The Future of Selling: Navigating Agentic Commerce with Jason Goldberg
Today, we’re diving into the world of agentic commerce with Jason Goldberg, who’s been riding the digital commerce wave since way before Amazon was even a blip on the radar. The big takeaway from our chat? Agentic commerce is rolling in faster than any previous disruption we’ve seen, and the tools we’ve got to navigate this shift are still figuring themselves out. But here’s the kicker: even if they’re not perfect, we’ve got to start using them to keep up. Jason breaks down why it’s crucial to lean into these changes now instead of waiting for everything to be fully polished. We also explore how this evolution is reshaping not just the retail landscape but the very way we think about buying and selling. So grab your headphones, and let’s get into it!
Takeaways:
- Agentic commerce is moving faster than previous disruptions, so businesses need to adapt quickly.
- Understanding the tools for agentic commerce is crucial, even if they aren't perfect yet.
- Companies risk falling behind if they wait for perfect conditions to adopt new technologies.
- The importance of leadership in guiding organizations through the transition to new commerce models is paramount.
Transcript
Hi, this is Adam Davidson, one of the co founders of FeedForward and your host of the FeedForward podcast. If you sell anything, and most of our members organizations do, this episode is for you.
Today I'm talking with Jason Goldberg, who goes by Retail Geek for very good reason.
He's the Chief Commerce Strategy Officer at Publicis and he's been at the frontier of digital commerce since, and I mean this literally before Amazon existed.
His big argument is, is that agenic commerce is arriving way faster than any prior disruption and that the tools we currently have to navigate it mostly don't work yet, and yet you should use them anyway. The reason why is the most interesting part. Here's Jason. All right, well, I am thrilled to welcome Jason Goldberg.
I'm realizing, Jason, I don't know your exact title.
Jason Goldberg:It's 87 words, so you should have studied for days.
Adam Davidson:Yeah, I should have studied.
Jason Goldberg:Chief Commerce Strategy Officer.
Adam Davidson:Great. At Publicis.
Jason Goldberg:Exactly.
Adam Davidson:Okay, great. So we're going to dig into agentic commerce. You've actually been playing around with digital commerce kind of before I knew, before the Internet.
Before the Internet, before I knew that was a thing. So we're going to get into AI and agentic commerce.
But I thought it'd be fun to talk about like we've been through a whole bunch of radical transformations and how people buy stuff because of zeros and ones. So maybe. Yeah, just walk us through how you first got into this.
Jason Goldberg:Yeah, well, I'm a fourth generation retailer. I grew up in a retail family.
I had several uncles that started an entrepreneurial kind of retail that was not very well thought of or lucrative at the time called video rental. And they, they started these video rental stores, they sold them to an entrepreneur and I was included in that transaction.
And so that the, that entrepreneur was a, an amazing gentleman named Wayne Heising. And the, the business he built by rolling up these video stores was called Blockbuster Entertainment.
And so I was inadvertently part of the, the founding team of Blockbuster. And you know, we of course were disrupted by digital quite a while ago, but we tried to do it ourselves.
We launched e Commerce in: Adam Davidson: ries. There's like memos from: ry. What were you thinking in:So that was still like going to the library and using something like what were you imagining?
Jason Goldberg: mber, no one was qualified in:Like, and, and so like the questions back then were, should we have a website? Right. And that was a totally valid question because it was really before there was a browser.
Like the browser at the time was like Spyglass, which was a text based browser at the time, Mozilla and Netscape and things hadn't even happened yet. And so, you know, you'd go, oh man, I can see the future. In the future, people are going to watch all these videos online.
They're not going to, you know, go get all these atoms at the store and drag the atoms home to watch content and we should be there early. Right? And so then you kind of kick off a project to solve that problem and some very helpful person in the room. And this is a true story in my case.
Hey, Jason, this new venture you're launching, just so we're clear, the total addressable market for everything you're trying to do is smaller than one of our stores revenue on one weekend because you're talking.
Adam Davidson:About like a few hundred geeks around the world who even could understand how to transfer. And at the time it's like with luck it would be like a 30 by 30 pixels, like micro. Yeah.
Jason Goldberg:And so, so again, like is, so is it going to be huge in year one?
Adam Davidson:No.
Jason Goldberg:Is it going to be this transformation that, that you know, we, we have the potential to miss and has the potential to disrupt us and that we need to be leaning in now to get good at this for when it does become more scaled. These are all good questions, right? And I would argue, like in the case of digital, obviously, you know, we were right. It was quite transformative.
