Meet Scout

This started as a navigation problem

I’ve spent years helping teams build their identities and stand up static sites to support their projects and brands. Just to be clear, a static site is a website that doesn’t require a backend database, user accounts, and too much programming to create a good experience. Think of them like digital storefronts. They’re fast, reliable, and cheap to host. They work great for portfolios and product/service marketing sites.

I’ve noticed a trend towards teams, especially those with lots of technical documentation, to host more content on their static sites. Teams will have user guides, blog posts, and case studies they want to feature on their sites, which is usually where these static sites break. The sites technically can handle the content, but it usually means adding another link, section or page.

And that’s where static sites usually break, when your visitors need to find something.

Static sites will give users a nav menu and wish them luck. I felt bad for my customers who had good things to share, but their website were more like dilapidated warehouses masquerading as showrooms. Even worse was when the content would just sit on some random page, be forgotten, grow stale, and then never really get to people. I write this because this is the same problem I had with my site.

Without having more advanced features like a database to index content, I was limited in how I could direct traffic to certain content. And at work my clients were facing the same challenge. They had information to share and needed help getting people to it.

That’s the problem I set it to fix with Scout. Yes, it is technically a chatbot, but I care equally about it’s ability to find existing information as I do it’s generative abilities. And yes, you can chat with it, but really I want you to put it to work by asking it questions about my experience, expertise and resources and have it serve as an actual stand-in when you can’t reach me directly.

In the age of AI (or more specifically semantic search) our preferences for finding content are moving more towards natural language and away from clicking through navigation menus, accordions, and drop-downs.

The rest of this case study will walk you through some of the things I’ve tried, the lessons I’ve learned, and maybe save you a couple weeks of back and forth with a chatbot if you are considering also making the jump from a static site builder like Squarespace/Wix to something a little more interactive.

Making sense of the infinite void

Imagine all of the information out there can be represented as stars in the sky. The value we can bring to others is to identify the constellations (patterns), celestial bodies (concepts), and other transient phenomena (current events, insights, trends) in the observable universe. Otherwise, people can easily get lost in the vast infinite void of information.

Okay, I can appreciate the space analogy might not have landed well. Just in case, this is the expression I’m suggesting:

$$ \frac{\text{Information} \times \text{Patterns} + \text{Key Concepts}}{\text{Context}} = \text{Knowledge} $$

This expression just makes sense to me as someone trying to get knowledge to adult “learners” at a time when, as my former colleague Bryan Woebkenberg would say, we are simultaneously thirsting for and drowning in information.

Great teachers today (whether actual teachers or website stand-ins) help learners make sense of the void. They function more as tour guides who help learners notice the patterns and key concepts relevant to their learning objectives. Then, these teachers provide context to help this information “stick.”

When I say context this could be case studies, anecdotes, or even simulations and exercises. The goal is to not simply give people an image of the universe. When it comes to digital information, we all have telescopes capable of observing more than we ever could want to observe. Great teachers point out “things” on the map worth observing, given our unique objectives. So I designed a chatbot experience that lets the learner determine the denominator…their context. The idea is to just let the learner tell me what it is they want to know, and Scout will do their best to point out what I think might help using the information and patterns I’ve provided.

An aside on graph theory

Back in 2018, a brilliant teammate of mine told me about this relatively new technology called “Neo4J” and started on about knowledge graphs. I wish I could say I jumped right in to the technology, but I didn’t. I listened, went to a couple sites and thought, “this is cool, but I have no idea what I can do with this yet.” He’s since become one of my company’s leading minds on Artificial Intelligence and I (kinda sorta?) think I know what he was trying to tell me all those years ago.

Graph database technology like Neo4J help application developers visualize knowledge graphs. And popular makers of graph database technology are still trying answer the question,what is a knowledge graph?

The technology still hasn’t caught on like wildfire, but is expected to hit mainstream by 2030, with the knowledge graph market set to more than triple in size by then.

The Knowledge Graph Market is expanding rapidly as organizations seek advanced data integration, semantic intelligence, and contextual analytics capabilities. Knowledge graphs enable structured representation of relationships between data entities, improving data discovery, reasoning, and decision-making. Enterprises adopt knowledge graph solutions to unify disparate data sources, enhance analytics accuracy, and support artificial intelligence initiatives.

My favorite online discussion about graph theory comes from a guide Sam Livingston-Gray https://think-like-a-git.net/sections/graph-theory.html

So I rebuilt the site

This is the part where people assume it got complicated.

It did—but not in the way you think.

I moved from Squarespace to a static site setup using Hugo.

Which sounds like a big leap.

And it is.

But the real change wasn’t the tool.

It was what I started doing with the content.

The tradeoff (let’s not pretend)

Here’s what you actually sign up for when you leave a site builder:

You lose:

• drag-and-drop editing
• built-in hosting + CMS
• one-click everything

You gain:

• full control
• near-zero hosting cost
• performance (it’s just files)
• the ability to build things your platform would never allow

That last one is the reason I stayed.

The unexpected shift: content became the system

Once I had control, I stopped thinking in pages.

And started thinking in chunks.

Every page on my site is now broken into knowledge chunks.

Not automatically.

On purpose.

Each chunk has:

• a title
• a topic
• tags

Why this matters (even if you never build a chatbot)

Most websites are designed like this:

page → scroll → maybe click → repeat

That works fine when someone is browsing.

It breaks when someone is searching with intent.

Chunking content changes that.

Now your site isn’t just pages.

It’s a collection of retrievable ideas.

This is where the chatbot actually becomes useful

Instead of guessing, the chatbot can:

• search those chunks
• understand what they’re about
• return something grounded in the actual site

Or…

It can decide not to answer at all. Not every question needs AI.

And just route you to the right place.

This was one of the bigger realizations.

If someone says:

• “Show me your projects”

You don’t need a model.

You need a link.

So the system I built checks that first.

Only if the question requires interpretation does it:

• search content
• assemble context
• generate an answer

Everything else is handled directly.

What this feels like as a user

This is the part that matters.

Not the stack.

The experience.

Instead of:

• hunting through menus
• scanning pages
• guessing where something lives

You can:

• ask a question
• get a grounded answer
• jump directly to the source

It turns the site into something closer to a guide.

What it costs (realistically)

This is where most people get nervous.

So here’s the honest version.

Money

• Hosting: basically free (Cloudflare Pages, etc.)
• Domain: same as before
• API usage: minimal if you design it right

Time

This is the real cost.

You will spend time:

• learning the system
• structuring your content
• maintaining your setup

More than Squarespace.

So… is it worth it?

Depends on what you want your site to do.

If your site is:

• a simple presence
• a digital business card
• mostly static

Stay where you are.

Seriously.

But if your site is:

• content-heavy
• meant to teach, explain, or guide
• something people come to with questions

Then this starts to make a lot more sense.

Where this goes next

I’m still figuring this out.

There’s a version of this that could be really powerful for:

• learning systems
• technical documentation
• performance support tools

Anywhere people are trying to find the right answer quickly.

That probably deserves its own write-up.

Closing

You don’t need to rebuild your site.

Most people shouldn’t.

But if you’ve ever felt like:

“This works… but it doesn’t really help people”

Then it might be time to think beyond pages.

And start thinking in constellations.