A Front Room, Not a Feed: Keeping ChrisJM Human When the Bots Come Calling

My kettle has a small instruction sticker on the base. Fill to the line. Don't switch on empty. I have read it roughly once, and that was because I was bored waiting for the water to boil. No search engine has ever read it. No AI summariser has ever paused to consider the line. And yet the kettle works, because it was designed for the person standing in front of it.

That is roughly how I think about ChrisJM in 2026. The site is not a datasheet for crawlers. It is a front room for people who have wandered in, probably via a link from a mate or a search result they half remember. Search engines, AI summaries and scrapers will turn up. They always do. But they are not the guests I set the table for.

The kettle test

When I am deciding whether a page on ChrisJM makes sense, I use what I have started calling the kettle test. Would this heading still work if someone read it out loud while waiting for a brew? If the answer is no, it probably needs another go.

This is not an argument against clarity. Search engines like clarity too. But clarity is not the same thing as coldness. A heading like Selected Work is clear enough, but it sounds like a portfolio template. A heading like Things I've Made tells a human what they are about to see and gives the page a bit of pulse. The bot can still index it. The person can still understand it. Everyone leaves with what they came for.

I have been trying to write headings that answer simple questions. What is this? Who is it for? How do I get in touch? How old is this thing? That last one matters more than it used to. The web is full of pages that look current but were last touched years ago. If a project note says 2026, it should actually mean 2026.

Contact routes should not be a maze

The contact link on ChrisJM sits in the main navigation. It is not buried behind an icon shaped like a paper plane. It is not hidden at the bottom of a page under three layers of social buttons. If someone has taken the trouble to find the site and wants to say hello, the least I can do is leave the door on the latch.

The same goes for the contact page itself. I do not want a form that asks for a budget before it asks for a name. I do not want a chatbot pretending to be me. I want a visible email address, a short note about what I am open to, and a rough idea of when I reply. That is it. Bots can scrape the email if they like. I would rather deal with a bit of spam than make a real person hunt for a way in.

Project notes need context, not just files

A project page with a screenshot and a link is fine if you already know what you are looking at. Most people do not. So I have been adding short notes to project pages: what it was, why I made it, what I learned, and what is still a bit wobbly. That last part feels important. The web is full of finished-looking things that were actually held together with tape and a deadline.

This is where human-first design and machine-readable design can actually agree. A crawler can read the text. An AI summary can pull out the key points. But a person gets the bit that matters: the context. A screenshot with no caption is just wallpaper. A project with no notes is just a file in a nice font.

What the bots can have

I am not trying to lock the doors. ChrisJM has clean HTML, sensible page titles, a sitemap and metadata that does what it is supposed to do. If a search engine wants to understand the site, it can. If an AI tool wants to summarise a post into three bullet points, it can have a go. I would rather be useful in a summary than invisible.

But there is a line. The personality stays in the original. The half-finished thoughts, the daft asides, the bit where I admit I changed my mind halfway through a project — that is for the people who actually turned up. A summary can tell you what a post is about. It cannot tell you what it felt like to write it at 11pm with a cold cup of tea going stale beside the keyboard.

Writing for machines makes a site feel empty

I have seen sites that feel like they are talking to a search engine rather than a person. Every sentence is optimised. Every heading is a keyword. Every page is a landing page. They are technically fine and completely exhausting. You can almost hear the spreadsheet behind the curtain.

ChrisJM is not that. As I write this in November 2026, the site is still a personal project. It is a place for notes, projects and opinions from one person in the UK. That means it can be a bit uneven. It can have a page about a bench and a page about a footer. It can have a contact page that reads like a doormat rather than a legal disclaimer. That is the point.

The human defence

Human-first design is not a soft option. It is the best defence against a web that feels like it is talking to itself. Every time I choose a plain word over a clever one, a visible link over a hidden menu, or a bit of context over a naked file, I am betting on people. That bet still seems worth making.

So the bots can come in. They can read the headings, index the projects and summarise the posts. But the kettle instructions are for the humans. And on ChrisJM, so is the rest of it.