We joke that all conversations should be recorded at Glenbrook because they often lead to some interesting podcast episodes. If you have 15 minutes, tune in for a fly-on-the-wall perspective in this bonus Fanning the Flames on recent developments in agent protocols.
Russ Jones and Yvette Bohanan reflect on the history of the Secure Electronic Transaction (SET) protocol and how it compares to modern protocols being developed for agentic commerce by companies like Google, Stripe, and Basis Theory.
They explore why protocols and standards are important, what makes them take hold or “get off the ground like a lead balloon”, and how recent announcements in this space are shaping our current view on agentic payments.
Take a deeper look at agentic commerce (and other industry trends) at our upcoming 1-day Advanced Payments workshop on December 4th in San Francisco. Space is limited – enroll today!
Yvette Bohanan: Hello, I’m Yvette Bohanan, a partner at Glenbrook and your host for this episode of Payments on Fire.
Well, Russ Jones, we meet again. Here we are. Fanning more flames. There’s a lot to fan out there these days.
Russ Jones: The fuel is plentiful.
Yvette Bohanan: It is. I feel like I’m sitting on the top of a hype curve with you looking out over a landscape of, kind of crazy landscape right now. One of the more interesting things going on is, and I’ve been like super absorbed in this world, not just from a payments perspective, but from a general perspective, no pun intended, agents. Agentic, agentic commerce, agentic payments, agentic. Now we have agentic protocols and a march towards making agents do more for us with less boundaries. It’s kind of interesting. And we started talking about this the other day and I said, stop. We need to just record a Fanning the Flames. And so here we are recording.
Russ Jones: I don’t know if you want to say it is sort of interesting or sort of confusing.
Yvette Bohanan: Well, yeah, I think it’s interesting, and interesting is interesting sometimes because it’s confusing and we’re trying to tease it apart and really kind of get to the bottom of what’s going on. And one of the benefits we have is we’ve been doing this for a while. I won’t say how long exactly, but it’s been decades, and we’ve seen this movie before when it comes to protocols and standards. And this is what we started talking about and our minds went to the same place, it happens every once in a while, and we thought of SET.
Russ Jones: Yes, SET. The Secure Electronic Transaction protocol.
Yvette Bohanan: Exactly. And I remember the first time I heard about SET, I was doing a startup called Pay Links. It was back in the nineties. It was actually even before we named the product and had it as a official startup. We were building this for one customer, USC. And I was on a plane reading, in paper, CIO magazine and in the margin on a sidebar in CIO magazine was “New Industry Standard for Payments Called SET”. And I read it and everyone thought this was going to be like the cat’s pajamas, like this was going to solve-
Russ Jones: Let’s put a little double quotes around “everyone”.
Yvette Bohanan: Okay. Everyone writing articles about it thought it was going to be amazing, that I was reading. When did you hear about this?
Russ Jones: I guess I was aware of SET, this would’ve been like 1999, maybe 2000, 2001, that sort of timeframe. And the context here, why this protocol was invented, so to speak, is this was the early stages of the internet for a lot of people and online commerce had just really started with Netscape and SSL certificates and there was a lot of consensus from the technologists that this was never going to grow-
Yvette Bohanan: This being the internet.
Russ Jones: Yeah, the internet. Buying things online would never take off if a buyer could refute a transaction and say, This wasn’t me, I didn’t do it, it was somebody else using a browser.
Yvette Bohanan: So consumer protection is going to keep internet commerce down was-
Russ Jones: Yeah, it was merchants were going to be hesitant to adopt it if they were going to be held responsible for effectively what turned out to be card not present fraud.
Yvette Bohanan: Right.
Russ Jones: And consumers were going to be hesitant to use it, being the internet, use the internet to buy things online if they couldn’t, with any certainty, feel confident that the website they were buying from was really the merchant’s website. So it was a trust issue, as you always go to whenever we talked. Go into trust. It’s a sort of a ground issue. So the technologists were like, Well, we can solve this problem with public key cryptography. Here our old friend math. So that was sort of the origins of it.
