The Undecided Future of Agentic Commerce

Sure. So, Signifyd, we're an e-commerce fraud protection platform that focuses on stopping bad actors, preventing fraud, and also optimising commerce for our merchants. And I sit on the side where we're looking at agentic commerce and payment optimisation, and how those two fit together in this new world where we're not talking about people making transactions, but potentially machines, or bots, or whatever you might like to call them. So I'm trying to figure out where we sit in this new agentic world, and how we can optimise things and protect our merchants at the same time.

It's definitely going to change, but I'm not really sure we've figured out quite what that is yet. Over the last two days there's been lots talked about, consumer trust, how this is going to affect regulation, particularly in the EEA and UK, but I've heard lots of questions and not a lot of answers. And for me, that's quite striking. So I think what we'll see is lots of change on the front end, consumers, user experience. There'll be this kind of middle change where things will have to meld around that, maybe from regulation to some process. But fundamentally, some of the stuff when we think about payments, the technology, the rails, a lot of that's probably going to persist. And I think it's how we bring those three layers together that's going to be the challenge in the coming six to eighteen months.

So, I'm hearing trust. Trust resonates. We're kind of in the trust business when we think about how we protect our merchants and make sure their interests are protected, from a Signifyd perspective. And so trust is important, because their consumers have to trust them, they have to trust us, and trust makes things work. Visa say their brand is trust, right? That's important. I've also heard a lot about transparency, auditability, all of these things that are inherent in what we do in a payments business right now anyway; it's just that they're going to change. So is it hype? Is it horror? I don't really know, and I don't think we're ready to find out just yet. We're at a point in time where people are going to start experimenting, start thinking about what they do. But I don't think we're at the edge of the cliff where we have to dive off just yet.

I'd be quite happy to. We know PSD3 is going to hit us very soon. We know there's not a lot in there that's going to help us in this agentic world. But I'm not panicking just yet, I think we have a little bit of time before we have to figure this out. It's significant: we talk a lot about identity and intent. When it comes to securing a transaction, it's about who's doing it and why they're doing it. Intent was talked about a lot on the Money Pot stage yesterday, the intent behind a transaction, understanding the consumer's intent. And I think that's something that's been a continuous thread through fraud and payments for the longest time. Now we're just at a point where the identity is the tricky part, and I think that's the bit we've got to figure out. There's an arguable case that identity is the new money, or identity is the bit that will pin this together. We've seen that most recently with the work the FIDO Alliance have been doing, and thinking about how they partner, Google are right in there, Visa are in there, everyone is thinking about the FIDO Alliance as the key to this identity piece. They do passkeys, and passkeys are proliferating right now across all types of identity, not just in payments. So between some of that stuff, the proliferation of wallets, network tokens, there's a lot that's going to happen in existing infrastructure. Some of it is already there; we're just not using it in the way we probably could for this.

Fraud's always going to be there, right? Because it's a business. There are two businesses competing head to head: you've got fraudsters, who are organisations banking on people making payments, and there are people like us at Signifyd on the other side, competing against them. Two businesses, head to head. But this is just another technology change, like we've had before. When we think about it, we're now in a position where, if we think bots kind of control the internet, Google indexing is a bot, the question is, are these bots friend or foe? That's what we've got to figure out. And some of that is something we can tell from the absence of data in a transaction, or maybe from the presence of data in a transaction. So for us, it's thinking about how we can identify whether we're dealing with a human or a bot. But at the end of the day, the problem is the same, the arena is the same, we've just got to be more clear on identification, and on understanding the intent behind that identity.

I think we're going to sit somewhere in the middle. If I think about how I feel when I make purchases, it's very different. I buy my groceries, my toothpaste, my deodorant, I know what it is, I know how much it costs, and do I really care about anything apart from getting it at the right price? So there's a certain band of transactions that will be perfect in an agentic world. There are others that will never go fully autonomous. Booking.com talked about it yesterday: the emotional investment we have in a holiday transaction, a honeymoon, do I really want to hand that over to an agent? No. They may help me figure it out, that's true. But a pair of shoes, a wedding dress, the clothing for a party, those kinds of things have an emotional, human element that I don't think you can remove from an agent. It's the discovery piece, making a human's life easier, where agentic will take off and see a lot of growth. So, probably business-to-business transactions, probably FMCG and those consumer goods we don't really care so much about, but there will always be a portion of human somewhere in a transaction, because we have feelings at the end of the day.

Yeah. But if you look at it, Amazon tried this a few years back with their Dash Buttons: "here's a button for when you want to order this." It didn't work; it's gone. That whole autonomous-shopping promise didn't really happen. Now, my printer ink gets ordered and delivered to me when it runs out. Perfect. That's agentic, without it really being "agentic." So I think we'll see it blend into our lives over time, and some of that will just be a technology change in general. But in three to five years, yeah, we're going to have a good portion of it. It'll become normal, but fraud is still going to be there.

The interesting thing is: if you look at the names around this room, and you think, how can there be so many people in the same business, making money at the same time? What does that tell you? There's a lot of interest, and payments aren't going away. For me, when we think specifically from a European perspective, the Wero team were on the Orbital stage talking about pan-European payments, bank-to-bank and account-to-account. I think it's likely we'll see an increase in account-to-account, or direct-to-bank, payments, particularly with agentic, as it becomes something normal in our lives. Now, Visa and Mastercard have a lot to say about that, and about how they'll fight against it to keep the coin in their pocket. So the interesting thing will be how that shift plays out, particularly in a regulated area such as the EEA and UK, where we've got the UK payments forward plan, and everything around trying to give consumers options around payment that maybe aren't Visa and Mastercard. So I think that's a very interesting space to watch. And we know that regulation drives innovation, we did chip and PIN first, we did contactless first. So I'm a firm believer that innovation, particularly in the payments arena, will come out of Europe and go globally. So we've got to see how that's going to go. But for me, the big piece is the alternative to the card rails for payments.

It is. But the interesting part for me is that if you look at each of those individually, where do they all come down to? It's this issue of who owns the liability. Agentic commerce adoption may go to die on the hill of who has liability. Bank-to-bank payments may also die on that hill. I think, from a consumer perspective, where the card networks win is that they have an existing model where consumers trust and know that if something goes wrong, they can talk to Visa or Mastercard and sort their problem out. We need to find the same kind of movement when it comes to agentic, to make sure that persists, and the same for any account-to-account or bank-to-bank payments. Because, not to be too cliché, consumer trust is the thing that will make adoption happen. End of. And as soon as we can figure that out, along with the technology, which will sort itself out, the other piece has to happen.

So this is my second Money20/20.

It's to plan ahead, look at what you're coming for. For me, it's a really good learning experience, but I've had to plan my time: between visiting partners, talking to prospects, and listening and absorbing that information. So plan ahead, and plan that you'll have three days wiped out. Don't plan anything else, give up all your other meetings, and just come, involve yourself, absorb it all, and dive in.

Thanks, yeah.