👋 Welcome back to LFL.

I host a lot of CFO dinners. And most of the seats at the table are occupied by SaaS people. ARR, ARPU, net dollar retention… it’s what’s for the appetizer, entree, and dessert. And I eat it up.

But the companies I find most fascinating are the ones just adjacent to tech. The ones where the business model doesn't fit neatly into a SaaS waterfall, and the CFO has had to think from first principles about how value actually accrues over time.

Like the business of greeting cards.

Mateo Bryant is the CFO of Minted - the design marketplace behind the holiday cards, wedding invitations, and birth announcements that make the rest of us look far more organized than we actually are (you know that I know that you ordered those cards two days before Xmas).

When Mateo started walking me through how Minted thinks about LTV/CAC, I stopped making SaaS small talk. Because there's a lot that SaaS CFOs can learn from a greeting card company in terms of how customer value actually compounds over time.

Specifically, four things:

If you want a primer on how LTV and CAC are calculated before we go deep, I've covered that here and here.

Let's get into it.

Let's start with retention, because it's the foundation.

When you use Minted to send holiday cards, you enter every recipient's name and mailing address. That list lives in your account. Next year you come back, add a few new names, upload a new photo, and you're done in ten minutes. Over time you’re creating a rolodex with a lot of data gravity.

Starting over somewhere else means rebuilding that list from scratch. Eighty addresses. Maybe a hundred and twenty by year three. Nobody does that. Not to save a few dollars on a card they buy once a year.

The switching cost isn't emotional and it isn't price-based. It's friction (the good kind), and it compounds every single year without Minted lifting a finger. The moat widens automatically as long as the product stays good enough that no one wants to rebuild.

Most LTV models treat the customer relationship like a straight line. Acquire them at point A, collect revenue at a steady rate, apply a churn assumption, done. It’s a very deterministic model with pre determined revenue events.

Minted's model doesn't work that way.

A customer who came in for a wedding invitation isn't going to spend at a flat rate for the next decade. They're going to spend in bursts, tied to what's happening in their life.

Each one represents a distinct transaction, and each one is triggered by an event Minted didn't manufacture from a recurring billing schedule.

Mateo has nearly two decades of cohort data. Rather than modeling assumptions about year 5 revenue, he's forecasting against a template he's already watched unfold across thousands of customers. The couples who entered through a wedding in 2010: he knows what they bought in 2013, 2017, 2022. The spending patterns around lifetime events are predictable. The only variable is timing.

Most B2C businesses are making educated guesses about long-term LTV. Minted is reading from the history of relationships that customers have with their recipients. That predictibility in life events changes how you think about acquisition outlay. A business with 18 years of cohort data can afford to pay more for a customer today than a business with 18 months, because the confidence interval on that future revenue is fundamentally different.

The cost to acquire a customer is fixed at the moment of acquisition. What happens after that is a denominator issue: every subsequent transaction spreads that fixed cost across more revenue, making the original spend look cheaper and cheaper in retrospect.

A wedding has four natural purchase moments:

That's four transactions before the couple has celebrated their first anniversary. The CAC is already getting amortized before Minted has done any additional marketing to that customer.

Then the baby comes.

By year five, a customer who came in for a save-the-date has likely generated eight or nine transactions. The CAC on that customer looks nothing like it did at the moment of acquisition. Which means the unit economics of the business look better every year a cohort ages. This isn’t because Minted is charging more (the antithesis to SaaS annual price hikes), but because life keeps generating purchase occasions.

There's one more force pushing CAC down that doesn't show up in a paid media budget. Every card Minted ships has "Minted" printed on the back, along with the name of the artist who designed it. That card lands in 80 homes. It sits on refrigerators. It gets pinned to bulletin boards. The person who received it files that brand away, and maybe next year, when it's time to do their own cards, they think of Minted. Maybe they use it for their daughter's wedding.

It’s kind of like the “Sent from my iPhone” signature.

This organic acquisition channel has been running continuously since 2007. The CAC for those customers is effectively zero, which makes the blended unit economics across the whole cohort look even better than the paid acquisition numbers suggest.

Minted doesn't have a subscription product. There's no monthly fee, no annual contract, no auto-renew. And yet their retention looks, in many ways, better than SaaS businesses that do.

The reason is that the human calendar is doing the subscription work for them. Christmas happens every year. It's not a question of whether the customer will need holiday cards; it's a question of whether they'll order them from Minted. And given the address book, the design quality, and the 18 years of habit formation, the answer is usually yes.

SaaS companies pay enormous amounts to manufacture this kind of recurring behavior through product stickiness, integrations, and switching costs. Minted gets it for free from the Gregorian calendar and the fact that human beings keep having babies and getting married at a remarkably consistent rate.

What this means practically is that Minted's revenue isn't really "one-time purchase" revenue, even though that's technically the model. It's better understood as episodic subscription revenue: predictable in aggregate, anchored to events that repeat annually or semi-annually, and reinforced by a structural retention mechanism that doesn't require a contract to enforce.

Most companies would kill to have a business model that self-renews without a renewal process. Minted built one. They just used life itself as the billing engine.

In this episode of Run the Numbers, I sit down with Mateo Bryant, CFO of Minted. We break down Minted’s life-event flywheel and decades-long LTV, managing extreme seasonality when half the year happens in one month, and balancing long-term CAC with short-term monetization. Mateo also shares lessons from scaling Uber and Amazon globally, localization missteps, and making marketplaces work in emerging markets.

Mostly Talent is actively placing a Director of Strategic Finance, Controller, FP&A Manager, and VP of IR across Series B through Pre-IPO companies. These searches move fast.

If you're hiring, work with us here.

If you're passively open to new roles, get into the ecosystem here.

Thanks for reading, and make sure to check out our sponsor, Abacum.

Register for their upcoming webinar series on the future of financial forecasting. I gotta say, the Scenario Studio they just dropped is straight 🔥... Idk why they had to show everyone my desktop tho…