Helium – AI automation agency logo
Helium – AI automation agency logo
Helium – AI automation agency logo
Helium – AI automation agency logo

The Work That Only Happens Because Someone Chased

Take one person out for two weeks and a surprising amount of your business stops. Not because they do the work, but because they remind everybody else to.

There is somebody in your business whose real function is remembering. They chase the supplier, they remind the client, they ask whether the report went out, they check the thing that was supposed to happen on Tuesday.

It does not appear in their job description and it is a substantial part of why the business works.

Chasing is a system that was never built

Every instance of chasing marks a place where something was supposed to happen and nothing guaranteed it would.

The document that should have come back. The approval that should have been given. The engineer who should have filed the report. The client who should have returned the form. Each one is a step with no mechanism behind it, and a person is quietly acting as that mechanism.

The work is invisible because it produces nothing when it succeeds. Nobody notices the four reminders that prevented a delay. They notice the one delay when somebody was away.

Why it is expensive

It is unbounded. Chasing scales with volume and there is no natural limit, so it grows quietly until it is most of somebody’s week.

It is done by the wrong people. Frequently the most senior person available, because they have the standing to push. Chasing a supplier for a delivery note is not what you are paying a manager for.

It fails silently under pressure. In a busy month the chasing stops first, precisely when the volume is highest, and nobody reports that they did not get round to it.

It damages relationships. Four reminders in a fortnight makes you the business that keeps ringing, and the person doing it knows it, which is why they hesitate and why the fourth reminder often never goes.

Count it before you fix it

Ask two or three people to keep a note for one week of every time they chased something. Not who they chased. What it was and how many times.

Two things usually come out of that week. The volume is considerably higher than anybody guessed, and it clusters: a handful of recurring situations account for most of it, which makes the problem far smaller than it first appears.

Where AI removes the chasing rather than the chaser

This is the cleanest possible case for automation, because the task is remembering, tracking and prompting at a volume nobody can hold in their head.

AI knows what is outstanding. Every document requested and not received, every approval waiting, every promise made with a date attached. Assembled from what has actually happened rather than from a list somebody maintains.

AI chases on the schedule, without hesitating. The fourth reminder goes out exactly like the first. That alone changes outcomes, because the fourth is the one people skip and frequently the one that works.

AI writes something worth reading. Referencing what was actually asked for and when, rather than a generic reminder. A specific chase gets a response and a generic one gets ignored.

AI escalates on evidence. Three attempts with no reply becomes a person’s problem, with the history attached, rather than continuing forever or being quietly dropped.

The person is not replaced. The remembering is, which is the part that was consuming them and the part they were never hired for.

Fix the cause of the biggest one

Automating the chase is the right answer for most of the list. For the largest item, ask a harder question first: why does this need chasing at all.

Frequently the answer is a form that is too long, a request that arrives without a deadline, an approval routed to somebody who does not understand why it matters, or a supplier whose process genuinely does not include sending you that document.

Those are fixable at the source, and removing the need to chase beats automating it. Do that once, for the worst offender, then automate the rest.

Not everything should be chased

One caution, because automating this makes it effortless to send far more reminders than a person ever would.

A system that chases everything equally will chase your best client about a trivial document with the same persistence it applies to an overdue invoice, and the volume itself becomes the problem. Effortless chasing produces too much of it.

Decide what genuinely warrants pursuit, and what should simply be recorded as outstanding and raised in the next conversation. The test is whether a person would have chased it a fourth time. If not, the system should not either.

Chasing your own team is different

Most of this article is about chasing outward, to clients and suppliers. A meaningful share of it is internal, and that needs handling differently.

Automated reminders to colleagues read as monitoring if they arrive without context, and a team that feels tracked will start managing the reminders rather than the work. Aim them at the process rather than the person: what is outstanding on this job, visible to everyone, rather than a message telling somebody they are late.

Where internal chasing is constant, the cause is usually workload or an unclear handover rather than forgetfulness, and no reminder fixes either.

The tell that you have this problem
  • Somebody is described as being good at keeping on top of things. That is a person doing a system’s job.

  • Things slip when a specific person is away, in areas they do not personally work on.

  • Your team keeps private lists outside your systems. Those lists are the map of everything that is not tracked.

  • Deadlines are met by effort rather than by process, and everybody can name the fortnight when it nearly did not happen.

Any two of those and there is a real number sitting behind it, usually a day or more per week across the business.

AI Optimize tracks what is outstanding and chases it on schedule, in language worth reading, so the person who has been remembering can stop. That work sits under Workflow Automation.

Related reading

WHAT WE BUILD

This is the part we solve