What we got wrong about our service list

Something in the business is not working and nobody can name it. So you call three suppliers, and each one names it for you. The web studio finds a website problem. The brand agency finds a positioning problem. The marketing firm finds a lead problem. Three diagnoses, no examination, and every one of them arrives attached to a quote.

We made that easier for a while. Then we made it harder in the wrong way.

The first version of this article argued that our service list had too many things on it. Seven, cut to four. A long list signals availability rather than capability, it said, and breadth invites the buyer to compare on price. We have since put sixteen things on a brochure, so either we changed our story or the first argument was solving the wrong problem. It was the second, and the error is specific enough to be worth writing down.

Is a long agency service list a problem?

Only if you can order from it.

That is the whole test, and it applies to any agency service list including ours. Can a client buy a single item off that list without the studio having looked at their business first? If yes, it is a menu, and the length of a menu is a real problem. If no, it is the range of things a practice can reach for once it knows what is wrong, and the length of that is not a claim about anything.

What was actually wrong with seven services

Not the count. The fact that each one was orderable.

They were seven products with their own names, sold as separate things a client could choose between. Someone could arrive, pick social media off the page, and never be asked what was wrong with their business.

That is what recruits the price comparison, and the count only changes how often it happens. Two studios selling a unit with the same name have exactly one variable left between them, and it is not craft. Cutting seven to four removed three chances to have the wrong conversation and left the mechanism intact.

The uncomfortable part of the first version stands, so it stays. One or two of those seven were thin. One we subcontracted almost entirely and put our name on. Nobody was defrauded and the work was fine, but we were selling access to somebody else’s craft and describing it as ours. That was a real fault. It just was not the fault the article said it was.

Why a narrow practice keeps finding the same problem

Because it can only write one prescription.

A studio that builds websites diagnoses website problems, and mostly not dishonestly. It is what the instrument is for. Theodore Levitt’s marketing myopia is usually read as a warning to the seller, but it reads as well from the buyer’s chair: a supplier who defines itself by what it makes will describe your situation in those terms.

The clearest evidence we have of this is our own, and it is evidence against us. Across 54 Singapore B2B sites we reviewed in August 2026, the marine group of 19 had a dominant failure that was not visual at all. Six of them published contact details in a state no redesign repairs: eight email addresses and eleven phone numbers with no guidance on which to use, and a placeholder sitting live beside a genuine one. A redesign tidies that page. It does not decide who owns the reply, or how fast it goes out. Two years ago we would have quoted the redesign, delivered it properly, and called it solved.

That is what our own headline, the website was never the problem, is reacting to. It is not an argument that websites do not matter. We still build them. It is an admission about what a narrow practice is able to see.

Sixteen things is a range, not a menu

The list is long now. Custom software, AI tools and products, web and mobile apps, chatbots, field service systems, AI training, websites, infrastructure support, brand identity, graphic design, film, photography, social channels, paid campaigns.

Nobody orders from it. Every engagement opens with a diagnostic instead: how work actually enters the business, who touches it, where it stalls, and what it costs when it does. The output is a written roadmap you own whether or not you build anything with us. Only after that does the list come out, and the client is not the one choosing from it. The diagnosis chooses.

Hong Hock Global is the plainest version of that. Water infrastructure work is awarded on prequalification rather than persuasion, so the answer pointed at a document rather than a brochure: certifications, live contract scope and technical method stated in the order a PUB officer or main contractor reads them, with emergency response on its own route, because that reader arrives with a burst main and no patience for navigation. That job happened to be a website. The one before it was a quoting tool that replaced a three-day email chain. A list has to be long enough to hold whichever answer turns out to be true, or the answer gets decided before anybody looks.

What a long list still does not excuse

The test cuts both ways, and we can fail it.

If a client can write in, buy social channels and receive it without anyone examining the business, it is a menu again and the first version of this article lands squarely. What holds the line is unglamorous: the diagnostic comes first, the roadmap is yours either way, and part of the job is saying the answer is not something we sell. We tell people a problem does not need AI more often than we tell them it does.

Depth still has to be shown. A range you cannot evidence is a list of things you are willing to attempt, which was true at seven items and is true at sixteen. Since 2023 we have run more than 100 projects across 70 clients, and that record is what makes the length defensible rather than boastful. Our pricing is published rather than quoted on request, web projects start from S$4,000 depending on how much strategy and content already exists, and every build carries 30 days of post-launch support, or 45 days on the larger ones.

Count the list if you like. The count will not tell you anything. Ask whether you can order from it.