But there's, you know, a lot of smart people had meetings and said we ought to be all in on the metaverse, right? And you know, who knows, maybe the metaverse will be a thing one day.
hype cycle and in, you know,: Adam Davidson:I feel like I've predicted one thing accurately in my life and that was how Podcasting would impact terrestrial radio for public radio. And I think I was broadly right, like maybe 70, 80% right.
Jason Goldberg:Yeah, I'm kind of sad that you were right, but yeah, yeah, yeah, yeah.
Adam Davidson: This was like: Jason Goldberg:I would argue it is the predictions are easy, that getting the timing right is, is quite difficult.
And again I, I hesitate to make fun of the predictions that are bad because the, the longer horizon of time may ultimately show that many of those things we make fun of today's bad predictions do come to pass. Right. Like it's a, you know, manner that the, the timing is, is very tricky. And that to me is the most interesting thing because.
So if you think about the kind of 30 year trajectory of my career, I worked for retailers when E commerce disrupted retail and then I worked for e commerce sites when mobile disrupted E commerce and then I worked for commerce companies when social disrupted commerce. And there were new challenges in each of those disruptions for sure. But there were, there were familiarities, there were things that rhymed. Right.
That were kind of common to me. The naysayers in those meetings I got quite familiar with throughout the iteration of those three disruptions.
And now we're going through the same thing again with agentic commerce. What's most different to me about agentic from the other ones is the pace of change. Right.
The, you know, I would argue most of us had 15 years to figure out e commerce and we had eight years to figure out mobile commerce. And we're, you know, we're still figuring out social commerce but you know, we've had a reasonable horizon.
It feels like the, the evolution of, of agent E commerce is happening at a completely different scale that breaks most of our old processes for dealing with disruption.
Adam Davidson:Yeah.
And that I really want to underline because I think a huge risk right now would be, you know, I'd say broadly the feed forward point of view is you don't want to be the laggard, you don't want to be the one who says someone else will figure it out and then we can just come in and clean up at that time.
Jason Goldberg:Yeah, I, I, I often, you know, joke about not needing to be a first mover and everything that you know, the fast followers tend to have better outcomes.
But the definition of fast follower in a world where, you know, it takes four days to get to a billion users instead of, you know, 10 years to get to a billion years looks and feels quite different.
Adam Davidson:Yeah, right, exactly. Also, by the way, like Blockbuster or for that matter Kodak or Xerox, the choice wasn't go all in or abandon it.
The other option was let's keep an eye on this. Let's devote some resources, let's have some kind of gate monitored system that sends a signal. There was a different choice.
Jason Goldberg:And in fairness to my colleagues, I won't defend myself, but in fairness to my colleagues, that, that actually was the play and it actually did work. Right. So the, you know, nobody was like, oh, let's close all the stores and figure out digital because that's, that's what's going to happen.
It's, oh man, this is going to be something we're going to have to learn how to do and get better at over time. And way before the Internet goes to 200 million households, it's going to be viable to go to 4,000 Blockbuster stores. Right.
So maybe we ought to be figuring out how to do digital distribution not to the home, but to our stores.
And so we, we went out and we, we tricked IBM into investing a bunch of money and we, we build the digital distribution network for, for stores and we started manufacturing products in stores instead of moving them there on trucks. And that turned out to be pretty lucrative. And we, you know, got, got some digital muscles and did all these things.
And then in fairness, you know, to my colleagues, we then sold the company to seller Redstone, who owned Viacom for, for the modern equivalent of $15 billion. And you know, everyone but me got to retire. So it actually was a success story.
Historians will always make jokes about how, you know, Blockbuster missed the, the digital revolution and didn't see Netflix coming and all these, these things.
Like the, the truth is all the smart people that founded Blockbuster, comma, and me like did, did see it coming and actually, you know, monetized it, executed it and exited. So the, you know, my joke is the railroads are not a very good business today. But the, you know, Rockefellers did, did pretty well.
Adam Davidson:Did pretty well. Yeah, fair enough. Yeah, that actually, that's a really good perspective.
And by the way, I, one of my favorite recent business books was Unscripted about Sumner Redstone. I actually wrote the review of it.
Jason Goldberg:And I got to meet him a number of times in that transaction. Transaction, yeah.
