MP Bills $650K a Month for HR Software It Didn't Build

Early in the interview Nathan Latka asked Jason Maxwell whether MP had engineers on the team. “We don’t,” Maxwell said, and then half-apologised for it: he had flagged it in his email, he wasn’t sure a company with no product of its own belonged on a show about software companies. At that point MP had billed $650,000 in the previous month, held just over 1,300 client logos and employed 70 people, not one of whom had written a line of the software those clients log into every morning.

Before the pivot

The payroll business that got commoditised

MP — formerly MassPay, introduced on the show as “wired for HR” — was founded in 2004 and sits in Beverly, Massachusetts on its GetLatka company profile. Maxwell described the first eight years in five words: “we were a pure payroll service focused on SMB.” Latka wanted the peak. Maxwell gave him the growth rate first — “our best year in terms of year-over-year growth was 2011, we experienced 44% year-over-year growth” — and then, pushed for the dollars, hedged: “that goes back a long way, so I’m gonna guess it was growing from, at that point, you know, two million to three million.”

That guess is the entire basis of the “$3m business” in the episode title. GetLatka’s record carries a $3m revenue row for MP dated June 2013, which is the transition period rather than the 2011 peak, but the two agree on magnitude. Nothing on the tape or in the record puts the old payroll line above that.

Payroll service was just getting increasingly commoditized, so, you know, there were a lot of players in the space.

The episode is titled Gusto Killed Their $3m Business. Maxwell’s own account is narrower than that, and the order in which he names things matters: “the pricing power of the large players in the space was really commoditizing the industry, and then you had, you know, the introduction of companies like Gusto, which came in with really low pricing, working a lot through the CPA channel.” The incumbents come first; Gusto arrives as one entrant among several forces, in a sentence about a whole category losing its pricing power. Maxwell also dates the squeeze precisely — “there was a big shift between like 2008 and 2012” — and MP signed with isolved in 2012, which means the exit was already underway. Latka’s closing summary compresses all of that into Gusto killing the business. The guest never says it.

What he says instead is that the demand had moved somewhere his payroll platform could not follow. Lower-mid-market companies, which he defines as 50 to 500 employees, were leaving server-based payroll for cloud suites “that did a lot more than just payroll”, because what they needed had changed:

• Healthcare reform — reporting obligations small employers had no system for.
• A remote workforce — the transition itself, not just paying people once they were remote.
• Electronic onboarding and offboarding — paperwork that had to stop being paper.
• Multi-state taxation — the compliance problem that arrives the moment a small company hires outside its state.

So MP did not shut anything down. Latka asked directly whether the payroll business was killed off in 2013. “No, no, we transitioned our clients off a legacy payroll software platform to isolved, and we started to build out the competency of our team to support other areas of HR.” It took two years to move the entire client base, and it came with a second move alongside it: out of SMB and up into the lower mid market.

• 2004 · Founded MP starts life as MassPay, a pure payroll service for small business.
• 2011 · Best growth year 44% year over year; Maxwell’s guess at the dollars is $2m going to $3m.
• 2012 · isolved MP signs on to sell, implement and support the isolved suite, then spends two years migrating clients off the legacy payroll platform.
• Jun 2013 · $3m the revenue row GetLatka carries from the transition years.
• Jun 2020 · $6m about $500,000 a month, after a year Maxwell says grew only 4%.
• Aug 2021 · $7.8m $650,000 billed in the previous month, 1,300 clients, 70 people.

What the isolved partnership actually is

This is the load-bearing detail of the whole business, and Latka got it out of him in two lines. “So you’re paying basically to white label their tech, so really you’re a distribution company?” Maxwell: “Yes. We sell, implement and support that platform, and we provide services around it.” Latka’s analogy — a Salesforce service provider who helps you install Salesforce — is the one Maxwell accepted.

It is worth being exact about what that is commercially, because “partnership” covers several very different arrangements. This is not a referral fee and it is not a cut of somebody else’s invoice. MP resells the platform under its own relationship, bills the client itself on a per-employee-per-month basis, and sells its own services on top. Maxwell’s word for the technology is “non-proprietary”, and he lists his competitors as Paycom, Paycor, Paylocity, ADP, Paychex and Zenefits — companies that all build what they sell.

0engineers behind a $7.8m run rate, recorded August 2021

One thing the tape never supplies is the other half of that deal: what MP pays isolved per seat. Without it the gross margin on the licence is unknown, and every margin figure below should be read as a whole-company number rather than a software one. Two figures Maxwell does give bound it — about 26% of revenue goes back into sales and marketing, and about 8% reaches the bottom line.

The other correction the episode title needs is smaller but worth making: MP does not sell an applicant tracking system of its own. The ATS is one module of the suite it resells, and it appears in this story for a different reason entirely.

The seat math

“We have just over 1,300 logos. Our average client has 29 employees. Our largest client has 3,200 employees.” Latka ran the arithmetic live — roughly 40,000 seats, and at the entry price of $8 per employee per month that is about $320,000 a month. Maxwell corrected him upward: the range runs from $8 to about $40 depending on services and functionality, and the average is $13. At the top of the book, one client pays MP about $400,000 a year and subscribes to essentially everything.

