Automate the Carrying, Never the Caring

At the edge the customer touches. The chatbot on the website, the phone menu, the automatic reply that acknowledges an email nobody has read yet, the portal where somebody can log a request instead of speaking to a person. That is where automation usually lands first, and there is a straightforward reason: it is the place where the cost of human attention is most visible. Somebody can count the calls. Somebody can count the people answering them. A number that can be counted is a number that can be reduced, and a project that reduces a countable number is an easy project to approve.

Behind that line, nothing changes. The fetching, the copying between systems, the chasing of a colleague for a figure that already exists somewhere, the retyping of an address from one screen into another. All of it carries on exactly as it did, because none of it was visible enough to be counted in the first place.

The result is the worst of both arrangements. The customer meets a machine at precisely the moment they had decided they wanted a person. And the staff, whose day was supposed to get better, still spend most of it as a message service between systems that do not speak.

The inversion is to leave the customer boundary alone and automate everything behind it. Every lookup, every hand-off, every piece of information that has to travel from one place to another before a person can act. Then, when a human being does get involved, they arrive already knowing, already able to do something about it, and already authorised to.

What that produces is not less human contact. It is better human contact, and less of the contact that never needed to be human in the first place. A business ends up spending its human attention on the moments that decide whether a customer stays, rather than on the moments a database could have handled without anybody noticing.

That is the whole principle, and it is short enough to be a rule: automate the carrying, never the caring.

Nothing on the left is a job anybody wanted. Nothing on the right is a job a machine does well, and the last line of it is the one that matters most: being accountable to a face is not a task, it is the reason the business gets to keep the customer.

It is the first thing people ask, and it is the right question, because that is exactly what most automation has done to most companies. The answer depends entirely on the order.

Automate the customer boundary first and yes, the business gets colder, and it gets colder in the most damaging way possible, which is at the moment of contact. That is the version everybody has experienced as a customer, and it is why the question is asked with suspicion rather than curiosity.

Do it in the other order and the opposite happens. The machine carries the information. The person handles the judgement, the exception and the apology. When somebody does pick up the phone, they already know who is calling, what was ordered, what went wrong and what they are allowed to do about it. That is the call nobody minds getting.

A technician arrives at a hospital to service a medical device. The job card on the tablet already holds the customer, the service history and the device itself, with the make, model and serial checked against the equipment register on the way in. The checklist for that exact class of equipment is already selected rather than chosen from a list. The parts offered are the ones that fit that device and no others. If the test equipment in the bag is out of calibration, the job cannot be closed, so a compliance problem that paper workflows leave open is simply not available.

Everything in that paragraph is carrying. None of it is caring. What the technician does is the actual work, the judgement about what the device needs, and then a signature captured on screen with a customer standing next to them. That is the Medeq field service system, and the line in the write-up that matters most is this one: consistency no longer depends on what a field agent remembers.

The client's own summary of the outcome, in a public review, was four words long. My customer loves me.

Smaller than anybody expects, and that is the point of putting this chapter first.

The instinct is to reach for a system. New software, a migration, a training plan, a go-live date. That is the version of this work that fails, and it does not fail for technical reasons. It fails because a business absorbs only so much change at once, and the cost of getting that wrong is not a delayed project, it is fear, a forced change of habits, and people leaving.

So the first thing is usually one piece of carrying that somebody currently does by hand. The document that is already sitting there waiting to be copied into an email. The spreadsheet somebody fills in every Friday from figures that exist in two other places. Nothing gets switched off, no new system arrives, nobody has anything to learn. One task that used to take a person twenty minutes stops needing a person at all.

We run this on ourselves, which is the version we can show you the inside of. When a piece of client work is marked finished in our project tool, the invoice raises itself, gets its number from the accounting system, and goes out with the PDF attached. Nobody retypes a line of it, and nobody has to remember it is Friday.

Because almost everything else in this series depends on it having gone first.

A business that has watched one piece of drudgery disappear without anything breaking is a different audience for the next conversation than one that has been asked to trust a plan. The relief comes before the change, and it buys the permission for everything after it. That is not a sales technique. It is what a business can actually absorb.

What comes next is the question that usually arrives about a week later, once the first bit of carrying is gone and somebody notices how much of it there was: why does growing feel so heavy in the first place?