Following up enquiries you already paid for

Before changing anything, measure the current state honestly.

Take the last fifty enquiries. For each one, record four things: when it arrived, when somebody first responded, how many times you contacted them in total, and what the outcome was.

This takes an afternoon and it is almost always uncomfortable. The usual findings are a first response time far longer than anyone believed, a large group that received exactly one contact and nothing further, and a group with no recorded outcome at all.

You cannot fix a process you have not looked at, and the measured version is invariably worse than the remembered version.

Speed matters more than almost anything else in this process, for a simple reason: somebody who filled in your form probably filled in two or three, and the first useful response tends to shape the conversation.

Pick a target you can hold on a busy day, not an aspirational one. If that is two hours rather than ten minutes, set two hours.

Then say it. Putting we reply within two hours on the form and then doing it reliably beats promising immediately and delivering by tomorrow. A met expectation is worth more than an impressive one that is missed.

Make it somebody specific job, with cover for when they are out. Response time collapses the moment it belongs to everyone.

The automatic acknowledgement is not a follow-up. It confirms receipt.

The first real contact comes from a person, references what they actually asked about, and tries to move things forward: a question that needs answering, two times you could speak, or a direct answer if the question was simple.

If you ring and get no answer, leave a message and send a short email saying you rang. Two channels, one attempt, and the person now knows a human is on it.

A process that lives in somebody head stops when they are busy, and dies when they leave.

Write it out. Something like: acknowledgement immediately, call within the agreed window, email the same day if no answer, second call day two, useful email day four, final message day eight, then move to the list.

For each touch: what happens, who does it, and when. Put it where the team can see it.

The value here is not sophistication. It is that on a chaotic Thursday, nobody has to decide what to do next.

Five emails to somebody who does not read email is not persistence, it is a loop.

Alternate. Call, then email, then text where you have permission and it suits the audience. Different people are reachable in different places, and you generally do not know which one this person is.

Two things worth noting. Text has very high open rates and a low tolerance for being used repeatedly, so use it sparingly and where the relationship warrants it. And check consent and local rules before texting, particularly for anything that reads as marketing rather than as a reply to their enquiry.

After the second contact, checking in is not a message. It asks the recipient to do work and offers nothing.

Give each later touch a reason to exist. A relevant example of similar work. An answer to something they raised. A short piece of information that helps them decide even if they decide against you. A genuine deadline if one exists, and no invented ones.

The last message in the sequence should be easy to reply to. Something like: assuming the timing is not right, I will stop chasing. If it changes, reply to this and I will pick it up.

That message gets more replies than the four before it, because it removes the pressure and requires one line to answer.

Every enquiry gets an ending: won, lost, not a fit, no response.

This matters for two reasons. Operationally, an enquiry with no outcome sits in a state where nobody knows whether it is being handled.

Commercially, it is the only way to find out which sources produce customers rather than leads. A channel producing forty enquiries a month that never close is worse than one producing eight that do, and without recorded outcomes both look the same in the report.

A spreadsheet is enough to start. The recording habit matters far more than the system.