For a large part of my career, finishing a website felt like reaching the end of a project.
You designed it, built it, added the content, tested everything and eventually changed the DNS.
Then came that slightly nerve-racking moment when the new website went live.
Job done.
Except it almost never was.
After more than 20 years of building websites, ecommerce systems and various bits of software, I've gradually come to think about websites very differently.
The pages somebody sees in their browser are important, but increasingly I think of them as the visible surface of something much bigger.
The interesting bit is often what happens underneath.
Web Design Used to Be About Pages
When I started building websites, the job was relatively easy to explain.
A business needed a presence on the internet.
You created a homepage, an about page, a services page and a contact page. Perhaps there was a gallery or a news section if they were feeling adventurous.
If the business was local, the website might be designed to attract people searching for things such as web design in Felixstowe, SEO in Suffolk or another service in a particular town.
The website itself was the project.
That changed surprisingly quickly.
Businesses started expecting their websites to do things.
Customers wanted to buy online.
Businesses wanted to manage their own content.
Orders needed storing.
Payments needed processing.
Enquiries needed tracking.
Google Ads needed landing pages.
SEO needed content.
Suddenly the website wasn't just a collection of pages.
It was becoming software.
Ecommerce Made This Impossible to Ignore
Ecommerce websites were probably where this became most obvious to me.
From the outside, an online shop can look relatively straightforward.
There are products.
There is a basket.
There is a checkout.
Somebody pays.
But underneath that simple journey is a surprising amount of machinery.
You have products, variations, stock, customers, addresses, orders, payments, emails, discounts, shipping rules and potentially thousands of individual transactions.
Then somebody asks a perfectly reasonable question such as:
"Can we automatically do this when somebody buys that?"
And suddenly you're building another system.
I loved that part.
The Best Improvements Often Had Nothing to Do With Design
When somebody hires a web designer, it's natural to think visually.
What colour should the buttons be?
Where should the photograph go?
Should the navigation be across the top?
What should the homepage say?
Those things matter.
But some of the most useful improvements I've ever made to websites would be almost invisible to a visitor.
Maybe an order automatically updates something that previously had to be entered manually.
Maybe a food order automatically prints out in the restaurant's kitchen.
Maybe information from one system appears in another.
Maybe a repetitive email gets generated automatically.
Maybe a dashboard turns several database queries into one useful answer.
Maybe a scheduled process does at six in the morning what somebody previously spent half an hour doing every day.
None of those things will win a web-design award.
They can still transform how useful the website is.
SEO Taught Me Something Similar
Search engine optimisation has its own visible surface.
People see the Google result.
They see the page title and description.
They search for something like SEO near me, Suffolk SEO, Felixstowe web design or ecommerce websites and choose where to click.
But again, what happens behind that simple interaction is much more interesting.
Google has to understand the structure of the website.
It needs to discover the pages.
It needs enough context to understand what those pages are about.
The website needs to respond quickly.
Internal links need to connect related information.
And most importantly, there needs to be something worthwhile waiting for the person who actually clicks.
The visible result is only the final few centimetres of a much longer process.
Google Ads Could Hide Bad Systems for a While
I also spent years around Google Ads and PPC.
Paid advertising is brilliant at exposing another uncomfortable truth about websites.
You can buy traffic.
You cannot buy a good experience after the click.
If the landing page is confusing, the traffic is wasted.
If the checkout is broken, the traffic is wasted.
If nobody follows up an enquiry, the traffic is wasted.
If the business can't tell which enquiries turned into customers, it becomes difficult to know whether the advertising worked at all.
That is why I gradually became less interested in traffic as an isolated number.
I wanted to know what happened next.
That Question Still Drives Most of What I Build
What happens next?
It is probably one of the most useful questions I've taken from years of web development.
Somebody visits the page. What happens next?
Somebody buys the product. What happens next?
A payment succeeds. What happens next?
A new piece of data arrives. What happens next?
A piece of content starts getting traffic. What happens next?
A customer needs attention. What happens next?
Once you start thinking that way, you stop seeing isolated web pages.
You start seeing workflows.
My Own Projects Gradually Became Connected Systems
This is probably why the things I build for myself today rarely remain simple websites for very long.
I might start with one small requirement.
Then I realise I need to store some data.
Once the data exists, I want to analyse it.
Once I can analyse it, I want the system to highlight what matters.
Then I notice a repetitive job and automate it.
Then another part of the business needs access to the same information.
Before long, what started as a page has become a small operating system for that particular project.
That progression feels completely natural to me now.
It probably wouldn't if I hadn't spent so many years building websites for businesses first.
Building Admin Systems Changed My Perspective
There is a strange satisfaction in building an admin screen nobody else will ever see.
Visitors don't care how efficiently I can manage something.
Google doesn't rank a private dashboard.
There is no reason for it to look impressive in a portfolio.
But if that screen saves me ten minutes every day, it matters.
If it shows me something I couldn't previously see, it matters.
If it stops me forgetting something, it matters.
If it turns hundreds of rows of database information into one obvious action, it matters.
I've come to value that sort of development enormously.
A Website Can Become Institutional Memory
Another thing I've learned is that a well-built system remembers things humans don't.
People forget when something happened.
Databases don't.
People forget what a number looked like three months ago.
Databases don't.
People develop theories about why something is happening.
Data can sometimes prove them completely wrong.
That is one of the reasons I've become increasingly interested in collecting useful first-party data inside my own projects.
Not because having enormous databases is impressive.
Because historical information creates context.
Today's number becomes much more useful when you know what a normal Monday looked like six months ago.
A search query becomes more interesting when you can see whether its impressions are growing.
A sale becomes more useful when you can connect it to the content that helped produce it.
This Is Also Why I Still Care About Search Data
I no longer look at phrases such as Suffolk SEO, SEO near me or Felixstowe web design and immediately think about creating a sales page.
But I absolutely still care that those searches exist.
They tell me something about how Google understands the history of the website.
They show which subjects it still associates with my work.
And sometimes they uncover old areas of experience that are worth writing about again from a completely different perspective.
Search data doesn't have to dictate what you create.
It can simply help you notice things.
I'd Build Differently If I Started Again
If I were starting a web business from scratch today, I would spend less time thinking about individual websites and much more time thinking about systems.
I would still care about good design.
I would still care about SEO.
I would still want ecommerce websites to convert and PPC landing pages to perform.
But I would ask more questions about what sits behind them.
- What information should we be storing?
- Which repetitive jobs could disappear?
- What decisions could the data help us make?
- Which systems should communicate with each other?
- What happens after somebody converts?
- What are we learning that will still be useful a year from now?
Those questions can create much more value than another redesign.
The Front End Is Still Important
None of this means I've stopped caring about what a website looks like.
I haven't.
A confusing interface can ruin an excellent system.
A badly designed ecommerce site can make brilliant technology irrelevant.
A page still has to communicate clearly.
People still have to understand where to click.
The difference is that I no longer see the interface as the entire product.
It is the part of the product somebody happens to be looking at right now.
That's Probably the Biggest Change in How I Build
Twenty years ago, I would have been quite happy to say I built websites.
Technically, I still do.
But it no longer feels like an accurate description of what interests me.
I like building the machinery around them.
The databases.
The automation.
The analytics.
The workflows.
The little tools that remove irritating jobs.
The systems that collect information today so I can make a better decision tomorrow.
The public website is still important.
It's just no longer where I think the interesting part ends.
More often than not, it's where it begins.
