks one question. Did a person move it.
Escalated after six checksOne hour without movement and the owner is named in a separate private channel, not the busy lead feed.
Breach written to a logTimestamp, deal, customer, service and cycles elapsed, appended to a sheet that survives outside the CRM.

What it sends

Two messages, two channels, on purpose.

Slack message showing a new lead request with customer, service, amount and booking number — Fires in the same workflow run that creates the deal. The owner has the record before they have opened an inbox.

Slack escalation message showing a lead not contacted in one hour with deal reference — Escalations go somewhere separate. A busy lead feed gets muted inside a week, and the message that matters most is the one most likely to be scrolled past.

Since it went live

What changed, and what I will stand behind.

4 to 9

Deals closed per week

My own business, my own count.

1 hour

Longest an enquiry sits

Verifiable. After that, somebody is named.

Every

Breach recorded

Logged whether or not the alert is read.

The customer count is his, given after the first week. One week is not a trend and I have not measured it myself, so it belongs here as an early signal and nothing more.

What is verifiable is the mechanism. No enquiry now sits without a named owner and a clock against it. When one goes quiet for an hour, somebody is told, and the breach is written down whether or not anyone reads the alert.

The log answers a question that could not be answered before: how many enquiries went quiet last month, and which ones.

Still being measured

Weekly counts are being kept. If the figure does not hold, it will be corrected here rather than quietly removed. The system has not yet run a full quarter in a business other than mine.

Where this applies

Headcount changes nothing about this.

This one was a single operator. The same failure runs at fifty people, and it hides better there, because everyone assumes somebody else picked it up.

What matters is not headcount. It is whether anything in your process can tell you how long an enquiry sat before a human touched it. If the answer is no, that gap is costing you and you cannot currently see it.

Start here

Tell me where your enquiries go quiet.

Thirty minutes. Bring where enquiries land, roughly how many a month, and who is supposed to answer them. On the call I will submit a lead into the live system and you watch the alert, the CRM record and the escalation happen in front of you.

Book a 30 minute call See the full system