Adam Davidson:And let's just say his specialty was not subtle investigation of transformation. His specialty was just being a brutal son of a.
Jason Goldberg:And yeah, it is interesting the skills that help, the core competencies that help People succeed and they're not always the ones that I admire.
Adam Davidson:Yes, right. But it wasn't like he's not the one you would expect. I mean, he was anyway, so. All right. So I would actually.
I kind of want to go through all of those beats of your story, but maybe we'll have to put that on hold.
Jason Goldberg:That'll be the episodic spinoff of this.
Adam Davidson: ,: Jason Goldberg:But we made it to the halfway mark. Congratulations.
Adam Davidson:We made it to the halfway mark. As I understand it, your job is to talk to clients who are a lot like the Feed Forward member clients.
These are large global companies who really, really, really want to figure out how to sell to folks in whatever way they sell to. And so I want to kind of get the. Where are we right now? What's going on?
So the thing that comes to my mind right away is I now, I mean, just this morning I had to buy something. I went to ChatGPT and to Claude and I said, hey, I want to buy this thing. Which one should I buy? And they largely agreed. I will say.
Ethan Mollick and Leonard Meinke from the Fade Forward team just did an academic paper that showed that these are extremely variable and not entirely reliable. This agentic product search. I know that, but I still use it, even though I know that.
And Leonard, who's like the scholar who proved this still uses it himself because it's just very tempting rather than spend two hours googling stuff to just say, hey, what should I buy? So where are we right now?
Jason Goldberg:Yeah, yeah. A couple things to unpack there. The.
So we are in the very early days, like I would, I would still argue we're in batting practice of this baseball game. The game hasn't really even started yet. Right. From a. From a overall perspective. But obviously, you know, if.
If we strain the baseball metaphor to it to its absolute limit that, you know, you still had to do a lot of work to become a major league pitcher and get. Get to the stadium and, you know, get a uniform on and get invited on the field to even have batting practice. Right. So the.
It's not to say there isn't a lot of interesting things happening that a lot's going on, but I have a feeling when the final history is written in on a jantic commerce, it's. I don't know what it's going to look like. It's probably not going to look a lot like we think it looks right now. Right.
And the things that are happening right now are super interesting and I spend my, my whole life following them and tracking them. They're not economically material. Right.
So one of the very first conversations I always have to have with, with my clients which you know, you know, very much overlap your, your membership the, is that hey, these conversations are super important for our future and it's a future that's coming faster than we've ever had the future come at us before. So, so well, well worth having these conversations. Well worth building some skills.
about making your numbers in: 's impacting their revenue in: Adam Davidson:Again, okay, that's really helpful because I, I mean maybe I can tell myself I'm a trend setter frontier person which I don't feel very often in my mid-50s because I use it all the time. But you're saying if we look at the overall market, I have a 14 year old son, so I know TikTok shop is the main.
Jason Goldberg:And that being said, you brought up a great point. You know, it's not completely reliable and you still use it every day.
So again, in my historical precedent, like the most important attribute in digital commerce, the one attribute more than anything else that will drive sales of your product, it probably won't come as a big, big shock to any listeners.
Ratings and reviews and so when I get a lot of ratings and reviews on a product detail page, that product sells quite a bit better than when I don't have ratings and reviews. When I survey users or when I talk to users, all users know that ratings and reviews are gamed and fraudulent.
Adam Davidson:Right?
Jason Goldberg:They, they all know they're wrong and yet they're still the most influential. Right?
Adam Davidson:It is crazy. Yeah.
Jason Goldberg:And, and even like point out that paradox to someone, you know what they say, they're like, oh yeah, I, I know they're wrong, but I can tell the good ones from the bad.
Adam Davidson:And how many of those people actually read all the reviews. And I mean you look at the.
Jason Goldberg:4.2, Which is of course statistically not possible. The. And so yeah, that in that same way the these robots answers to commerce questions are based on the available data about the commerce products.
The product feed, we often call it the attributes. And our databases of attributes are still wildly inaccurate and wrong. Right.
And you think about it, when we used to sell stuff through stores, the store needed to know like eight pieces of information about a product, right? So you know, how many cookies are in the bag, how much does the bag weigh, you know, is the bag edible and what price should we sell it for? Right.
So then we launched E Commerce and people are like, oh, you know what? We need a bunch of things you didn't used to give the store.
