Website, web app, or SaaS platform?
These three words get used interchangeably and cost wildly different amounts. Picking the wrong one is the most expensive mistake available at the start of a project — in both directions.
Last updated 3 August 2026 · about 7 minutes
The short answer
If nobody logs in, you need a website. If people log in and it remembers their data, you need a web app. If strangers can sign up and pay to use it themselves, you need a SaaS platform. Those two questions — does someone log in, and can a stranger buy it without you — separate almost every project correctly.
The three shapes
Ignore the technology for a moment. What separates these is not the stack, it is what the thing does and who it does it for.
A website
Explains something to a visitor.
Everyone who visits sees the same thing. It has pages, not screens, and nobody logs in. A brochure site, a portfolio, a restaurant menu, a landing page for a campaign. The measure of success is whether a stranger understands you and gets in touch.
- Nobody needs an account
- The content is the same for every visitor
- You want to be found on Google
- Success looks like an enquiry, a call or a booking
A web app
Does a job for the person using it.
People log in, and it remembers them between visits. It holds data, changes what it shows based on who is looking, and usually replaces a spreadsheet or a pile of paperwork. An internal dashboard, a booking system, a client portal, a stock tracker.
- Users log in and have their own data
- It stores and updates records over time
- Different people see different things
- Success looks like work getting done faster
A SaaS platform
Is the business, not a tool for it.
A web app that strangers can sign up for and pay to keep using. Everything a web app has, plus the commercial machinery: sign-up, subscriptions, billing, plan limits, an admin view, and support for many separate customers who must never see each other's data.
- Strangers sign up without you involved
- Customers pay you monthly or yearly
- Multiple unrelated companies use the same system
- Success looks like recurring revenue
The two-question test
Almost every project sorts itself with two questions, asked in this order.
One: does anybody log in? If no, you need a website, and you are probably about to be quoted far less than you feared. If yes, keep going.
Two: can a stranger sign up and pay for it without you doing anything? If no, it is a web app — built for your team or your existing clients. If yes, it is a SaaS platform, and the billing, plans and account management are not details to add later. They are half the product.
A useful sanity check: if you cannot describe who logs in and what they see when they do, you do not yet need a web app. You need another week of thinking, which is free.
The expensive mistake in each direction
- Building an app when a website would do. By far the more common and the more costly. Someone describes a business that needs to be explained well and sold to strangers, and ends up commissioning a platform with accounts and dashboards nobody asked for. Every login screen is work that a brochure site does not need, and none of it helps a customer find you.
- Building a website when you needed an app. Rarer, and it usually shows up as a site propped up by spreadsheets and manual email. If your team is copying form submissions into a spreadsheet every morning, you have already outgrown the website and are paying for it in staff hours instead of software.
- Building a SaaS before anyone has said yes. The most dangerous version. Subscriptions, plan tiers and multi-tenant data separation take months, and none of it proves anyone wants the product. Sell it to one customer as a web app first. If they pay, you have a business worth building the machinery for.
What each one costs in time
Rough shape, because the ranges are genuinely different rather than slightly different:
| What it is | Typical build | Priced as |
|---|---|---|
| Website | Days to a few weeks | Fixed price, fixed date |
| Web app | Several weeks to a few months | Quoted per project |
| SaaS platform | Months, then continuously | Quoted per project |
Websites can be quoted as a fixed price against a fixed date because the work is well understood. Apps and platforms are quoted per project because the honest answer depends on what yours does — anyone offering a flat number for “a SaaS” before asking what it does is guessing. For the website numbers, see what a website costs and how long one takes.
Start smaller than feels right
The three shapes are not a ladder you have to climb in order, but they do stack neatly. A website can be extended into a web app later. A web app serving one client can become a SaaS product once several want it. Nothing is wasted by starting at the smaller end, and quite a lot is wasted by starting at the larger one.
The reverse is where money disappears. Building the platform first means spending months on billing, permissions and admin screens before a single customer has confirmed they want the thing at all. If you are unsure which of these you need, the answer is almost always the smaller one — and you will know within a few months whether you have outgrown it.
Still not sure which one you need?
Describe what you want it to do in plain words. We will tell you which of the three it is, and quote the smallest version that actually solves the problem.