The short version
- Price tracks the build method, not the page count. There are four methods and they cost very different amounts.
- Fitting a template runs about KES 10,000 to 35,000. Coding a site from scratch runs about KES 90,000 to 350,000.
- Cheap is often the right answer. It stops being the right answer the moment the site is supposed to bring in work.
- Ask how it will be built before you ask what it costs. The first answer explains the second.
If you run a business in Nairobi, Mombasa, Kisumu or anywhere else in Kenya, you have met the gap. One developer will build you a website for KES 15,000. Another quotes six hundred thousand. Nobody explains the difference, so most owners assume somebody is overcharging and pick on price alone.
Almost every price guide sorts websites by type: brochure site, business site, online store. That is close to useless, because two “business sites” can differ tenfold in price and neither seller is lying. The number that actually moves is how the thing gets made. Sort by that instead and the market stops looking irrational.
Why the gap exists at all
Because the word “website” covers at least four different trades, and the market sells all of them under one name.
Buying a finished template and swapping in your logo is a two-day job. Designing, coding and testing something that did not exist before is a four-week job. Both end with a site at an address, and both get quoted as “a website”. The person charging KES 15,000 is usually not cutting corners on the four-week job. They are doing the two-day job, honestly, at a fair rate for it.
Your risk is not being overcharged. It is buying the two-day job while expecting the four-week outcome.
What it costs, by how it is built
These bands are drawn from prices published across the Kenyan market for 2026 and sit at the lower end of what those sources report, because published ranges tend to be optimistic at the top. Treat them as a map, not a quote. Nobody can price your project without seeing it.
| How it is built | Typical range (KES) | What is actually happening |
|---|---|---|
| Theme fitting | 10,000 – 35,000 | A finished template, your logo and text dropped in |
| Page-builder assembly | 35,000 – 90,000 | Blocks arranged by drag and drop, features added as plugins |
| Custom studio build | 90,000 – 350,000 | Designed for your business, then written as code |
| Product engineering | 350,000 – 1,200,000+ | Accounts, databases, payments, things that must not fail |
The four methods, and what each is good for
1. Theme fitting
Someone buys a template built for a generic business in your industry, replaces the demo photographs and text with yours, and hands it over. Fast and cheap, and for a business whose current alternative is a Facebook page, genuinely a step forward.
What you are trading away: the layout was designed for someone else, so your actual strengths sit wherever the template happened to put them. The site usually cannot be extended, and whoever fitted it often understands it no better than you do once something breaks.
2. Page-builder assembly
Built by dragging blocks around inside a builder, with each extra capability bolted on as a plugin. More flexible than theme fitting and you can often edit it yourself, which is a real advantage.
What you are trading away: speed and durability. Every plugin adds weight and another thing that needs updating, and a stack of them is why so many Kenyan sites take five seconds to open on mobile data. Plugins also go unmaintained, and an unmaintained plugin is somebody else's door into your site.
3. Custom studio build
The site is designed around what your business actually does, then written as code. No template underneath, so nothing is inherited and nothing is carried that you do not need.
This is where being found starts to be part of the build rather than a service sold afterwards: clean structure, real headings, structured data, and pages fast enough to hold someone on a weak connection. It is also where a content manager usually appears, so your team can publish without calling anyone.
What you are trading away: time and money up front. Weeks rather than days.
4. Product engineering
Now there are user accounts, a database, payments, and rules the business depends on. Different people see different things. Money moves. This is not a website with extras bolted on, it is software, and it is priced as software. We drew that line in detail in website versus web app.
What it costs, by what you need
The same money viewed the way most people actually shop. Find your row here, then check which method column it lands in above. Where the two disagree, the method decides the price.
| What you need | Typical range (KES) | Usually built by |
|---|---|---|
| One-page or personal site | 15,000 – 45,000 | Fitting or assembly |
| Small business site, up to 6 pages | 45,000 – 120,000 | Assembly, or a small custom build |
| Business site with a content manager and SEO | 120,000 – 300,000 | Custom studio build |
| Online store | 80,000 – 350,000 | Assembly at the low end, custom above it |
| Platform, portal or application | 350,000 – 1,200,000+ | Product engineering |
How to tell which one you are being sold
Quotes rarely say which method they are using, and the difference is worth several hundred thousand shillings. These questions surface it in about a minute.
- “Can I see this exact design on another site?” If a template is involved, it is somewhere. A reverse image search of the portfolio piece often finds it.
- “What happens if I want to move a section?” Fitting and assembly answers sound like the tool's limits. A custom answer sounds like a design decision.
- “How many plugins or add-ons will this need?” A specific number means assembly. Confusion at the question usually means custom.
- “What will it score on PageSpeed when you hand it over?” Anyone who has not thought about performance will not have a number.