We need a picture of the cookies and you know, that ingredients list, that's a picture on the package. Well, we need that as text and it would be, you know, and the nutrition information. It would be great to get a table of all of that.
And now customers are asking, is it gluten free? Right.
So, so kind of the world had to go from sending eight attributes to a retailer to sell something to sending 30 attributes to a retailer to sell something. Today, if you try to sell something on Amazon and you want to completely fill out the record, they're probably going to ask you for 300 attributes.
The robot would like 3,000 attributes. And I suspect by the time we're done, it'll be 30,000 attributes. Right.
And so you know, the shirt that, you know the, that the traditional retailer wanted to know, is it a large or an xl?
The robot wants to know how many millimeters the top button is from the neckline and, and how big the sleeve opening is and you know, does it come pre shrunk and all of these, these nuances that would help it give you a better answer that literally the manufacturer of the product likely doesn't have. And for sure if they have, they haven't published publicly, they haven't made available in a, in a uniform format.
You know, we don't even have have like controlled nomenclature for things like gender, right. So you know, the gap might say men's and women's and, and Abercrombie and Fitch might call it male and female or whatever.
And so when the robot says, oh, I'm looking for, for a shirt for a man, you know, first problem, right.
Adam Davidson:Like, and then if we're getting at like somehow it knows, even if I don't, that I like bigger buttons or smaller buttons based on my previous purchases or something, there's no as Jerry Seinfeld.
Jason Goldberg:Taught us in one of the episodes, the whole secret to the shirt is, is how close that button is to the. How much. How much chest you're showing when you button that top.
Adam Davidson:Yeah, right. Least possible is my. My vote, but I think that's most.
Jason Goldberg:People's vote for you and I. But for, for other people, it might be a different answer.
Adam Davidson:Yeah. Yeah. A few things are going through my mind.
So one is a model of the world based on today would tell me it's going to be a lot like Google or Amazon just with agents. So, like, I will go to a search engine. Instead of going to a search engine, I'm going to go to cloud or chat or I already do.
And I'm going to say, hey, I want a new camera. Here's what I'm interested in. Tell me about it. I actually did that very thing yesterday.
's the difference between the: And it said, basically, the:And so my channel.
Jason Goldberg:And you're smarter. Google.
Adam Davidson:Yeah. And my challenge, if I'm Canon or Nikon or whatever, is how do I get that agent to say the Canon camera is the best one?
Or how do I convince that agent to upsell or whatever. But you're saying to me that's, that's like probably question, but that's not tomorrow's question.
Jason Goldberg:Yeah. Or, or tomorrow's question is going to be that that'll be a tiny subset of tomorrow's. That that may still be valid, but it'll be a tiny subset.
And to be clear, I talk with clients every day about that exact tactic. Right.
And again, the world can't even agree on what to call it, but we'll call it either Generative Engine Optimization, Geo or Answer Engine Optimization. Aeo. Right, right.
Adam Davidson:Which is like, people want to think of it as SEO, but for agents.
Jason Goldberg:Search Engine Optimization for agents. Right. And so a couple problem. So AI encourage everyone to do that. Right.
And I say, I do want to throw one small caveat at while I'm Asking you to spend money and implement an SEO pro AEO program or geo program at your organization.
You should probably know up front that the tools I'm going to recommend you buy to tell you how you're doing are wildly inaccurate and probably wrong and everything you learn will be invalid tomorrow and probably won't work. Right.
Adam Davidson:So wait, wait, wait. Let's just stop for one second because this is a big moment.
Yeah, you're saying yes, spend money on this, spend time and effort on this and it doesn't work at all and we'll waste your money.
Jason Goldberg:Yeah, I actually don't think of it as a waste of money. I think of it as an investment. But so two problems.
First of all, so if you're a investor and you want to invest in a jantic commerce right now, the companies that you found are the companies that tell all of your members how often they are or are not showing up in the robots prompts and what citations the robots prompts are giving for the answers they are giving so that you can try to show up in those citations more often. Right. So you know, you may there the.
Adam Davidson:Robots might be reading the wire cutter and rating that higher or Consumer Reports and then you know, there's also a ton of slop out there. They may or may not be reading listicles and whatever you want to know.
Jason Goldberg:What are the leverage points?
So in the SEO era you bought a tool that said here's how you're in the early days, Google told you how you were doing and then Google goes, oh yeah, actually this is not helpful to advertising. We're going to stop telling people. And so a bunch of third party tools were born.