So the technologists really got out ahead of the marketplace and they were developing a protocol that was going to standardize how trust is established in online commerce so that buyers could feel confident about sellers, sellers could feel confident about buyers, and people wouldn’t have to use cards to buy things online. Their digital certificate would prove, issued by their bank, would bind them to a card at their bank. And it was, if people haven’t heard of this, there’s a reason why. It got off the ground like a lead balloon.
Yvette Bohanan: What was interesting and one of the reasons people thought it would take off and not just sink like a lead balloon was because it was a consortium based model, right? There were participants from all over the industry.
Russ Jones: Visa and MasterCard were right in the thick of it. And SET was sort of a, it was a consortium protocol. Visa had their protocol. MasterCard had their protocol. They did the same thing in incompatible ways, and so they came together with other industry participants like Netscape, at the time, and Microsoft and IBM and GTE and RSA, Verisign.
Yvette Bohanan: Everybody. Yeah.
Russ Jones: Everyone who was involved in early days of internet security were actively involved in this protocol. And it turned out to be super complex and a lot more heavyweight than the market wanted to adopt.
Surprise, surprise. It turns out merchants were willing to sell things online and develop what now, you know, is a multi-trillion dollar marketplace by taking on that incremental risk of developing confidence in who the buyer was.
Yvette Bohanan: Right, and of course we’re still trying to solve those risks, solve for those risks and control them today, But we’re still willing to take the risk too.
Russ Jones: So the reason that both of us went there was we were looking at the new agentic commerce protocol from Google. It’s got the backing of 50 companies, so it kind of feels like a big technology consortium. And they’re using public key cryptography to solve the same problems so that buyers can feel confident that sellers are really representing what they have to sell and they’re not hallucinating about their inventory.
And sellers can feel confident that buyers are really, have the right intention to buy and there’s not an agent there that’s hallucinating what the buyer’s really looking for. And it just struck us as, we’ve kind of seen this.
Yvette Bohanan: We’ve seen elements of it for sure. And I think, there were these comparisons and we were like, Wow, is this a replay or is this a new way of doing this, right? Because consortiums do work, too. EMVCo standards work. W3C, all of these things work. ISO. It’s good. It’s good. This is all good stuff and it’s super important, right, for a lot of reasons, standardization, end user experience, consistency, adoption helps everyone. We like to say, in the words of Carol Coye Benson, cooperation below the line, competition above the line, right? So standards help.
Google realized that they had a big problem to solve out there. And then since then there’s been other announcements, Basis Theory, different groups coming out with sort of their standard. And the old joke is, the problem with the word standards is there’s two S’s. It’s the one on the end that gets you every time. Right.
So, now we’re starting to see sort of this like who has the pen kind of thing. I think Google’s interesting because where SET was like this so rigid, complicated, you had to do a million things. And honestly, the math, the algorithm, the theory behind SET was fine, but at the time, you’ve got to go back to the nineties. We had dial up modems, we had different tech, right, for the internet. That protocol was too heavy for the infrastructure.
Russ Jones: And it might be right today,
Yvette Bohanan: Yeah. You never know, right? But at the time, you have to hit the sort of magic triangulation in order for things to take off. I think that’s one of the big learnings, first of all. Simplicity versus complexity.
Russ Jones: Yeah.
Yvette Bohanan: And the algorithm has to match the infrastructure for the vast majority of people that need to use it. And you have to have a fairly, you have to have flexibility in the right places. And I think that’s what Google with AP2 is trying to do with the mandates, this sort of idea. I wish they would’ve used a different term because mandate means a lot of things in payments already, but it’s sort of the same, the way they kind of allow you to have some flexibility in how you’re implementing things across the protocol and what you’re communicating.
And SET didn’t have a lot of that, right? It had unmatched infrastructure to tech lift. It was pretty rigid. It was super rigid because it was a security protocol. It was a bit confusing, honestly, to end parties trying to implement it. Remember, you had to go sign up and get a passcode from your card issuer, I think it was, right? It was pretty arduous.
Russ Jones: Yeah, in the world of innovation, it required like seven party adoption.
Yvette Bohanan: Yeah.
