Back before software ate payments (although payments is now eating software), the payments world looked very different.
Banks, who technically still represent merchants for acquiring, would partner with processor/acquirers to do their bidding for them in exchange for shouldering the risk and making a small fee.
The acquirers would then punt much of this sales work to ISOs, or independent sales offices.
The ISOs would then wrangle tens of thousands of independent, 1099 contractors, colloquially referred to as payment agents.
We wanted a simpler image but this is what Nano Banana gives us, so whatever:
Our data shows that, at its peak, ISOs represented about a quarter of all US payment processing revenue, so not exactly a trivial channel.
But the compensation model for agents was always a head scratcher to anyone coming from SaaS.
Assuming we’re dealing with transparent players for a moment, the agent would often command 50%-80% of the residual revenues of a merchant account in perpetuity.
Many of these agents were not getting a VAR sheet, loading it in gateway, programming devices, training a merchant, or even supporting the account.
Agents were doing even less work as a proportion of the overall effort if a software needed to be sold: the agent was definitely not learning the features of Square, as an example, to appropriately qualify and sell the merchant.
Expect the agent to train the merchant on Square?
That’s a fever dream, bro.
Turns out you need to sell software (or software integrations, which is just another kind of software) if you want to keep the payments business.
It’s why software companies directly partner with modern gateways like Stripe and Adyen – with relatively great developer experiences – when they think about monetizing payments.
Software recognizes software.
(As the software company’s understanding of payments matures, they may deviate from this path, but the aforementioned gateways definitely own the brand for software-led payments.)
Within a decade the model of ISV-led acquiring will be nearly all acquiring volume.
And part of us wonders how much the existing payments compensation model contributed to this dynamic.
Let’s look at SaaS, a field that was crushing the market until this inconvenient thing called AI popped up.
In SaaS, just as in other lines of business, there has long existed the idea of sales channels.
A core rule of SaaS partnerships is that compensation scales directly with the amount of customer acquisition cost (CAC) the partner offsets.
If a partner does the heavy lifting of marketing and selling, they sensibly get a larger cut.
Basic Introduction/Referral
A partner passes a lead over the fence and the SaaS company’s people nurture, pitch, and close the deal.
- Compensation: Typically a 10% share of the first year’s Annual Contract Value (ACV).
Qualified Opportunity
Partner does the preliminary work to educate the prospect and delivers a qualified lead that is practically ready to buy.
- Compensation: Typically 20% of the first year’s ACV.
Full Sales Cycle”On Partner’s Paper”
Partner handles the entire sales motion—prospecting, demoing, negotiating, and closing the customer completely on their own—they are effectively replacing the SaaS company’s internal sales team.
- Compensation: Partners who close deals autonomously can capture 25%- 40%+ of the first year’s ACV. They can even migrate up the value chain by charging the end-user directly for implementation and onboarding services.
One note: perpetual revenue payouts are less common. We almost never see them in enterprise as the customer success team is compensated on net revenue retention and doesn’t want a partner screwing that up. We see it for smaller deals, and often it tiers down to zero after 2-3 periods, so like 10% of first year’s ACV, 5% ACV for year 2, then nada.
SaaS companies were grown in a very sophisticated way.
Raised lots of money.
Had intelligent business stakeholders around the table who could all do math adeptly.
Nobody once raised their hand in a board meeting and said,
Hey let’s give referral partners 50-80% of the total revenue of the deal so we can do all the ACTUAL work of implementing, training, supporting, and still run an expensive R&D operation with highly compensated technical professionals!
The math doesn’t math.
That’s because software has VERY real costs, unlike payments which have zero COGS (and if you argue interchange are COGS, you’re an imbecile).
Software companies trade on a multiple of revenue.
That multiple is directly influenced by gross margins (GM).
Some rules of thumb:
- 75% – 85%+ GM: Commands the full market multiple baseline.
- 60% – 70% GM: 15% to 30% multiple compression.
- < 50% GM: Severe multiple destruction. Trade at a 50% – 70% discount, closer to a 1.5x – 3.0x revenue, similar to IT services.
If a software company is giving away > 30% of its revenue to a channel while still eating all the COGS, they’ve just taken their software business from a market multiple to a services model, cutting the value of the business by 70% and earning the management termination notices for breaches of fiduciary responsibility.
Literally not even worth taking the sale as it is wholly destructive to the value of the company.
Which gets us back to the agent model.
We suspect the compensation model for agents worked when the barrier to winning an account was someone with a local presence and a stronger relationship than an outsourced call center (although depending on the ACV, software companies like Toast justified putting feet on the street to build local relationships).
20 years ago payments were completely fungible as they existed on stand-alone terminals, or the softwares themselves allowed multiple gateways so virtually anyone could integrate if it were even required.
Today generating ROI matters more than ever.
And you’re not generating ROI by offering a commoditized payment solution: you’re needing to bundle that with software.
If the payment agent’s value is simply “Bob’s my cousin”, how many times will Bob choose his cousin over a solution that keeps food in his family’s mouth?
The payment compensation model has to be rewritten.
Maybe the short answer is you can no longer be a full-time agent (although that sounds ironic since 99% of agents are the a-deal-a-year types).
If the upstream solutions provider has to sell software, that’s just infinitely more expensive than slamming an overpriced piece of hardware on a countertop.
Payment softwares like Rectangle Health have already modernized the agent model when software is involved: the agent refers the deal, Rectangle takes 50 bps.
Too much?
Rectangle is closing, installing, training, and supporting.
What is the agent doing beyond a referral?
Can the agent talk about how Rectangle works?
Know its features and shortcomings?
Understand errors?
Your average agent needs to get comfortable with the idea of earning 10% of ACV in year one.
Any income in year 2 is an absolute gift.
Can’t sell standalone payments anymore.
Along with everything else, markets change.
And the payments compensation model is due for an overhaul.
Add comment