I Ran an MSP. Here's the Growth Problem Nobody Warned Me About.

Why I left, what I learned about AI and revenue, and why the fear conversation is pointing at the wrong thing entirely.

I Knew the Revenue Was There. I Just Couldn't Get to It.

Let me describe a feeling I think you'll recognize.

You're sitting across from a client you've had for three years. Good relationship. Solid contract. And somewhere in the back of your mind you know there's more work you could be doing for them. Projects worth scoping. Gaps in their environment you've noticed but haven't had time to formalize into a proposal. Budget they're about to spend with someone else.

But your PSA data is one thing, your RMM is another, and your vCIO is still two hours into pulling together the QBR deck. By the time you have the full picture, the moment has passed.

I lived in that moment for years. And I want to be clear: it wasn't a sales problem. It was a signal problem. I didn't have bad people or bad clients. I had disconnected data, and it was quietly costing me more than I knew.

Why I Left the MSP Business

I didn't wake up one day and decide to build a software company. I spent years running an MSP, doing the work, living the operational chaos that comes with scaling a managed services business. Server rooms, client escalations, vendor negotiations, team management. The whole thing.

And I was good at it. But there was a ceiling I kept bumping into, and it wasn't about effort. I was working harder than I probably should have been. The problem was that I kept making decisions with incomplete information, and the cost of that incompleteness was invisible in the day-to-day but very real at the end of the year.

The insight I kept circling back to: the revenue I was looking for wasn't out there with new clients I hadn't met yet. It was sitting inside my existing relationships, in data I already had, waiting for a system smart enough to surface it.

What Running an MSP Actually Taught Me

The revenue leak is invisible until you go looking

I didn't have a good sense of how much project revenue I was leaving on the table until I started measuring it. Industry benchmarks from the Service Leadership Index are clear on this: healthy MSPs should be generating 20 to 50 percent of their annual MRR as project revenue from existing clients. Most of the operators I've talked with are sitting closer to the floor of that range than the ceiling.

That gap doesn't mean your sales team is failing. It usually means your systems aren't talking to each other, so the signal that would prompt the right conversation never arrives in time. I saw this in my own business before I had language for it.

QBR prep was killing us, quietly

I'm going to say something that might sting a little: your QBR process is probably costing you more than you think, and not because of the prep time.

Yes, a traditional QBR requires roughly 5.5 hours of internal time across your vCIO, your engineers, and your sales team. At fully loaded cost, that's around $750 a meeting. I know that number, because I optimized against it. I made decisions about QBR frequency based on that prep cost.

What I wasn't measuring was the other side of the equation. A well-run QBR typically surfaces one to two meaningful projects per year at $15,000 to $30,000 per project. Skipping or shortcutting those meetings wasn't saving me $750. It was costing me $20,000 per client cycle in revenue I never saw.

Tribal knowledge is a liability

I had people on my team who knew everything. Every client's quirks, every system's history, every past incident that explained why things were configured the way they were. That knowledge was invaluable right up until the moment it wasn't accessible.

When critical context lives in someone's head instead of a connected system, every gap in their availability is a gap in your service delivery. Scaling while that's true is more fragile than it looks from the outside.

The Question That Changed How I Thought About the Business

There's a version of MSP growth that looks like success from the outside. More clients, more headcount, more tools. But if your margins aren't expanding, if the operational overhead keeps pace with every new dollar of revenue, you're not really scaling. You're running faster on the same treadmill.

The question I couldn't stop asking was not how to work harder. It was: what would it be worth if I actually had the right information at the right time?

That question, and years of experience working alongside a leadership team with 35+ years in data, managed services, and professional services, led to what became SKAIA. Not as a product I wanted to sell, but as the solution I genuinely needed when I was still operating.

I built it because it didn't exist. That's the whole story.

What I Mean When I Talk About AI for MSPs

I want to be careful here, because the word AI has been attached to a lot of things that don't deserve it. Automated reports aren't AI. Better dashboards aren't AI. I've seen MSP vendors use the term to describe features that were already table stakes five years ago.

What SKAIA does is different in a specific and practical way. It connects the data you already have across your PSA, your RMM, your knowledge base, your SharePoint, and your financials, and it surfaces the insights that currently require hours of manual work to find. It makes the signal visible before it expires.

In practice, that means:

• QBR prep that used to take my team 5.5 hours can happen in closer to 15 minutes, with richer context than we had before

• Revenue opportunities inside existing client data get surfaced before the moment passes

• Project capture rates can move from 25% toward 40% and beyond, without adding headcount

• The knowledge that used to live in one person's head becomes accessible to the whole team

This isn't AI as a novelty. It's AI as operational infrastructure. It works inside the tools you already use and it doesn't ask your team to learn a new way of doing their jobs. It just gives them better information when they're doing the jobs they already do.

The Fear Conversation Is Pointing at the Wrong Thing

When I talk to MSP owners about AI, I hear a version of the same hesitation pretty often. Concern about client data. Worry about what it means for the team. Skepticism about whether the ROI is real.

Those are fair questions. I'd ask them too.

But here's where I think the conversation goes sideways: we treat the risk of adoption as visible and the risk of not adopting as hypothetical. The truth is the opposite. The cost of not changing is just harder to see on a spreadsheet.

Every month you run a manual QBR process is a month you're leaving the same money on the table. Every month your client knowledge lives in tribal systems instead of connected ones is a month your operation is more fragile than it needs to be. That cost is real, even when it's invisible.

I'm not saying that to be dramatic. I'm saying it because I made that trade for too long before I did something about it.

Questions Worth Sitting With

If you're running an MSP and this is landing, here are the same questions I eventually had to answer for myself:

• Do you know exactly how much project revenue you're generating per dollar of MRR? If not, you're likely closer to the floor of that 20-50% benchmark than the ceiling.

• How long does your QBR prep actually take? Not what it's supposed to take. What it takes. Multiply that by your client count and your quarterly cycle.

• Where does your critical client knowledge live right now? If the honest answer is 'mostly in a few people's heads,' you have a scale risk that hasn't surfaced yet.

• What is one missed project per client actually worth to your business? Run it across your top 20 accounts. That's the number AI for MSPs is really about.

I spent too long not asking these questions clearly enough. When I finally ran the numbers, the case for change wasn't difficult. The difficulty had been making the numbers visible in the first place.

What I'd Tell Myself If I Were Still Operating

The MSPs that will define their market segment five years from now won't necessarily be the biggest or the most funded. They'll be the ones who stopped accepting that disconnected data and manual processes were just the cost of doing business.

I built SKAIA because I couldn't find it when I needed it. If what I've described sounds like your operation, I'd genuinely like to talk. Not to sell you on a product, but to have the conversation I wish I'd had earlier, with someone who has sat in the same seat.

Thirty minutes. Honest look at your numbers. That's where it starts.

Let's continue the conversation:

Connect with me on LinkedIn: linkedin.com/in/denniskao1

Book a 30-minute call: correlatio.io/#book-a-demo