Russ Jones: Simultaneous seven party adoption, which always spells doom. Anytime you get beyond more than one, it becomes a challenge. You get beyond two, you’re not-
Yvette Bohanan: You’re on a downhill spiral.
Russ Jones: Yeah. And so it was very complex in that sense, and it required a lot of simultaneous adoption.
And that’s another thing that kind of, something I have my eye on in agentic commerce, not agentic technology. I think a lot of people have belief and have some hands-on experience that the AI, the generative AI tools can automate a lot of things and streamline a lot of things. But they’re, to some extent one sided. They can simplify the vast world of everything. But when you get into commerce, commerce is two-sided, a buyer and a seller. And we often talk, in stablecoins, the canonical question is, do we need 2000 of them?
Yvette Bohanan: Right.
Russ Jones: In agentic commerce, the canonical question is, does the seller have to do anything? Do they have to take action? Do they have to change their web presence? Do they have to adopt new, do they have to get new vendors? Do they have to change the way they present their merchandise? Do they have to, do they have to, do they have to?
Because if the answer is yes, there’s 12.8 million standalone websites in the world. And if we’re waiting for the world to adopt, those 12.8 million websites to adopt MCP servers, it’s one of these things. There’s too many points of friction, too much adoption challenge.
So, you know, Stripe and OpenAI announced their commerce protocols, I think in the last 30 days. And it was a little bit hopeful, I think, what they were proposing because it did not require, they didn’t have as much assumptions on the seller side about needing to establish and speak a new agentic protocol on the seller’s behalf.
Yvette Bohanan: Everyone’s kind of coming at it, not everyone, several large entities are coming at it, but from different perspectives, right. The networks have their perspectives.
Russ Jones: Most of the emphasis is on the buyer side. And it’s easy to understand why. It’s the most tangible and if it falls short, if the vision falls short, it falls short in a way that’s still incredibly useful. If we just had agentic agents helping people buy, we’re better off than where we are today.
But you look at what Visa’s doing, what Mastercard is doing, they’re very focused on sort of making sure that the buyer side of agentic commerce is secure and has integrity and all the things they’re known for.
Yvette Bohanan: Yeah, so paying attention to the seller is actually the key here, in our opinion.
Russ Jones: Well, the full vision, if you’re an advocate, if you’re getting ready to bet your career on this, you’re betting on agentic buying agents working while you sleep with agentic selling agents. And that’s a two-sided challenge.
Yvette Bohanan: Yeah. Always.
Russ Jones: We’ve been doing a lot of work in this area, sort of teasing apart the landscape, what the differences are between all of these different competing protocols. Who are the people to keep your eyes on to see how the landscape is shifting.
And there’s sort of a new category of, I don’t know if it’s a stakeholder, but there’s a new player on the field that needs, not a company, so to speak, you couldn’t invest in this. But an important part of this is the standards committees who are trying to standardize how product data is described and how it’s presented in a very structured way so that it can be agentically compared across offerings and whatnot.
That’s something that’s always been very merchant specific. How a merchant manages their product catalog is sort of up to them and how they present it on their website is really up to them. And agentic commerce is going to require a lot more standardization of product catalogs and the way they’re presented to agentic agents.
Yvette Bohanan: In a way, yeah.
Russ Jones: So anyway, it’s a great topic. Once again, a moving target. And we’re going to be diving into this at our Advanced Payments workshop in December in San Francisco.
Yvette Bohanan: Yeah, I’m looking forward to it. I think people are going to be super interested in all things agentic these days. So, Russ, thank you so much for taking the time. Always fun to talk with you. Hope everyone who’s listening out there has enjoyed this conversation.
If you’re interested to hear more and learn more come to the Advanced Payments workshop. We’re going to be in San Francisco in December, and we will be talking about this, we’ll be talking about stablecoins, we’ll be talking about all kinds of stuff. And if you think that’s a little bit advanced for where you’re at, show up for the boot camp. There’s two days before that where we can get you up to speed pretty quickly on so that you’d be well versed walking into an Advanced workshop. Understand fundamentals so that we can really dig into the advanced stuff that’s coming out now.
Russ Jones: Yeah.
Yvette Bohanan: Alright. Until next time, keep up the good work. Bye for now.
