Pull up the last ten new clients you signed. The most recent ten, not the biggest ten. Next to each name, write down how the first conversation started.
Use one rule. A client counts as non-referral only if they arrived without a referral, an intro or an existing relationship behind them. The accountant who sent you a dental practice is a referral. So is the old colleague who changed jobs and called you. What is left are the clients who found you, or whom you found, with nobody vouching for you first.
Count them. With your own book in front of you it takes about five minutes, and the number you get, out of ten, tells you more about where your managed service provider (MSP) goes next than your marketing plan does.
Most owners I meet land on zero or one. The usual diagnosis is a marketing problem. I read it as a demand-source problem. Demand source means where a new client's first contact came from. A channel is the tool you used to reach them. You can run five channels and still have one demand source, if every signed contract traces back to somebody's recommendation.
For ten years my job was pricing businesses from the buyer's side, and today I am fractional Chief Revenue Officer (CRO) at a national MSP in the United States. In both seats, the first thing I want to know about a business is where its next client comes from.
If you want the case against hiring an outbound agency as the quick fix first, read why MSP outbound agencies disappoint. If you want to see how a buyer scores the same weakness, read the MSP moat framework.
Why the best leads cap your growth
Referrals are the best leads you will ever get. They arrive already trusting you. That is exactly why nothing else gets built. When the best leads show up on their own, there is never a quarter where building a second source feels urgent.
The cost shows up later. A referral-fed MSP grows at the rate its existing clients happen to talk to people who need information technology (IT) help this year. You do not control that rate, you cannot turn it up in a slow quarter, and all of your demand is borrowed from other people's conversations. I call this the referral ceiling: the growth limit set by how many people your clients mention you to in a given year. Good service raises the odds they mention you. It does not add conversations they were never going to have.
You know you are under the ceiling when the symptoms sound familiar. The pipeline is the golf calendar. Nobody can forecast next quarter. You cannot take a month off, because the relationships that bring in work are yours. One referral partner retiring is an existential event. And you cannot hire ahead of demand, because you do not know when demand is coming.
None of this means you should stop asking for referrals. Referral programs, picking a niche, showing up at the local chamber of commerce: that is the standard advice, and all of it is fine at every stage. Keep doing it. Just do not expect it to move your count much, because most of it still depends on someone else deciding to mention you.
The four stages, by the count
Here is how I stage an MSP using that one number. Every count maps to exactly one stage.
| Stage | Non-referral clients in your last 10 | What it feels like | What you can forecast |
|---|---|---|---|
| 1. Referral-Dependent | 0 or 1 | The pipeline is the golf calendar | Almost nothing |
| 2. Accidental | 2 or 3 | Something is working and nobody can say what | Not reliably, because you cannot repeat it on purpose |
| 3. Predictable | 4 to 6 | Spend goes in, pipeline comes out | Next quarter's pipeline, and what a new client costs you |
| 4. Compounding | 7 or more | Winning clients gets cheaper every quarter | Pipeline, cost per client, and which way that cost is moving |
Stage one, Referral-Dependent, is the zero or one from the opening. The business can be profitable and well loved by its clients. It just cannot tell you where next year's clients come from, and the owner is usually the one who brings them in.
Stage two, Accidental, is the confusing one. Two or three clients came in some other way: a web form, a post someone shared, a listing a buyer stumbled on. Something works. Nobody can say what, so nobody can do it again on purpose.
Stage three, Predictable, is the first stage where growth becomes a decision. You know your customer acquisition cost (CAC), which is what you spend on sales and marketing to win one new client. Money goes in, pipeline comes out, and you can defend a forecast to your banker or your spouse.
Stage four, Compounding, is where what you built keeps working long after you paid for it. Content, search rankings and your reputation in the places buyers look all appreciate, so your CAC falls quarter after quarter.
None of the growth frameworks I have seen stage MSPs by where demand comes from. They stage by company size, or by what a marketing firm does for you. I prefer demand source because it is the lens that predicts whether you can plan.
If you are happy at your size and have no plan to sell, stage one can be a good business for a long time. The ladder matters when you want growth you can plan, or when someone else will one day put a price on what you built.
What moves an MSP up a stage
Two engines and a floor.
Engine one makes you the default before they shop. When a business owner starts wondering whether their IT provider is good enough, they ask around, search online, or ask an artificial intelligence (AI) assistant. Engine one is whatever makes your name come up in that moment: search engine optimization (SEO), visibility in AI search, a place in the lists buyers read, and a founder whose name people recognize. It is slow, and it compounds.