Semrush a rep like there's some geek in your, in every one of your members organizations with SEO in their title that has an expensive license to one of these tools and they're finding out what keywords people are typing and how we do in those keywords and when they make changes to their content, how, what impact that has on how they rank. Right. And that's.
Adam Davidson:And that did work, right? I mean, yes, SEO works and I think of it as kind of it was a bit of a war between SEO tricks and Google. Like Google didn't want them to start.
Jason Goldberg:Out very black hat. There were a lot of bad things you could do and, and over time Google has been able to, you know, knock.
I'm not saying there aren't still some black hat things but knocked down the vast majority of that. And so you know, now people mostly behave in the way Google wants them to behave.
And you know, good SEO versus bad SEO can materially impact how you show up in, in Google. And so that's a good investment and the tools to do it are quite reliable. Right.
So when we, we go, oh, now people are going to a robot instead of Google, let's do this all over again with the robots. Right. And so the first tranche of investments is venture capitalists invested in the, the tools that measure how you show up in the prompts. Right.
Well, so first of all, couple, couple problems with that. Like the, those prompts are obfuscated from us way more than than Google searches are.
Like Google publishes a lot of data about what the most popular keywords are and the headword keywords, people type the same ones quite a bit. There's a long tail. But the, the keywords that are materially relevant like are, are very consistent and knowable.
By, by the way, the average number of words people type into Google is 3.8 words. The average number of words that people type into a a robot is, the data is a little more foggy, but it's something like 40 words. Right?
Adam Davidson:Right.
Jason Goldberg:So there's a lot more context. So Google quite intentionally does not personalize those results. So the results you get are very similar to the results anyone else gets.
And that's for a very tangible reason. They want to sell an ad and that ad wants to reach a lot of people.
If everyone got different results, then the ad would only be relevant to a few of those people. Right. So the one size fits all advertising model drives Google to have very consistent results.
Adam Davidson:So if I type in best low.
Jason Goldberg:Cost camera for $2,000, camera for 2,000.
Adam Davidson: d you say I want a camera for: Jason Goldberg:Use the we're going to get similar results. There'll be some nuanced differences based on, on our proximity to a Best Buy store and some things.
But like for the most part the they'll be the same. The robot is going to give a wildly different result every single time.
And there the results are less aggregated because again 40 words of context instead of 3.8 words of context. And like if you go into a private mode and do a search on Google, the results will look a lot like the public mode.
Like it's not like it learned a lot from what you did in the past. The robot is using the context from the last 4,000 prompts you typed to the robot, what we call the memory.
So for all of those reasons, the result set from the robot are wildly More the standard deviation is far greater than Google.
And once you figure out where you show up in Google and how to fix it and you win Google, you're the winner for the next six months until Google changes the algorithm. The robot changes the algorithm every minute. Right.
And so if you, for example, figure out a nefarious way to, to disproportionately win the robot and you invest a bunch of money in executing that strategy, the robot's going to see that strategy and change the algorithm. Right.
Adam Davidson:So you've got like 30 seconds to exploit it.
Jason Goldberg:Yeah, so the academics would call it. It's, it's. Google is deterministic. AEO is probabilistic.
Adam Davidson:And so, yeah, and a simple way to see this that I've shown some people is like, search, you know, best camera. Claude, GPT. It doesn't matter. Gemini, say like, I want a camera for 2,000 bucks. Then in a fresh session, say, I hate Mondays.
I want a camera for:But you could type anything like aardvarks are cute or whatever you're going to have. It's not just that it's going to recommend different cameras. It's like the whole structure of the response.
Like, in one case it might be a table, in another case it's a narrow.
Jason Goldberg:Have to be quite as, as clever a detective as you're proposing can actually go to the robot and go, what do you know about me that would be relevant to buying a camera?
Adam Davidson:Wow.
Jason Goldberg:And, and it, it will, you know, tell you a bunch like, of shocking facts that you probably didn't realize you had disclosed to the robot that the robot is using.
Adam Davidson:Yeah. Wow, that's. That's really cool. So, okay, so you're convincing me that.
Jason Goldberg:There was one other. Yeah, so, so the results change all the time.
So unlike SEO, where you could kind of win and then you got a durable advantage from winning when you win it at Geo, you won for one consumer for one search, and the rules completely change for the next consumer and the next prompt. That's half the problem. The other half the problem is the tools are lying to us. Right?
