Why Payments Is Eating Software

Not long ago venture firm Andreessen Horowitz celebrated that software was eating the world.

Makes sense: software is scalable, accurate, and can be logarithmically cheaper than manual processes.

So over the past several decades, trillions of dollars of enterprise value has been created in modernizing every corner of the earth.

(Retailers have avoided it as best they can, obviously.)

That’s all well and great, but Andreessen missed an even more intoxicating elixir:

Payment processing.

In the early 2010’s, software companies got a taste for just how lucrative payment processing was.

No product.

Minimal support.

EBITDA out the wazoo.

And, depending on your moral stance of the day, infinite EBITDA.

What do we mean?

Oh, we don’t know, how about ripping out more than 12% of a merchant’s revenue and having full legal immunity in doing so?

Sure, the merchant goes bankrupt, but you can find another sucker to take his place.

See, while your software might have earned you a few hundred dollars a month, what if you could make 10x – or 100x – that amount by doing nothing?

For example, Toast rips a little over 50 bps in payments EBITDA (called net take rate, but it’s just EBITDA by another name) from their customers, or about 2.5X what a fair market rate would be for the size of merchants that use Toast.

The aforementioned link shows Heartland ripping 10,000 bps (not a typo) in payment EBITDA on their restaurant victims.

If Toast followed in Heartland’s footsteps (and like all payment companies they are, albeit more gradually so as not to totally blow out a quarter and then struggle to manage future expectations), Toast’s ARR would balloon from $1.7B today to $17.7B without so much as adding a new location.

That’s the power of payments, baby!

And here’s the thing.

Unlike software, which shows up on a monthly bill and can be audited, credit card payments are virtually impossible to decipher.

This is intentional.

See, if you purchase software, you get a clean invoice and a bank statement that matches.

While the line items on the bank statement are not always immediately decipherable, they are neatly broken out and quantifiable.

See the below sample bank statement.

We could all agree that this person spent $710.49 on mortgage payments in November of 2009.

This is precisely how a software purchase would show up: you’d have an invoice for $100/mo let’s say, then you’d have a corresponding payment for that $100 in your bank ledger.

Very transparent.

Payment processing is the antithesis of transparency.

Why?

Because it’s a commoditized service.

Hard to make money in a commodity business.

But if you had to?

Make it so complicated that nobody knows what they’re actually paying.

Look at how complicated interchange has become since inception:

The interchange fee is the percentage tacked on to each sale when a credit or debit card is used. Two decades ago, there were a handful of interchange rates. Today merchants face about 200, and the percentage changes depending on the sort of plastic the customer uses.

If you tried to plot it, the graph would be up and to the right harder than Nvidia’s share price over the past three years.

This isn’t an accident.

Because in the chaos and confusion money is made.

And handsomely, at that.

Here’s a sample credit card statement from Vantiv.

Can you tell us what the hell is going on?

Here’s an even better litmus test: can a person with a 140 IQ and no payments experience tell us what’s going on?

The answer is no.

Hell, 99% of people in payments acquiring can’t even read statements accurately.

What’s a real interchange category?

Is it being downgraded?

Is it being padded?

And this is why payments will conquer software.

Here’s a list of math scores by country.

The USA is ranked 34.

And then here’s our spending on education per capita.

We suck at math.

And as bad as the averages are, they skew even worse for merchant demographics, who are bottom quartile at best.

So now you have a buyer who can barely brush their teeth but is somehow legally allowed to enter into a binding contract.

No surprise when they’ll sign any piece of paper to solve an immediate pain.

And that’s what’s happening.

Merchants get locked into a software, maybe recognize that switching costs are high, but get juiced through the gills on payments.

Rates go up.

Support and features go down.

Because, hostage situation.

But if the merchant is really trying to understand if their rates are fair?

Impossible to audit.

Their effective rate keeps increasing, sure, but they don’t know if it’s because the duopolies are increasing prices, card types are changing, or their acquirer simply wants rocket rides paid for.

Because of this dynamic, software companies are learning how to become payments bros.

That payments is quite literally infinitely easier than software.

And so the game has now become how can I own the merchant acquiring with something that is really hard for the merchant to dump.

Software, Andreessen should note, is that mousetrap.

… But only in certain geographies.

Some markets – like Brazil and India – have instant, free rails for money movement, which means merchants are “forced” to transparently pay for value.

What a crazy notion.

So long as the duopolies persist in the good ol’ USA, domestic merchants will have the privilege of paying 10x what their international counterparts pay for the same software.

How’s that for a tax on mathematical illiteracy?