Engine two makes you the answer when the trigger fires. The server dies. The cyber insurance renewal asks questions nobody in the office can answer. Engine two is paid search and outbound, aimed at the moment someone needs an MSP now. Both engines run at every stage. What changes from stage to stage is the mix.
Then the floor, which is where I see most of the money lost. The sales floor is everything between a lead arriving and a contract signed: speed to lead measured in minutes, qualification, a discovery call, a second call, nurture for the prospects who are not ready yet, a proposal chase, and a customer relationship management (CRM) system that someone actually keeps current. In the $2M to $10M MSPs I meet, sales is usually the founder, fitting follow-up between everything else. More leads poured onto that floor do not produce more contracts. They produce more leads that went quiet.
This is also where the fired-agency stories come from. Closing the lead is the MSP's job. The trouble is that most MSPs I meet do not know how to sell well, because nobody taught them and the founder learned on the phone between tickets. That is why my agency's system runs from lead to revenue rather than stopping at the form fill. In my experience the best results come when the agency also coaches the MSP's sales team, or puts its own sales team in, because the floor is where the leads were dying.
At an MSP I work with, won deals averaged 2.2 conversations. Lost deals averaged 1.3. Deals that reached a second conversation closed at a far higher rate than deals that stopped at the first. The same pipeline showed something less comfortable: a third of inbound callers had no clear reason to call, and none of those closed. That is one MSP, not an industry benchmark, and I would not treat the decimals as law. The shape is the point. At that MSP, deals were won or lost after the lead arrived: in whether anyone got the prospect to a second conversation, and whether the calls coming in were ever qualified.
one MSP, not an industry benchmark
The last piece is the scoreboard. Attribution means following each click or call all the way to the signed contract, so you can say which source produced which revenue. Without it you cannot compute your own stage, except from memory, which is what you just did. Memory is good enough for today's count. It will not tell you next quarter whether anything changed. So measurement comes first, before any new spend.
The first move at each stage
One move for your stage, not four. Each one assumes the stage below it is already in place.
Stage one: measure, then test your own floor. Add a lead-source field to every new opportunity in your CRM and fill it in every time, starting today. Then run the Friday 4pm test. At 4pm on a Friday, submit the contact form on your own website from a personal email address, and time how long it takes to hear back. I pick Friday afternoon because that is when a fast reply is hardest to keep. The standard to aim for is that someone whose job it is calls within minutes. If your first non-referral leads land on a floor where nobody answers until Monday, building an engine only buys you more of them.
Stage two: find the accident and run it on purpose. Take the two or three clients who did not come from referrals and trace each one back to its first contact. If two came from the same place, that is your channel. Give it a named owner and a modest budget, measure it, and do not start a second channel yet. My rule is to start with one channel, get results from it, and put the money those results produce into the next one.
Stage three: add the engine that makes you the default. If you reached stage three through paid search or outbound, you can buy pipeline when a trigger fires. What you do not have yet is the slower asset that puts your name in front of buyers before they shop. Fund engine one from what engine two earns, one channel at a time, the same way you built the first. If you got to stage three on search and content instead, the missing piece is engine two, added the same way.
Stage four: protect the floor. At stage four the risk moves from finding demand to handling it. More leads mean more chances for a slow reply, a skipped second call or a proposal nobody chases. Watch the scoreboard for those first: time to first reply, and how many leads reach a second conversation.
What a buyer reads in your count
When a buyer's team does diligence, meaning the weeks they spend checking everything you told them, one of the questions they ask is who brings in new clients. If the honest answer is the founder's network, they see growth that walks out the door with you. Read from the buyer's side, referral-dependent revenue trades at a discount when you sell, and a predictable pipeline is part of the story that earns a higher multiple (the number your profit is multiplied by to set the price). Buyers weigh your growth trajectory separately from the size of your profits, which I cover in what moves an MSP's multiple. Your count out of ten is a fast read on both questions a buyer cares about here: how much of your growth depends on you, and whether anyone could plan it without you.
What to do this month
Do the count this week, with the rule: no referral, no intro, no existing relationship. Write the number and your stage on the first page of whatever you plan the business in. Then do the one move for your stage and leave the other three alone. Next time you count, count from the lead-source field instead of memory, and see whether the number moved.
If you want to work through your number with other owners, Know Your Number is a private room for MSP owners with a regular call.