Like, so, so of the thousand tools you can buy for, for AEO, about 997 of them. Here's how they work. They, they type one prompt into one robot and get a result and say, this is the result from that prompt.
But as you and I just discussed, if you ran that prompt 100,000 times, you'd get a Hundred thousand different results. So the, the enterprise tools that no one wants to pay for, when you type a prompt in, run it a hundred thousand times.
And so then they tell you on average, this is how you land. And then you go, ooh, okay, I don't like where I landed. Why did you propose my competitor more than me?
And then, you know, you, you get into this thing called citations. And citations are saying what the robot quoted when it gave the answer. Citations is not the same as sourcing.
Like citation is something the robot read while they were giving the answer. It does not mean it was the thing that decided what it gave is the first answer. So that's problem number one.
We conflate the fact that the robot used you as a citation with the fact that you were the persuasive citation. Right?
Adam Davidson:Right. Because all it's doing is, oh, he wants a citation. I'll give him something that's called a citation.
Jason Goldberg:Even if it's not the actual is anecdotal. It's not comprehensive.
Like, there could be many, many factors that you, we can implicitly see the, the robot use data from this source because that's, that data is unique and that source is not one of the citations. Right? So, so the citations are, are not super reliable.
And we don't know whether the, the, the robot took a positive or negative inference from a particular citation. The citations aren't comprehensive.
And so we all today, we, we disproportionately react to the, the results the, the tool gives us, even though the results probably aren't. And we, we disproportionately chase the citations even though the citations aren't as influential as we'd like to believe.
And when, when we use these tools, the tools last week said Reddit was super important. This week, Wikipedia is super important also, access is super important. You know what these tools don't cite very much in product content is Amazon.
And you go, huh, that's interesting. Amazon seems like the, the, the most authoritative source about product data, and they are. Well, Amazon blocks all the bots.
Amazon doesn't let the robots read them so they don't show up.
So you would, you know, you can still find unique data that has leaked into the robots from Amazon, but the robots will never cite Amazon as a source, for example.
Adam Davidson:So you've done a very good job of convincing me that AEO and GEO is a scam. Is a scam. And you're also though you have not convinced me that you should do it anyway, so here's why.
Jason Goldberg:Because again, what we talked about is we have to be at the minimum, prepared to be a fast forward in a world that's changing way faster than ever. So we have to develop new skills, we have to develop new vernacular, we have to develop new processes.
And the process of adopting a EO helps us develop all those things. And what I, what I'm absolutely certain of from my, my life experience history is the hardest thing with any of these disruptions.
It's not figuring out what to do, it's not learning the new tech, it's not inventing the new solutions to the problems. It's getting all the people that were selected for their skills in the old model to adopt the new model.
It's the, it's the human software, it's organizational change management. Right? And so more than anything else, we need people getting used to doing their jobs differently.
And when you launch a AEO initiative at a major CPG brand, suddenly the whole board of directors getting a dashboard about how the robots show up. And that's prompting a bunch of great conversations in that board meeting. Ooh, should we be investing more? Should we be caring more?
Should we be joining the, the consortium of that's developing standards for the robot?
Should we be weighing in on this legislation and these lobby, all of these good muscle memories that help us prepare to be a fast forward get developed by doing projects like aeo? So my thing is we ought to be doing this. It's a training exercise. It's not necessarily on its own gonna yield a super lucrative outcome.
Adam Davidson:If there was something that worked, you'd say do that. There presumably will be something that works. And so we have to be in the arena.
A joke I often make in our meetings whenever we have a member meeting and the message is, we don't know, nobody knows, is I always say to the members, I am willing to provide you absolutely certain, 100% definitive and completely wrong answers for $1 million. And you just describe my job.
But okay, yeah, no, that's, that's why I'm raising it because without getting into any particular names, there might be others in that market and certainty is very attractive. I guess I, I'll just ask this, like why, why don't you just sell the publicist AEO Geo 100% solution?
Like what, why is it in your interest to say, say what you're saying?
Jason Goldberg:Yeah, because obviously like, you know, we're, we're in the long term relationship business. Like if I, you know, if, if I didn't think I was going to be around in five years and I just wanted to monetize the, the fear, uncertainty and doubt.
