Notes on building the AI-native web.
Page 8 of 17
Where AI actually belongs in your product
Bolting a chatbot onto a homepage isn’t AI strategy. The wins come from wiring models into the workflows that already cost you money.
Putting an AI employee inside your CRM
The CRM is where the context lives and where the mess lives. Both facts shape what you should let it do there.
Writing about competitors without lying
Comparison queries are among the most commercial prompts there are. A page that only flatters you won’t be cited for them.
Building on someone else’s model
Your feature depends on a system you don’t control, that changes without asking. Four ways to hold that dependency safely.
Writing documentation developers will actually read
Nobody reads the wiki. They read the four things they need at the moment something is broken or new.
How to review an AI employee’s work
Reviewing everything doesn’t scale and reviewing nothing is how it decays. Sampling well is the skill in between.
The anatomy of a site that converts in 2026
Speed, clarity, and a point of view. A teardown of what actually moves the needle on a modern marketing site.
Permissions: what your AI feature is allowed to see
Retrieval ignores your access rules unless you build them into it. The question to ask in week one, not week twenty.
How long does AIEO take to work?
Parts of it move in weeks and parts take quarters. Knowing which is which stops the programme being cancelled at month two.
Dates, times and the bugs they cause
The most under-estimated source of quiet data corruption in business software. A short list of rules that prevent most of it.
The page that used to rank and doesn’t any more
Before rewriting it, work out which of four things happened. Three of them have completely different fixes.
How to explain an AI project to a board
They are not asking how it works. They are asking what it costs, what it risks, and how you will know if it failed.