Every managed service provider has a client who stopped paying. Not the one who disputed an invoice and settled it, but the one who went quiet, then apologetic, then quiet again, while the work kept going out the door because nobody wanted to be the person who cut them off. The balance crossed thirty days, then sixty, then ninety. Somewhere in there the conversation stopped being about an invoice and started being about whether the relationship was worth the money. That drift is almost never a finance problem. It is a decision the owner has already made by not making it, and it shows up later as the single ugliest line on the aging report. The same instinct that lets an owner absorb a late payment is the one that lets scope creep go unbilled and rate increases go untaken, which is why the damage is rarely isolated to one account and why agreement gross profit is usually the first place it becomes visible. Owners who have already learned to bill for the work that falls outside the agreement tend to be the same ones who collect on time, and the ones who eventually conclude that revenue is vanity when the client is unprofitable usually got there through an AR conversation they wish they had held two years earlier.
What makes non-payment worth an executive's attention is that it is a rare business problem where the correct action is well understood, cheap to take, and almost universally deferred. The playbook is not secret. Send the notice. Suspend on the terms you already wrote. Convert the promise into an instrument. Enforce it or decide, deliberately, not to. Almost none of that is legally complicated. What stops it is that each step requires an owner to say something uncomfortable to a person they like, on behalf of a business they are not fully separating from themselves. The firms that get this right do not have better lawyers. They have a system that fires without requiring an act of courage, and a leadership posture that treats a payment term as a term rather than an aspiration. This piece is about how that system is built, what it costs when it is missing, and why the contract you signed eighteen months ago has already decided most of the outcome.
Tom Fafinski, who has spent thirty five years as an attorney to managed service providers, explains late payment by talking about curfews. You tell your kid to be home by midnight. They come in at 12:15 and you let it go, because 12:15 is close enough and you do not want a fight on a Tuesday. Next time it is 12:20. The time after that, 12:30. Nobody renegotiated the curfew. It simply moved, one small concession at a time, until the stated rule and the operating rule were different things. Alexander Caplan, who runs an international MSP out of London, described the client-side version of the same drift more bluntly: when you let a balance accumulate, clients do not conclude that you are being generous. They conclude that you do not need the money, and they begin using you as a credit facility.
This is the part owners consistently misread. A late invoice feels like an isolated event involving one client having one bad quarter, and it gets handled as such, with an accommodation that seems small and humane. But an accommodation is not a neutral act. It is a disclosure. It tells that client, and eventually their peers and your own staff, exactly which of your written terms you will actually stand behind. The MSP that carries a client to ninety days has not made a one-time exception. It has published a payment policy, and the policy it published is not the one in the agreement. Everything downstream, including the awkward collections conversation the owner is dreading, is a consequence of a policy nobody consciously set.
Josh Peterson framed the scenario in numbers that most owners will recognize immediately: a three million dollar MSP, a client who owes seventy five thousand, more than thirty percent of that balance sitting past ninety days. Seventy five thousand does not kill a three million dollar business. But for a low margin MSP it is not nothing either, and by the time it is visible in that form the useful decisions are already behind you. The answer both guests converged on is not a tactic. It is a process that runs on its own schedule, so that no individual has to decide, in the moment, whether today is the day to be the bad guy.
Caplan's version is a dunning sequence that starts with polite reminders and escalates on a defined cadence through to service withdrawal. Fafinski's version starts even earlier, at the contract, with payment terms and a suspension right that both parties agreed to before there was any money in dispute. The two amount to the same discipline. The escalation is written down, it is automatic, and when a client asks why they are receiving a notice, the honest answer is that the system sent it. That framing matters more than it sounds. It converts an accusation into an administrative fact, which is the only version of this conversation that does not damage a relationship worth keeping.
Josh put a deliberately uncomfortable hypothetical to both guests. You resell your client's Microsoft licensing. They are sixty days late. You have communicated repeatedly. Can you simply turn off their email? Fafinski's legal answer was that you can, with notice, on terms your agreement already established. But it was Caplan's reframing that made Josh stop the conversation and back up. Synergy does not turn services off. It declines to renew them. When a client stops paying, the MSP is no longer in a position to keep purchasing third party services on their behalf, so auto-renewal is disabled and the resold licenses lapse. The practical effect on the client is similar. The posture is not remotely the same.
The distinction is worth sitting with, because it is the difference between an act of aggression and a refusal to extend further credit. Shutting off a service is something you do to a client. Declining to fund a renewal is something you stop doing for them. One invites a story in which the MSP held a business hostage, a story that will be told to every prospect in that owner's network. The other is simply the end of an unreciprocated arrangement, and it is very hard to be aggrieved about. For a business where reputation is a primary acquisition channel, and Caplan was emphatic that it is, the framing is not cosmetic. It determines what gets said about you afterward.
There is a specific scenario Fafinski wanted on the record, because it is where most MSPs quietly lose the money. The client has fallen behind, has caught up on current invoices, and still carries a past due balance. They want to pay it down over time. Five hundred or a thousand a month until it clears. The owner, relieved that the relationship is intact and the number is moving in the right direction, says yes and shakes on it. What they have actually accepted is an unsecured promise from a party that has already demonstrated it does not pay on time, with every defense about service quality still fully available to them.
