The Pageloop Blog

The best Mintlify alternatives for teams that have outgrown it.

Eight IT documentation software tools for storing what your team knows, and the one job most of them leave to you.

The full run of support and docs conferences from now through December

Text ages (kinda) gracefully. Screenshots don't.

Five platforms, one shared blind spot.

What knowledge base automation means in practice, why most tools only cover half the problem, and what to look for if you're evaluating one.

Writing the articles was never the hard part. Keeping them from contradicting each other is.

These aren't the articles you know are broken. They're the ones that look fine.

The help center that worked fine for humans breaks in specific, predictable ways when a chatbot reads it.

Migration is rarely the answer. Here is what to add on top.

Bad documentation was always costing you. AI just made the bill visible.

Zendesk Guide is a good start. Here is what to do when it is not enough.

And that's costing you customers.

Seven knowledge base platforms for SaaS teams, split by the jobs that matters most

Launching takes a week. Keeping it accurate takes a system.

Most documentation was written for one audience. It now has two.

A knowledge base is a garden, not a monument. Here's how to keep it alive after launch.

What we've learned about getting the foundation right.

It's not the bot. It's what the bot is reading.

A help center's accuracy is decided on the org chart, long before anyone writes a sentence.

Your help center is always behind your product because the people building the product aren't the ones documenting it.

A practitioner's guide to finding and fixing what drags Fin down.

From blank page to published article, in under five minutes.

Because your docs won't maintain themselves.

(and what they still can't do)

Two years ago, nobody wanted AI in support. Now everyone wants it yesterday.

The questions to document first, why deflection decays over time, and how to tell whether self-service is actually cutting your ticket volume.

Because "AI knowledge base" means three different products.

It happens to every help center. Here's how to slow it down.

Where GitBook holds up, where teams outgrow it, and what to switch to.

When your deflection metrics look healthy but your ticket volume hasn't changed, the problem is in the measurement itself.

The content problems that make AI agents give wrong answers, and how to find them before your customers do.

A more capable model writes a more convincing wrong answer.

The articles that look fine are usually the ones that aren't.

How a decade in support and years in AI led to Pageloop.

And learn where it still falls short.

Because Fin is only as good as what it reads.

Not all of them are worth your time