MSPs: Your Best KB Articles are Hiding in Closed Tickets

A client can't get on the Wi-Fi after a router swap. Your tech walks them through it. Twenty minutes. Ticket closed.

Two weeks later, a different client hits the exact same thing. Different tech. Same walkthrough. Same twenty minutes.

By now your team has written that answer six times. In six tickets. Where nobody will ever see it again.

The answer already exists. It's just in the wrong place.

Here's the thing. Most MSPs don't have a documentation problem. They have a location problem.

The fix for nearly every recurring issue is already sitting in your PSA. It's in the resolution notes, the back and forth with the client, the note somebody left at 4:45 on a Friday. It's real, it's specific to your clients, and it's yours.

It's also buried. Nobody searches closed tickets for fun. So the answer stays put until the next tech happens to remember it. Or doesn't.

Why nobody writes the article

It's not laziness. It's timing.

The moment one ticket closes, the next one opens. Writing the article becomes the thing you'll do when things calm down. And we both know how that goes.

Then there's the blank page. Turning a messy ticket thread into a clean article takes real effort, even when you know exactly what the fix was.

And even when someone carves out an afternoon, which article do they write first? Most teams guess.

So the KB stays thin. Clients can't find answers, so they open tickets. Your techs answer the same questions again. Round and round.

Start by finding out what's actually missing

This is where CloudRadial ServiceAI earns its keep. It sits on top of your PSA and your documentation and works from your own ticket history.

Before it writes anything, it can tell you what to write.

ServiceAI's regression test takes your most recent tickets (the latest 20, by default), feeds each one in as if it just arrived, and checks whether your current knowledge base could actually answer it.

When it can't, and the reason is that the article simply doesn't exist, ServiceAI writes a Documentation recommendation into its Recommendations tab. Each one names the topic, explains why the answer fell short, and outlines what the article should cover. The weakest results go to the top of the pile.

So instead of guessing which articles matter, you get a to-do list built from real tickets. Run it weekly and watch the list get shorter.

Let your tickets do the first draft

Now the writing part. ServiceAI gives you a few ways to skip the blank page.

• Straight from your tickets. When ServiceAI's chat recommends new KB topics, each one gets its own Generate button. Every article is written from that topic's own tickets and articles, so it reflects how your team actually fixed the problem.
• From a prompt. Paste in a recommendation, or describe what you need in plain English, add a source URL if you've got one, and Generate Article returns a full draft with a title and body.
• From the Idea Library. Starting from zero? There are ready-made starting points for common Microsoft 365 and IT topics.

Whichever way you start, you preview and edit before anything is saved. Then you either add it to your connected knowledge base as a draft, or save it as an HTML file to upload wherever you keep your docs.

That changes the job. You're not writing anymore. You're editing. Ask any tech which one they'd rather do.

Fix the articles you already have

Your KB probably isn't empty. It's just uneven.

ServiceAI scores every article twice, on a scale of 1 to 10. One score is for grammar: how well it reads for a person. The other is for readability: how easily ServiceAI can pull an answer out of it. Each article also comes with specific suggestions for making it better.

Find a weak one and hit Rewrite. You'll see the original and the cleaned-up version side by side, and you decide whether to update it in your KB, download it, or walk away.

There's also a Similar Articles view that shows how much two articles overlap. A match of 90% or more usually means you've got duplicates, and duplicates are how a client ends up with two different answers to the same question.

You still review every one

A draft is a draft. Somebody on your team reads it, fixes what's off, and decides whether it's ready.

That step matters more than it sounds. A ticket thread can hold a workaround that was right for one client and wrong for everyone else. It can hold a detail you'd never want published. The AI saves the typing. You keep the judgment.

Treat every draft like notes from a sharp new hire who's never seen your stack. It's a solid start, and it still needs your team's specifics before it ships.

Your data stays yours, too. ServiceAI works from your own tickets and documentation, and each MSP's data is kept separate. Nothing gets pooled across partners.

Put the answers where clients look

An article nobody can find isn't much better than a closed ticket.

That's where the CloudRadial Unified Client Portal comes in. It's the branded portal your clients already use for requests and tickets, and it has a knowledge base built in. So the answer lives right where the client goes when something breaks.

The client with the Wi-Fi problem has somewhere to look before they open a ticket. And when they open one anyway (they will), your tech can send a link instead of retyping the walkthrough for the seventh time.

If you want to see how the portal handles the ticket side, we covered that in Where's my ticket? and in request forms that produce complete tickets.

The same articles make ServiceAI sharper, too. It grounds its answers in your documentation first, so a better KB means better help for the techs using it. One body of knowledge, doing several jobs.

A few honest notes

We'd rather you hear these from us now than find out in month two.

It's only as good as your tickets. If your resolution notes say "fixed" and nothing else, there's not much to draft from. ServiceAI does best with months of recent history behind it, so when you set it up, plan to sync six to eighteen months of tickets, not just the last few days.

It doesn't replace a person. Every draft needs a human read before a client sees it.

Someone has to own the loop. Run the test, review the recommendations, edit the drafts. It's a little time each week. Put a name on it.

A KB won't make anyone read it. It gives clients a better place to look. It doesn't make them look. Some clients will email you forever, and that's fine.

Which tier you need

The knowledge base is part of UCP, starting on UCP Starter.

ServiceAI is sold separately, in Professional and Enterprise tiers, and the article tools are part of the platform. Check pricing for current plans.

Bring your three most-repeated tickets

The easiest way to judge this is with your own history. Book a demo and bring the three issues your techs solve most often. We'll show you what ServiceAI says is missing from your KB, what a draft built from your own tickets looks like, and where your clients would find it.

Frequently Asked Questions

How can an MSP build a knowledge base without writing every article by hand?

Start from the work you've already done. Resolved tickets hold the real fixes, so drafting articles from them turns writing into editing. In CloudRadial, ServiceAI can recommend KB topics and draft each one from that topic's own tickets and articles for your team to review.

How do I know which KB articles to write first?

Test your knowledge base against real tickets and see where it falls short. ServiceAI's regression test replays your most recent tickets, checks whether your KB can answer them, and writes Documentation recommendations for the gaps, with the weakest answers first.

Should AI-drafted KB articles be reviewed before they're published?

Yes. A human should read every draft before a client sees it, because ticket threads can include client-specific details or workarounds that don't belong in a general article. In ServiceAI you preview and edit every draft, and articles added to a connected knowledge base go in as drafts.

Where do AI-drafted articles end up?

In your knowledge base. ServiceAI can add reviewed articles to your PSA's knowledge base, Hudu, or CloudRadial, or save them as HTML files you can upload anywhere. IT Glue is a source ServiceAI reads from, not one it writes to.

Can AI improve the KB articles I already have?

Yes. ServiceAI scores each article for grammar and for how easily AI can pull answers from it, suggests improvements, and offers a Rewrite with a side-by-side comparison before anything changes. It also flags articles that overlap heavily.

Does ServiceAI share my data with other MSPs?

No. ServiceAI works from your own ticket history and documentation, and each partner's data is kept separate. There's no shared corpus across partners.

How much ticket history does ServiceAI need?

More recent history gives better results. Six to eighteen months is a common starting range, and the quality of your notes matters as much as the quantity.

Which CloudRadial tier includes the knowledge base?

The knowledge base is included starting with UCP Starter. ServiceAI is sold separately in Professional and Enterprise tiers. Check pricing for current plans.