Nobody's Job Is Remembering
Most lost customers were interested. Nobody was reminded. A look at what happens when following up depends on a person remembering, with real numbers.

Review notes: Problem-aware. Contains the strongest quantified pain in the archive (40–50% first-stage drop-off, described by the customer as industry-normal). Evidence: two international school groups, hotel operator, property brokerage.
Ask a service business why they lose customers and you'll usually hear something about price, or competition, or timing.
Then you look at the pipeline, and the truth is duller and much more fixable.
An admissions manager at an international school, describing where applicants go:
"There are people in this stage of application that we could follow up better. But because you do not have that visibility or a reminder — for example, notification, this person has applied at this time, has it come back to us? If a system could do that, normally it can notify: it has been three days, do you want to follow up? Those kinds of things."
Nobody lost interest. Nobody was reminded.
The number nobody wants to say out loud
At a five-campus international school group, we asked where applicants drop off. The answer:
"It drops off about — it ranges probably about 50%, 40%."
Where? "After initial inquiry. At the very beginning."
Between four and five in every ten people who raised their hand disappear at the first step. And here's the part worth sitting with: he described this as normal for the sector. "I believe it's quite similar to the same industry as well."
He's probably right. Which means an entire industry has normalised losing half of its interested prospects at the earliest and cheapest point to save them.
What "following up" actually rests on
At the same group, we asked what moves a lead from one stage to the next. The answer wasn't a workflow:
"We don't move them along. So we have certain SLAs in place at different stages. If it doesn't move or it doesn't get picked up within a certain period of time, the owner of that week gets a problem."
Read that carefully. There is no mechanism that advances anything. There is a human who is supposed to remember, and a consequence if they don't. The system is a person, and the backup system is a telling-off.
This is not a criticism of that team — they'd built the most reasonable process available to them with the tools they had. It's an observation about what happens when memory is load-bearing.
Two more shapes of the same problem from our calls:
The lead who says they never enquired. "We come across leads which will say, oh, I didn't submit an inquiry. Or there are times when we try to reach out to them, there's no response." When contact comes days after the form, the prospect has genuinely forgotten. Delay doesn't just reduce your chances — it makes you look like a cold caller.
Warm-but-not-ready has nowhere to live. "We call them and they say, oh yes, I am keen on the school tour, I'm not ready to book a slot yet. So where does that update happen?" This is the most valuable state a prospect can be in — interested, just not now — and in most service businesses there is no field for it. So it lives in someone's head until it doesn't.
The diagnostic
Three questions. They're uncomfortable on purpose.
- What happens to an enquiry that nobody touches? If the honest answer is "nothing," you have a leak, not a conversion problem.
- Who decides that day four is the day to chase? If the answer is a person's judgement, it will be inconsistent — not because they're careless, but because they're busy.
- Where does "interested, not now" live? If the answer is "in a colleague's memory" or "we just try again eventually," you're losing your warmest prospects to the least sophisticated part of your process.
What fixing it looks like
Deliberately unambitious, because ambitious versions don't get implemented.
Automate the reminder before you automate the message. The school above didn't ask for AI. They asked to be told that three days had passed. A nudge to a human is easier to trust, easier to switch on, and captures most of the value.
Give "not now" a home. One state, with a date attached. When the date arrives, someone gets prompted. That single field recovers more revenue than most automation projects.
Make speed structural, not heroic. The reason first-response times slip is that they depend on someone being free. Any first response that doesn't require a free human — even an honest holding reply with a real commitment attached — stops the "I never enquired" problem.
Then measure the stage, not the total. The 40–50% figure was only useful because it was attached to a stage. An overall conversion rate tells you nothing about where to intervene.
Honest assessment
- Automated follow-up applied to a bad list makes you a nuisance at scale. Fix intake quality first.
- Some drop-off is healthy. If a prospect was never going to buy, losing them early is efficient. The goal is to stop losing the ones who would have.
- This is one of the few problems where the software genuinely is most of the fix — the reminder either fires or it doesn't. But it only helps if someone acts on it, and if your team has no capacity, a reminder is just a new source of guilt. Capacity first, then automation.
LinkedIn native post (link in first comment)
"It drops off about — it ranges probably about 50%, 40%."
That's a school group telling me where applicants disappear. Where? "After initial inquiry. At the very beginning."
Four to five out of every ten people who raised their hand, gone at the first step.
Then he said the thing that actually stopped me: he described it as normal for the sector.
He's probably right. Which means an entire industry has normalised losing half its interested prospects at the cheapest possible point to save them.
So I asked what moves a lead from one stage to the next.
"We don't move them along. We have SLAs at different stages. If it doesn't get picked up within a certain period, the owner of that week gets a problem."
There's no mechanism. There's a person who's supposed to remember, and a consequence if they don't.
The system is a person. The backup system is a telling-off.
Meanwhile, an admissions manager at a different school: "You do not have that visibility or a reminder... it has been three days, do you want to follow up?"
She didn't ask for AI. She asked to be told three days had passed.
Three uncomfortable diagnostic questions 👇
LinkedIn article angle
Centre it on "memory as load-bearing infrastructure" — applies to far more than lead follow-up and is a strong think-piece. Title: When Memory Is Load-Bearing.
CTA
The three-question diagnostic → website.
Yalla is the chat-native CRM where the AI does the work, not just the talking.
See how much of your business it can reach.
or book a free demo