Fafinski's alternative costs nothing and changes the instrument entirely. Put the payment plan in a promissory note. A note strips out the defenses, because the note obliges the client to pay a sum and obliges the MSP to do nothing at all. The argument "your service was bad" has no purchase on a promissory note. Where the situation warrants it, secure the note with a confession of judgment, so that a missed payment converts directly into a filed judgment without the cost of litigating anything. From there the remedies are real ones: garnishment, levying receivables, and a public judgment on record. None of this is exotic and none of it is expensive. It is simply the difference between documenting a debt and hoping for one. And if it does go to collection, Fafinski's expectation-setting is worth internalizing: a seventy five thousand dollar balance handed to a contingent fee attorney may settle around forty and net closer to twenty five. You pursue it anyway, and the reason is not the recovery. It is the message, sent to the market and more importantly to your own team, that your terms are terms.
By the time a client is ninety days late, the range of good outcomes was set months or years earlier, in a document nobody has read since signing. Fafinski's practice is built around that fact. Rather than sending a twenty eight page master services agreement into a prospect's inbox, where it will be summarized by AI, misunderstood, and forwarded to a lawyer, he drips it: a four or five page MSA covering payment terms and limitation of liability, agreed while no money is at stake, then a short services addendum that bolts onto it. Clients rarely escalate a short document that costs them nothing. The terms that matter get agreed in the abstract, which is the only condition under which they get agreed easily.
The same logic runs through the rest of the agreement. Caplan's review of Synergy's own contracts found the firm had constrained its ability to bill for ancillary work by being insufficiently explicit about what was and was not included, and Josh's consulting experience puts that squarely as the largest driver of poor agreement gross profit he encounters. Fafinski's fix is visual and almost crude: columns of services with checkboxes for included and excluded, so that a declined penetration test is documented as declined rather than assumed. Two years later that log is both a liability shield and a sales asset, because the easiest revenue in any MSP is the service a client already said no to once. Add an annual escalator defined as an inflation adjustment and you have removed the third quiet margin leak. Tom's arithmetic on that last point deserves to be pinned above every owner's desk: a company at twenty percent margin facing two percent cost inflation with no automatic increase does not lose two percent. It loses about eleven percent of its margin, dropping to roughly 18.8, and if the reported margin holds anyway, it is holding because new higher margin clients are subsidizing old ones. Tom's word for that structure was Ponzi scheme, and Josh's reaction was that far too many MSPs would recognize themselves in it.
Whatever period your agreement specifies, applied consistently. The common structure is a written suspension right exercisable on thirty days notice, with the notice itself issued on a fixed trigger such as a balance reaching sixty days past due. The specific number matters far less than whether the trigger fires automatically. A thirty day rule enforced on day thirty is stronger than a fifteen day rule enforced whenever the owner finally loses patience.
Generally yes, where the agreement grants a suspension right and proper notice has been given. Fafinski's guidance is to issue a formal notice citing the contractual right, follow up with a reminder several days later, and send by certified mail where the balance is significant. The stronger practice, per Caplan, is to structure resold services so the MSP simply declines to renew them rather than affirmatively disabling them. This is not legal advice for a specific situation; the enforceability of any suspension depends on your agreement and jurisdiction.
It is a provision under which a debtor agrees in advance that, on default, the creditor may obtain a judgment for the full balance without litigating the claim. Used with a promissory note securing a past due balance, it converts a missed payment directly into an enforceable judgment. Availability and enforceability vary considerably by jurisdiction, so it is a question for counsel in your state rather than a clause to copy.
Materially less than the face value. Fafinski's working expectation for a seventy five thousand dollar balance handed to a contingent fee collections attorney is a settlement near forty thousand and net recovery closer to twenty five. Owners who go in expecting full recovery tend to abandon the process at the first counteroffer, which undercuts the deterrent value that made pursuit worthwhile in the first place.
Largely because of document length and the ease of AI summarization. A twenty eight page agreement arriving for a three or four thousand dollar a month service reads as disproportionate, gets pasted into an AI tool for a summary, and the summary raises questions the buyer cannot answer alone. Shorter staged documents rarely trigger the same reflex, which is the operating principle behind the drip approach.
Both guests advocate defining the mechanism rather than debating the number each year. Language framing the increase as an inflation adjustment, defined as the greater of a published index or a stated floor, holds up better under client review than a bare percentage. The economic case is straightforward: at a twenty percent margin, two percent of unrecovered cost inflation removes roughly eleven percent of that margin in a single year.
Tom Fafinski is co-founder of Virtus Law, PLLC in Minnesota and has practiced law for thirty five years. He describes the firm as a general counsel's office for managed service providers, covering master services agreements, mergers and acquisitions, governance, synthetic equity, employment matters, and estate planning for MSP owners. His work in the channel traces to the 1990s, when he handled technology matters for staff augmentation firms and began advising his brother's break-fix company as it transitioned to a managed services model. By 2008 the practice was almost entirely MSP work.
Connect with Tom on LinkedIn →
Alexander Caplan is the founder of Synergy Associates, an international managed service provider headquartered in central London and operating for more than twenty years. Synergy supports mid-market businesses with international footprints, with offices and teams in Paris, New York, Los Angeles, Hong Kong, and the UAE, and roughly fifty staff and contractors worldwide. He brings the operator's perspective to every question in this conversation, from how a redline negotiation should actually be assessed to how a dunning process escalates in practice.
Connect with Alexander on LinkedIn →
Josh Peterson is the CEO of Bering McKinley and host of The BMK Vision Podcast. Since 2004, Josh has worked with hundreds of MSP owners to build operationally sound, profitable businesses through consulting, peer teams, and direct coaching.
Connect with Josh Peterson on LinkedIn →
Collections discipline is rarely a finance problem and almost always a leadership one, which is why it shows up alongside unpriced scope, deferred rate increases, and margin that erodes without anyone deciding to let it. The BMK Vision Operating System exists to make those decisions structural rather than personal.