A website is how people find you, understand you and get in touch. A system is what happens after that — the records, the rules, the roles and the things that need to happen without anyone remembering. One faces outward at strangers; the other faces inward at your own work. Most small businesses need a good website and no system at all.
First, two different things are called a "business operating system"
This trips people up before the real question starts, so it is worth clearing up.
Search the phrase and most of what comes back is about management frameworks — EOS, Traction and their relatives. Those are ways of running a company: meeting rhythms, scorecards, who is accountable for what. They are about people and habits, and they are not software.
The other meaning is software that runs the operational side of a business — your records in one place, with the rules and the routines built around them. That is what this page is about, and it is what we mean when we say "a system". We do not use the phrase "operating system" for it, because in practice nobody buying one calls it that.
If you came looking for the meeting-rhythm kind, this is not that page. No hard feelings.
What each one is actually for
A website has one job: turn a stranger into an enquiry, a booking or a call. It explains what you do, makes you look like a firm worth trusting, and gives an obvious next step. It is public, it is read once by each visitor, and nobody logs into it.
A system has four properties rather than a feature list, which we set out in full in our guide on moving from a spreadsheet to a system: one set of records everything else reads from, rules that hold so a typo gets refused rather than saved, roles so ten people can look and two can edit, and things that happen without anyone remembering — the reminder before the date, the record created when a form is submitted.
Notice these are not competing. They answer different questions. "Nobody knows we exist" is a website problem. "We know exactly who our customers are and we still missed the renewal" is a system problem. Firms get into trouble by spending on the wrong one.
The two questions that separate them
| A website | A system | |
|---|---|---|
| Who is it for | People who do not know you yet | You, your team, sometimes your clients |
| Does anyone log in | No | Yes — that is rather the point |
| What it holds | Pages that rarely change | Records that change all day |
| Succeeds when | More of the right enquiries arrive | Less goes wrong, and less is retyped |
| Cost shape | From £750 one-off, or from £2,950 for a larger build | A build, then a running cost |
| Fails quietly by | Looking dated and losing you enquiries | Nobody using it, so the spreadsheet comes back |
The quickest test is the login. If what you are describing needs someone to sign in and see only their own slice of it — a client portal, a booking diary, a member area, a staff tool — you have left website territory. We make the same point in our website cost guide, because it is the single thing that moves a quote from hundreds to thousands.
The second test is where the work actually hurts. If your admin pain would survive a beautiful new website completely untouched, a new website will not fix it.
Which do you build first?
Nearly always the website, and the reason is not aesthetic. A website is cheap, fast, and it brings in the work that pays for everything else. A system built before there is enough work to strain against is a system built on guesses.
We have watched this line get crossed three times, and can name all three because they are ours.
Our own business. Dopara started because Tiber Course needed a website. The site came first. The backend that runs the courses behind it came second, once there were courses to run.
Our first client. A UK property firm, for whom we built the brand and the website at the front and a full portfolio database behind it. Same firm, same studio. The public pages and the private system are two halves of one thing, which is exactly why we can describe the join honestly rather than in the abstract.
The third crossing. What we learned building that became TenureBook — the same job, but as a product you can simply buy instead of commission. Which is the strongest argument on this page against commissioning anything.
The exception to website-first: if the business cannot physically operate without the system, build the system. A firm whose whole service is the portal does not need a brochure about the portal.
What actually happens at the join
This is the part no one writes about, and it is where the money is either saved or wasted.
Done properly, the two are one thing with a door in the middle. The enquiry form on the public site writes a record straight into the system rather than an email into an inbox someone has to re-key. The login sits on your own domain, so signing in does not feel like being handed off to a stranger. One brand runs across both, so the private side looks like the public side rather than like software you bought.
Done badly, you get two projects that happen to share a logo: a site built by one supplier, a system built by another, and a person in the middle copying between them. That person is the thing you were trying to remove.
Our own bias, stated plainly: we build both, from the same desk, which is why we think the join matters. If you use two suppliers, agree in writing which one owns the handover before either starts.
What each costs
Websites start at £750 for a rebuild as a one-off, or from £2,950 for a larger build, always with a fixed quote agreed before anything is built. The full ladder, including what other UK agencies charge, is in the website cost guide.
Systems split in two. A Proprietary Database — your records, rules, roles and reminders in one private system — is £7,000 to build and £100 a month to run, one price, no per-seat charges. Anything larger or stranger is a bespoke system and has no shelf price; we scope the first working version with you and agree the cost in writing before you pay anything.
Budget for the running cost of both. A site nobody maintains goes stale as surely as a spreadsheet nobody owns.
When the answer is neither
Three cases where we would tell you not to commission anything, and we would rather say so now than after a deposit.
A product already does it. Booking, invoicing, payroll, helpdesk, email marketing — solved, cheaply, by companies who do only that. Buy it.
Your process is still moving. A system makes a process permanent. If yours changes monthly, a spreadsheet or a no-code tool is the cheapest way to find out what it should be. Our three-way comparison puts real prices against that middle option, and concedes that below a certain size it beats us on cost.
The pain is annoyance, not money. Price the annoyance in hours before you price a build. Annoying is not a business case.
Working out which you need, in ten minutes
- Write down the last five things that cost you money. If most are "nobody found us", that is a website. If most are "we knew and still missed it", that is a system.
- Ask whether anyone needs to log in. If nobody does, stop — you want a website.
- Ask whether a ready-made product already does the job. If one does, buy it and build neither.
- If you need both, do the website first unless the business cannot run without the system.
If you would like a second opinion from a studio that builds both and will happily tell you that you need the cheaper one, book a short call or tell us what your business tracks. If you are still working out which half of this you are in, everything we build is here.