Monthly platform revenue = seats × per-employee-per-month fee40,000 seats at a $13 average is roughly $520,000 a month. MP billed $650,000. Our arithmetic on his figures, not his — the gap is where the services sit.

The $6,500 average deal value is the number to hold onto, because it is also the annual contract value that has to pay back the cost of winning it.

Where the $10m comes from, and where it doesn’t

Asked what the previous month looked like, Maxwell said $650,000. That is the figure GetLatka wrote down: a revenue row of $7.8m dated 4 August 2021, which is $650,000 annualised. A year earlier he put MRR at about $500,000, and the record agrees — $6m, recorded June 2020. In between sits a year he does not dress up: “2020 was an interesting year, really because a lot of our clients downsized, so we only realized 4% growth last year, which was our lowest growth year.”

Then came the sentence the headline was built on: “Our average deal value is 6,500, and we’ll do 10 million in revenue this year, so we’ll cross the 10 million in recurring revenue mark in 2021.” Read it against the meter and the tension is obvious. $650,000 a month is a $7.8m run rate; crossing $10m of recurring revenue by December means getting MRR to about $833,000, a climb of roughly 28% in under five months, from a business that grew 4% the year before and about 30% in the twelve months to August.

GetLatka’s next reading, recorded November 2021, is the same $7.8m. No $10m row was ever written to MP’s record. So the honest version of the headline is this: $10m is Maxwell’s stated target for the year, said out loud in August; $7.8m is what the company was actually running at when he said it, and it is the last figure on this business the record confirms.

Churn is the moat MP didn’t have to build

A distribution business with no product should be easy to leave. It isn’t, and the reason is the same one that makes the pivot work at all.

It’s really hard for companies to transition from one HCM platform to another, and really the reason for that is companies hire us for our applicant tracking system, payroll, employee benefits administration, onboarding, offboarding. So to transition from one platform to another is a significant amount of work.

Jason Maxwell, founder, MP

Maxwell splits churn in two. Controllable churn — clients lost to rivals — averages about 4% a year. Non-controllable churn, meaning clients who get acquired or go out of business, runs 6% to 8%. Against roughly 10% gross churn, expansion of about 15% put MP at 105% net revenue retention in 2021, which Maxwell called positive “by a modest amount” and his best showing in recent years. GetLatka’s company record carries the same pair, 10% gross churn against 15% expansion. The average client stays more than six years.

The cash gap underneath it

The economics are not comfortable, and Maxwell did not pretend otherwise. “Our average deal value is, you know, 6,500, and it costs us about 11,000 to bring on 6,500 of revenue, so typically we break even at like the 20-month mark.”

Payback = $11,000 to acquire ÷ $6,500 of annual revenueAbout 20 months, Maxwell’s own figure. It is a revenue payback rather than a gross-profit one — the tape never says what the platform costs MP.

Asked whether spending $11,000 to buy $6,500 puts stress on the business, he said it does: “It puts stress in the business. As we’ve grown, we’ve been able to absorb that cash flow impact more easily.” The shape he keeps it in is deliberate — “we’re investing about 26% of revenues back into sales and marketing, and we’re bringing 8% to the bottom line” — which on Latka’s read is $50,000 to $60,000 of net income a month, and GetLatka’s record carries a $50,000 cash-flow figure to match. A twenty-month payback on a six-year customer is a good trade; it is just one that has to be funded out of this month’s cash, which for a bootstrapped company is the whole difficulty. Anyone sizing that trade-off is really pricing customer acquisition cost against contract length.

The engine buying those contracts is 12 account executives, 5 sales development reps and 2 customer success people — GetLatka logged 17 in sales in August 2021 — against quotas that average about $400,000 and reach $650,000 of annual recurring revenue for the experienced AEs.

Why he won’t sell for $25m

Latka asked the obvious question about a channel partner this size: why hasn’t isolved simply bought them? Maxwell’s answer confirms the risk without flinching from it. “isolved has approached a lot of companies like us in their ecosystem, so they are growing through M&A,” he said — before setting out the alternative: “we’re pursuing a path of 25 million dollars of recurring revenue in the next five years.” The differentiation he is betting on is the services layer, not the software.

He bootstrapped MP and still owns 100% of it. On price, he said businesses like his are valued on a multiple of revenue rather than earnings — “typically companies like us trade anywhere from two and a half times revenue to, you know, five times revenue” — a claim Latka pushed back on twice, because top-line multiples on a services company are not what he expected to hear. Offered $25m all cash on the spot, Maxwell said no.

I like growing the business. I feel like we can position ourselves well. I feel like there’s a ton of opportunity to be a stronger regional provider and make some noise in our space.

Sources Nathan Latka’s interview with Jason Maxwell, recorded August 2021; the GetLatka company record for MP (revenue and team rows, each dated above, plus the churn, expansion and cash-flow fields on the company profile).