Five questions for a system you already own

Every so often a business surfaces from a busy stretch and looks at its operations with fresh eyes. Revenue gets reviewed. Targets get set. And somewhere on the list is that thing built two or three years ago — the website, the portal, the reporting stack, the internal tool.

Should we do something about it this year?

It is a reasonable question. It is also usually answered on instinct. It looks a bit dated, so yes. It still looks fine, so no.

Neither answer is reliable, because neither is about whether it works. Here are five questions that produce evidence instead. All five can be answered in an afternoon, by you, without engaging anybody.

1 · When did it last produce the thing it was built to produce?

Every system was built to produce something. Enquiries. Applications. Bookings. Approved invoices. A number somebody uses on a Monday. Whatever it was, when did it last do it — not through a referral, not because somebody phoned, but through the thing itself?

If you can recall an instance from last month, it is working. If you have to think back further than a couple of months, that is a signal. If you genuinely cannot remember, that is the answer.

Something that has been live for more than six months and is producing nothing is not neutral. It is either not being found, not converting the people who find it, or not reaching the people it was built for. All three are fixable. None of them fixes itself with time, and this is the question most likely to be answered with a story rather than a fact — so answer it with a fact.

2 · Does it work in the conditions people actually use it in?

Not on the office connection, on a laptop, by someone who built it. On a phone, on mobile data, in the conditions the actual user is in when they need it — which for a great many South African users is a mid-range handset on a patchy signal.

There are free tools that will score this for you honestly, and the number they return is less interesting than the experience of running it yourself. Open the thing on a phone, away from the office, and try to do the one thing it exists to let someone do. If you find yourself explaining to nobody in particular why it is slow today, you have your answer.

Slowness is not a cosmetic problem. It is the most common reason a system that tests fine in a demo underperforms in the field, and it is very often fixable without replacing anything.

3 · Can you change anything in it yourself?

Your service descriptions, your contact details, your operating hours, your team list, your rates — can you change any of them right now, from a browser, without sending an email to somebody?

If not, you are not operating a system. You are operating a document that requires a specialist to edit, and which is therefore almost certainly out of date.

The consequence is not just inconvenience. Something that cannot be updated does not get updated. Descriptions that were accurate three years ago are still on the page. People who have left are still listed. This is visible to everyone who looks, and it undermines exactly the credibility the thing was built to establish. Where I find this, it is usually the most consequential finding of the five — and it is architectural rather than cosmetic, which means no amount of redesign addresses it.

4 · Can the people who need it find it?

Open a private browsing window, so your own history does not flatter the result, and search the way somebody who has never heard of you would search. Not your name — the problem they have and where they are.

Where do you appear? First page? Third? At all?

If you do not appear for the terms that describe what you actually do, you are invisible to the only group that represents new business: the people who do not already know you exist. This takes two minutes to establish and the result is rarely ambiguous.

The same question applies to internal systems, in a different costume: can the people who are supposed to use it find the thing they need inside it, without asking somebody? If the answer is a person's name, the system has a findability problem and that person is the workaround.

5 · Do you know who uses it, and what they do?

Measurement is free and has been for years. If yours was built in the last four or five years there is a reasonable chance something is already recording this. If it is, look now.

It will tell you immediately which of two very different problems you have. A lot of people arriving and nothing happening is a conversion problem — something about the experience is stopping interested people from acting. Very few people arriving is a reach problem. These need opposite responses, and choosing the wrong one is how a rebuild fails to change the numbers.

If nothing is recording anything, that is itself the finding: it was built without measurement, which means nobody has ever been in a position to say whether it works. That is worth fixing before anything else, because it is what makes every other decision evidence-based rather than aesthetic.

What to do with the answers

Three or more confident yeses and it is working. The conversation is about refinement — a design pass, fresh content, one new capability — not replacement.

Three or more no or I don't know and it is invisible, unchangeable, or unmeasured. Any of those means it is not doing the job it was built to do.

But the next step is still not automatically a rebuild. It is an honest look at what is actually broken, which of those problems are worth solving, and what the right size of intervention is. Sometimes that ends in replacing the thing. More often it ends in something smaller, cheaper and considerably less disruptive than the conversation started out assuming — and the businesses that get the best return are the ones willing to find that out before committing.

Answering these five is most of the diagnosis. It is worth doing before anyone, including me, tells you what you need.