Two quotes on the desk. One is roughly half the other, and the cheaper one lists more pages.
The obvious reading is that you are looking at the same product at two prices, and that one supplier is greedier than the other. That reading is wrong often enough to be expensive. A cheap website is not a smaller version of a good one. It is a different object with a shorter life, and the difference does not show up in month one. It shows up around year three, to a buyer you will never meet and who will never tell you.
We know what year three looks like because we went and checked. Across 54 Singapore B2B sites we reviewed in August 2026, covering marine, construction, engineering and logistics, every fault described below was verified on the live site. These are not bad companies. Most of them are very good at the thing they actually sell, which is rather the point.
What does a cheap website actually cost?
Three numbers, and the quote only shows you one.
The build. The figure everybody negotiates. It is also the only one of the three with an end date.
The running cost. Hosting, domain renewal, TLS certificates, plugin and theme licences, and the labour of keeping all of it current. This one never stops. On a cheap build, nobody has been made responsible for it, which is not the same as it being free.
The cost of not working. The enquiry that did not arrive because a page errored, a certificate expired, or the contact details were wrong at the moment somebody checked. You never see this number, and that is exactly what makes it dangerous. A website that is quietly failing looks identical from the inside to one that is quietly working. Both of them sit there. Only one of them is answering.
Cheap builds are priced against the first number and judged, eventually, on the third.
Year three, in the field
Here is what a site with nobody minding it looks like after a few years. Every figure in this section is ours, from the review described above, and each uses its own denominator because the batches did not all check the same things.
12 of the 54 sites were running an outdated or end-of-life CMS or plugin. Not a version behind. Versions whose security support ended years ago, including two separate installs of Joomla 1.5, which has been unsupported since 2012.
7 of the 54 had a contact page or a navigation link returning a server error or a 404. The contact page. The one page a buyer goes to only when they have already decided to talk to you.
6 answered on only one of the www and non-www forms of their own domain, so half the ways a person might type the address led nowhere.
5 threw a browser certificate failure, which in a current browser is not a small warning icon. It is a full-page interstitial telling your customer that your site may be trying to steal from them.
3 still carried live template placeholder content, the kind that ships with a theme and is supposed to be replaced on day one.
2 had a domain that did not resolve at all, while still ranking in Google under the company name. Somebody searched, clicked, and got nothing.
And of the 23 sites whose footer we examined, 19 carried an out-of-date copyright year. The oldest read 2012.
Not one of those is a design problem. Every one of them is an upkeep problem, which is to say a budget problem that was settled at the quote stage and then forgotten about.
Software does not sit still, and browsers move faster
The reason a website decays rather than simply ages is that everything underneath it keeps changing while the site stays where you left it.
WordPress ships security fixes to old versions for a long time, but not forever. In 2025 the WordPress security team ended security updates for versions 4.1 through 4.6, and that line will keep moving up. PHP, the language nearly every one of these sites runs on, gives each branch 2 years of active support and 2 more years of security-only fixes, after which it is done. Anything older than 8.2 is now end of life.
Certificates are about to get harder. The CA/Browser Forum has agreed a schedule that takes the maximum life of a public TLS certificate from 398 days down to 47 days, phased in between 2026 and 2029. A certificate that used to need renewing about once a year will need renewing close to 8 times a year. If that renewal is automated and monitored, you will not notice. If it is a calendar reminder on the laptop of a freelancer you last spoke to in 2022, you will notice, and so will everybody who visits that week.
None of this is exotic. It is just ordinary infrastructure moving at its ordinary speed, and it is completely indifferent to what you paid.
The line that is missing from the cheap quote
Read a low quote closely and the omission is rarely in the page count. It is in the sentence that should say who owns the site after handover.
A serious quote tells you what happens the day after the support window closes. Ours includes 30 days of post-launch support covering bugs, minor content edits and questions, and then hosting from S$200 a year and maintenance from S$200 a month if you want us holding the updates, backups and monitoring. Those are starting figures, and what moves them is how much the site actually carries.
You do not have to buy that from your web supplier. You can run it internally, or hand it to your IT vendor. What you cannot do is not decide, because the default answer is nobody, and nobody is how a site ends up with an accreditation certificate that expired nine years ago as the newest dated item anywhere on it. We found exactly that, sitting above a template from a platform that went out of support in 2012.
What the difference in price is actually buying
Not more pages. Usually fewer.
The gap between a cheap build and a considered one is mostly work that happens before anything is designed: deciding what the site has to prove, gathering the technical content that proves it, and cutting everything else. That work is invisible on a quote and it is the first thing removed when the number has to come down.
When we rebuilt THK Engineering, an M&E contractor running electrical, ACMV, fire protection, plumbing and ELV, the useful decision was not visual. It was structural: present the five disciplines as one package a main contractor can hand over whole, because the question being asked at tender is whether any of it will get sublet to someone nobody has vetted. The same principle drove Hong Hock Global, where certifications and BCA-certified personnel moved to the front because the work is awarded on prequalification.
Neither of those is expensive because of the design. They are expensive because somebody had to sit down and work out what the buyer was actually deciding.
Where to spend if the budget is genuinely tight
Buy less scope, not less care. This is the honest version of the advice and it is not a sales line, because it makes the project smaller.
Five pages built properly, with real content and someone responsible for upkeep, will outperform twenty pages of filler every time, and it will still be working in 3 years. Twenty pages of filler is twenty pages to keep current, which nobody will, and it is also twenty chances for a buyer to hit a broken one.
The other honest thing to say is that a rebuild is not always the answer. If the structure is sound and the content is current, a refresh does the job. Our pricing bands are published rather than quoted on request, starting from S$4,000 for a refresh where the content already exists, and we have written separately on what a B2B website actually costs in Singapore if you want the full breakdown.
The point of this piece is narrower. Whatever you spend, spend it knowing that the build price is the smallest number in the project, and that the one you never see is the one that decides whether the site was worth building at all.