Skip to content

Let’s be real. Most companies don’t need something bespoke and complex. Even large corporations sometimes need “just a website” to establish their online presence.

This, of course, largely depends on the purpose their online presence serves. Sometimes it’s only there so people can read information about their services or products. A slider here, an image there, marketing copy everywhere, done.

But then you’ve got the other kind: the ones that use their website as a service. As a tool. As an application. They’ve got member logins, dashboards, integrations with third-party systems, automations, you name it.

These are not “websites”. They’re complex applications appearing as simple pages to the untrained eye. And what they need to stay afloat is much more than what a simple website does.

Usually we’re talking about sites with tons of information flowing into them from multiple sources: CRM systems, data files, proprietary applications with custom endpoints. Data gets pulled into the CMS, transformed, adapted, and presented to users in various formats, views, and even multiple languages.

This is not “web pages” anymore. This is much more.

And with that comes the responsibility of maintaining the system, keeping it secure, adapting it to changes that may come from any direction, and preserving its readiness for expansion.

Some examples that we’ve seen so far:

  • A foundation listing thousands of grants originally coming from their internal CRM database, absorbed into the CMS, transformed where needed, with relations between entities (such as grantees and categorizations), and enriched within the CMS with more data not available in the original CRM entries, while maintaining integrity so they can be updated at any time.
  • An online lecture system where editors link videos to (sometimes multiple) speakers and slides so each slide is presented at specific video points, with multi-level categorization and tagging.
  • A financial institution that automatically updates hundreds of documents from external sources daily, using taxonomies and placing them into sorted tables for easy reference.
  • A product listing pulling product information and media from external databases maintained by other teams, transforming them into visually appealing information for website visitors while maintaining a complex filter system that adapts to the current set of product attributes. (Worth saying: it was not an e-shop).
  • An educational organization running multi-question quizzes with different grading systems and completion milestones, allowing editors to create new quizzes on the fly with a variety of question and answer types, from multiple-choice questions with relative answer weights to text-only inputs to image selections. Each student tracks their own progress and selects the next test to take.

If all of the above sound familiar, it’s because they are. Many agencies work with such complex CMS implementations every day.

What’s not obvious here is what happens after delivery.

The “after delivery” maintenance nightmare

For a plain brochure website, things are simple: do minor updates and patches every now and then to keep it secure, make sure it runs smoothly, ensure editors can work without issues (that is, if they change their content at all). Done and done.

For a complex implementation, though, things get a bit more complicated. You’ve got to take care of:

  • Content architecture
  • Software architecture
  • Scaling when needed
  • Maintainability
  • Technical debt
  • Upgrades
  • Security
  • Accessibility
  • Performance
  • New features
  • Teams of editors with specific permissions
  • Content approval workflows
  • GDPR (where applicable)

Sure, you can find some of these things on simple websites as well, but to a much lesser extent.

What’s certain when you have such a project in your hands is that you need continuous maintenance and support. Which, in turn, brings you the most desired thing in the software world: stability.

And this doesn’t come cheap. In fact, maintenance costs can be higher than development costs in some cases. Not without reason, since websites are living systems - they must change, evolve, and expand, as well as follow framework and system developments over time.

The AI trap

The AI hype, at the time this post is written, is over the limit - and while this helps AI growth, it can create false expectations when it comes to such complex implementations. The most common ones are:

“Code is now cheap, developers are extinct - AI can do the job at a fraction of the cost, very quickly.”

Let’s be realistic: AI can indeed speed up mundane tasks and even free you from some complex ones, but an implementation like the ones laid out above depends on code only by a relatively small percentage. Maintenance and support need people: calls, brainstorming, decisions, strategic planning, cost efficiencies. AI is not there yet, and might never be, at least not in a cost-effective way.

“AI can spot security weaknesses and suggest fixes.”

Yes, it actually can - but suggestions are one thing, and implementation of those suggestions can vary greatly in cost, time, and scope. You need a human in the loop to evaluate what can be done now, what later, and what maybe not at all. AI can point you in the right direction - you still need to follow that direction yourself.

“AI can do my front end, so I don’t need designers or front-end developers.”

This is good if you’re building a prototype or a brochure website. But leave AI alone with your front end on a more complex project and you’ll soon face the ugly truth: what you’ll get will be “nearly there”, but getting it “entirely there” will cost you much more than what you’ll save. And “nearly there” is not good enough for enterprises.

I’m not saying AI is useless in this scenario, but it must be carefully used and controlled in order to yield acceptable results. And this needs humans - and money.

The bottom line

In a complex CMS implementation, you deal with a, well, complex system worth a lot of money and costing even more in the long term. You need assurance that it will continue to function, remain secure, and stay easy to extend when needed.

There are no middle-ground solutions here. You need people who know what they’re doing, are able to make decisions, and can even challenge you. This is the healthy approach to keeping the engine running. Everything else is just a tool in your arsenal - this includes cost cuts, AI, and even platform changes. Using each tool must be justified and proven to benefit more than it costs.

So no, not everything is a “website”. Some websites are more… websites than others. And this justifies the difference between a €2k once-off budget and a €60k recurring one.

Considering Umbraco for a complex website?

At DotSee, we have developed reusable libraries and internal tooling that cover many of the recurring needs of large websites. Redirects, navigation structures, sitemap visibility, editorial safeguards, and more are already solved problems for us.

This allows us to focus on the things that actually matter for each organization: content structure, integrations, security, and long-term maintainability.

If you are building or redesigning a complex website and considering Umbraco, we would be happy to talk with you.

Let's discuss your project