Right now I would just run around selling AEO and say this is, you know, you should absolutely take all your super bowl money and put it in a EO because the super bowl is dead and AEO is the only thing that matters in the future. I know that's unlikely to be true. First of all, it's going to be an end.
I would actually argue if we start gaming this out top of funnel marketing and brand building, being a loved brand is going to be more important in the future than less. I would actually argue all this digital so far has pushed us to the bottom of funnel and commoditized things.
You care less about the brand than you're used to. This started going to get really complicated. But the brands were originally invented as, as trust symbols for safety.
The one of the very first brands was Gold Medal Flower. They developed that brand because they won a gold medal. Right. And sunflower was fake and cut with a bunch of chalk and would make you sick.
And the gold medal flower was, was verified by a third party to be quality. And so if you bought flour from the barrel that had a gold medal mark on it, you knew it was safer. Right. So brand was very important to safety.
Today we have objective information about how safe every product is. Right. Like so the, the brand name on it matters less. We know all the TVs in the store are made but at the same factory in China.
And so you know, we care a lot more about what wirecutter says is that is the best value in TVs than we do whether it has a Sony or Samsung logo on it. So I would argue that the shift in available data in the digital ecosystem has made brand still important, but less important than it once was.
So now that the robots are answering that question for us, here's an interesting thing. What prompt do you want to win on the robot? Best Phone for a 10 year old or best iPhone for a 10 year old?
Adam Davidson:Best phone.
Jason Goldberg:If you're Apple, one of those you would way rather everyone type best iPhone for a 10 year old. Oh, right, right.
Adam Davidson:Yes. Yeah, right.
Jason Goldberg:And so it's now more important than ever that people don't think oh I just need a phone, it's more important than ever that phone be a loved brand.
And before I even asked the robot for its opinion, I want to make it clear I'm not going to embarrass my 10 year old by sending him to school with an Android phone. And he'll be Forever scarred. Right.
So, so, you know, brands are going to, I think in the long run, invest more in being that loved brand that, to avoid the commoditization. Because I do think, you know, one very obvious thing here is some products we're going to want the robot to help us.
We'll call those considered purchases. We want the robot to do it with me. You, you didn't just say to the robot, pick the best camera and send it to me.
e difference between thousand:And I think that's likely going to be true for a long time. It's certainly going to be true for your house. It's certainly going to, you know, be true for your car.
But if I said to you, you don't ever have to worry about running out of toilet paper again, you would probably accept that, that premise, right? And say, robot, never let me run out of toilet paper again.
And the robot's going to figure out how much toilet paper your household consumes and have toilet paper show up in the right quantity just before you need it. It's way smarter than subscription, by the way, which mostly didn't work. But auto replenishment of these commodities is likely going to be a thing.
And so if your brand, you have to say, wait, do I want to be the loved brand that people specify? Or am I just going to be the commoditized solution to a problem that people are going to outsource to the robot? Right.
And so you know that a bifurcation of those two things, auto replenishment versus considered purchases, is probably likely one of the ramifications of this new agentic commerce era.
Adam Davidson:By the way, I was distracted for a second there because I did type in to Claude, what do you know about me that would affect a camera camera?
And it first listed all the chats I've had about buying cameras and pointed out that sometimes I say price is no object and sometimes I say price is a huge object.
Jason Goldberg:You're inconsistent.
Adam Davidson:Yeah, yeah. But then I said, what about things that aren't directly related to cameras?
And it actually had, it was really interesting is that it's saying I shouldn't buy the $2,500 Canon I was thinking of buying. It says, you run a novelty. You drop things when they're tedious, you want to get outdoors more, you want to explore More of Vermont.
You want to become a more present parent. And essentially I'm going through all this.
You don't want to spend a lot of time learning the complex menus of some top of the line device made for professionals. And it recommended a camera I wasn't considering at all as a much easier but fun and really high quality camera. So that's kind of cool. I like that.
That is what happened by the way.
Jason Goldberg:The absolute correct answer is get the $5,000 camera to make your, your brain feel good and then buy the, the, the easy to use low impact thousand dollar camera to actually use.
Adam Davidson:To actually use.
Jason Goldberg:Right, Right.
Adam Davidson:We've been mostly focused on B2C. Are there B2B? How would you say this to a B2B?
Jason Goldberg:Like yeah, it's bigger and more urgent.
Adam Davidson:Yeah.
