Referrals are a growth channel, not a lucky accident
Most MSPs win their best work through referral and treat it as weather. It is the only channel where you arrive already trusted — and almost nobody runs it deliberately.
Ask an MSP owner where their best customers came from and the answer is usually some version of “word of mouth”. Ask what they do to generate more of it and the answer is usually nothing, or something that stopped two years ago.
There is a reason for that. Referrals feel like a consequence of doing good work rather than an activity, and asking for them feels like admitting you need the help. So the channel that produces the highest-quality, fastest-closing, least price-sensitive business in the entire company is the one left entirely to chance.
Good delivery makes referrals possible. It does not make them happen.
The same deal, arriving two different ways.
You start by proving you are credible
Several meetings establishing that you are real, competent and safe. Procurement involved early. Two or three competitors in the frame. Price becomes the comparable field because trust has not been established yet.
Longer cycle, lower margin, lower win rate.
You start with the credibility borrowed
Somebody the buyer already trusts has vouched for you. The conversation starts at the problem rather than at your credentials. Often there is no competitive process, because the referrer’s judgement did the shortlisting.
Shorter cycle, better margin, higher win rate.
Good delivery makes referrals possible. It does not make them happen.
The reason nothing happens is rarely unwillingness. Ask a happy customer whether they would recommend you and almost all of them say yes and mean it. Then nothing follows, because you have asked them to do the hard part.
They have to think of somebody, judge whether it is appropriate, work out how to raise it, and find the words for what you actually do. That is real cognitive work on behalf of somebody else’s business, and it loses to the next email in their inbox every time.
Nobody wants to be the first.
AI work carries a risk that ordinary managed services do not. A customer buying a process redesign and an agent deployment is making a visible decision that their peers will hear about, and there is no established reference price or established norm to hide behind.
Which makes a peer reference worth more here than anywhere else in your portfolio. “We did this at a business like yours and here is what changed” removes more risk than any amount of technical credential. An MSP with two or three referenceable AI engagements has an advantage that is very hard to compete with, and it does not depend on being the best engineer in the room.
That advantage does not accumulate by itself. It accumulates if somebody decides, at the end of each engagement, to turn it into a reference and a referral.
Four things that turn goodwill into pipeline.
None of them is a referral scheme. Incentives usually make it worse, because they turn a favour into a transaction.
Ask at the moment of value
Not at renewal, and not in a quarterly survey. The week something visibly worked — a project landed, a problem disappeared, a number moved. Goodwill is at its highest and it decays quickly.
Be specific about who
“Do you know anyone who might need us?” asks them to search their whole address book. “Do you know another operations director dealing with manual order processing?” asks them to check one drawer.
Do the hard part for them
Draft the introduction. Two sentences they can forward without editing. You are not being presumptuous — you are removing the reason it never happens.
Give it an owner and a rhythm
Which accounts, at what point, asked by whom, reviewed when. Without that it is a good intention that survives about three weeks.
Four things worth remembering.
- Referrals produce the highest-margin, fastest-closing, least price-sensitive business you have — and are usually the only channel nobody runs.
- Customers do not refer because you asked them to do the hard part: thinking of somebody, judging the fit, and finding the words.
- AI work carries visible risk, so a peer reference removes more of it than any technical credential.
- Ask at the moment of value, be specific about who, draft the introduction, and give it an owner.
How many referrals did you ask for last quarter?
A Growth Review is a structured conversation with you and your leadership team: where your AI practice is constrained, which parts of your pipeline are being left to chance, and what running them deliberately is worth. Ninety minutes, no cost, and an honest answer — including when the answer is that you do not need us.
Win the business case before you win the build.