- “Which single item here would you cut to save me money?” Anyone who genuinely scoped your project answers in about four seconds.
None of the answers are wrong answers. You are establishing what you are buying, so the price has something to attach to.
Where we sit, so you can place us
The tables above are the market, not us, and we did not put ourselves at the top of them. To be straightforward about it: ByteFrame does custom studio builds and product engineering. We do not fit themes and we do not assemble page builders.
That is about method, not about project size. Small is welcome. Our entry tier is a single-page site, typically delivered in one to two weeks, and we build those for consultants, creatives and people who just need one sharp page that loads fast and is actually theirs. A small project built properly is a much better starting point than a large one built badly.
What we will not do is quote you for more than the job needs. If you come to us describing a platform and a good landing page would serve you better this year, we will say so before you spend anything.
What comes as standard on everything we build: a design made for your business rather than adapted from a theme, a PageSpeed score of 90 or above tested on throttled connections rather than our own machines, full handover with no vendor lock-in so the code and every credential is yours, a fixed quote agreed in writing before work starts, a 30-day post-launch warranty, and the two founders doing the work rather than briefing someone else to. The tiers are laid out on the services page.
The costs that never appear on the quote
Not hosting and domains, which everyone lists. These are the ones that do not look like costs until you add them up.
The enquiries you never knew you missed
This is the largest cost in the whole article and it is invisible by nature. A site that does not surface in local search does not produce a bill or an error message. It produces silence, and silence is indistinguishable from a quiet month in your market. Businesses can run for years without ever learning what the site should have been bringing in.
Your own hours
Count the time you personally spend chasing a developer who has gone quiet, re-explaining a change that came back wrong, or logging in to a system that fights you every time you update a price. It is not on the invoice, and for an owner it is usually the most expensive input in the project.
Paying for the content twice
Most quotes assume you are supplying copy and photographs. Most businesses discover in week three that they have neither. Whatever you then spend on a photographer and a writer is part of the true cost of the site, and it lands after you have already committed.
Losing the keys
If your domain sits in your developer's account, they control your web address and your email, not you. When that relationship ends badly, recovering it ranges from tedious to impossible. Register the domain yourself, in your own account, before anything is built.
The rebuild you are already funding
A site that cannot be extended has a fixed lifespan. The day you want a booking flow, a portal or payments, you are told it has to start again. The original fee was not a discount on a website. It was a deposit on the next one.
Five questions before you sign anything
Ask these of everyone you are considering, including us.
- Which of the four methods is this? Everything else follows from the answer.
- What exactly do I own at handover? The code, the domain, the hosting account, the analytics, the design files. Ask for the list in writing.
- How does this site get found? If the answer is an SEO package sold separately later, structure was not part of the plan.
- What does a change cost in month six? A quote that ends at launch has not told you the real price.
- Who is actually doing the work? Whether the person answering your questions is the person writing the code changes how much gets lost between the two.
Common questions
How much does a website cost in Kenya in 2026?
It depends on how it is built rather than how many pages it has. Fitting a bought template runs about KES 10,000 to 35,000. Assembling a site in a page builder runs about KES 35,000 to 90,000. A site designed and coded for your business runs about KES 90,000 to 350,000. A platform with accounts, payments and data starts around KES 350,000.
Why do some developers charge KES 15,000 and others charge half a million?
Because they are doing different work. At the low end someone is fitting a template that already existed, which takes a couple of days. At the high end someone is designing, coding and testing something that did not exist before, which takes weeks. Both are honest trades. The problem is that the market sells both under the same word.
Is a cheap website in Kenya worth it?
Sometimes, genuinely. If your alternative is no website at all, a template site is a real step forward. It stops being worth it when the site is expected to bring in enquiries, because being found and being convincing are the parts that template fitting does not include. If you are weighing the two, ask someone who builds both kinds and will tell you which your situation calls for. We would rather scope you a small project that works than sell you a large one you did not need.
What are the ongoing costs of a website in Kenya?
Budget annually for a domain at roughly KES 1,000 to 3,000 and hosting at anywhere from nothing for a static site to about KES 25,000 for shared or cloud hosting. Business email on your own domain is a few hundred shillings per user per month. An SSL certificate should be free and automatic in 2026, so a yearly charge for one is worth questioning.
Decide the method, then the budget
Most people start with a number and try to find out what it buys. Start at the other end. Work out what the site has to do for the business, which method that requires, and let the budget follow. Done in that order the figure is usually less alarming than the guessing was, and occasionally the honest answer is the cheap one.
Want a real number for your project? Send us the shape of it: what the business does, who the site is for, and what you want a visitor to do. Tell us about it here and we will come back with a scoped, fixed quote within a business day, including the times we think you need less than you asked for.
Wilfred Gweth
Wilfred co-founded ByteFrame in 2025 and handles design and front-end engineering on every build. Meet the team.