Jason Goldberg:So for a bunch of fun like the, the, the consumer E commerce industry in the United States of America is slightly over $1 trillion a year in sales. The B2B E commerce sales in the United States are about 2 1/2 trillion dollars.
Adam Davidson:Really? So I didn't know that. Yeah. So that. Wow, that's counterintuitive to me.
Jason Goldberg:Yeah. So. So B2B is a bigger market. People spend more now often a transaction has both. Right.
A retailer bought something from a manufacturer and that was a B2B transaction. And then the same thing got sold as a retail transaction from, from Walmart to a customer.
But there are a lot of things that businesses use that don't turn into consumer products. Right. So the volume of purchases in, in the B2B world is much larger than the volume of purchases in the B2C world. The B2C world is more sexy.
Everyone wants to go to a dinner party and say like I invented the thing that everyone used. Not the, the invisible thing that made everyone's life better that we don't know about. Right.
And so, so in my industry, the platforms that help you sell the consumers and the tools and the processes have always been had greater voice and been talked about quite a bit more. But the, the B2B solutions are more lucrative. Right.
And when we look way before there was E commerce for consumers to buy stuff digitally, there was a thing called EDI for businesses to do transactions digitally. It was rudimentary and sucked and it's now all been replaced by E commerce. Businesses now use E Commerce, not edi.
But they were absolutely first mover.
And a lot of companies, including some member companies of yours, made a lot of money like facilitating B2B EDI transactions for businesses way before There was E Commerce.
And so I would argue the same thing is happening in agentic like the, that we worry about using the robot to help me pick a camera, but that purchase manager is also using the robot to figure out what toilet paper they should buy for their 27 commercial properties. And they have. That tends to be a more rational, less emotional decision.
And so you're more willing to outsource that without consideration to a robot. So there's a bunch of those things. And then just evidentiary, we look at what the volumes of transactions in China and the.
The AI agent of commerce has infiltrated B2B much faster and at a higher rate than it's infiltrated B2C.
Adam Davidson:Interesting.
All right, I want to dig into that, but because you've mentioned toilet paper and B2B toilet paper, I have to tell a quick anecdote that I, that I experienced that I just find fascinating. I was a business and economics reporter in Iraq for a year in that first year of the US occupation.
And one thing I was interested in, there are some very wealthy families that have, typically they'd have a bunch of businesses, construction, trade, you know. But I had read these articles about how there is a real ceiling on these kind of family led, almost tribal led businesses.
And, and the example that always came to my mind when I thought about this was I was once in the office of the richest man in Iraq, probably worth 150, $200 million, which in the US doesn't make you the richest, but in Iraq that's a lot of money. And I was sitting there waiting to talk to him and there's just dozens of people coming in, coming out, coming in, coming out.
And one of the people came in was asking him which toilet paper to buy for the guest bathroom downstairs. And he took a second and I was like, all right, I now have a rule.
If the CEO is making decisions about toilet paper, then you have a ceiling on your growth. And that might even be true for procurement. Right.
So it's actually, I find it easier to picture the future of B2B agentic commerce than B2C in a lot of ways.
Jason Goldberg:Yeah, this auto replenishment threshold, if you go, oh man, consumer spending US is $7 billion. If what percentage of that is going to be replaced by auto replenishment? You know, maybe it's a trillion dollars, right?
Like maybe it's 1/7, there's, you know, 6/7 that we still want to be involved in. In the B2B space. It's more likely to be like 50, 50 wow.
Adam Davidson:Okay, so, and that's.
Yeah, it's like the little screws or whatever that go into your manufacturing plant, there's the cups for your Starbucks store, like whatever that is.
Jason Goldberg:You actually in B2B already outsource it to a robot. It just that robo. Right, so.
Adam Davidson:Right.
Jason Goldberg:So you run lots of RFPs and you very often use a third party. In my industry, brands hire a third party to run an RFP for creative agencies.
Adam Davidson:Right.
Jason Goldberg:And so that, that third party is just a wildly inefficient version of the robot. Right. That goes out, does all the research, comes back to you and says you should hire pubosis. You don't need to get in all the details.
Just trust us. We did a super comprehensive, thorough investigation. And pubis, this is the best fit for your. Your needs. Right.
Which is objectively true, by the way, for all the listeners. Right.
Adam Davidson:Obviously is the best.
Jason Goldberg:But you outsource that already and felt okay with the answer. Right. So if you believe a robot's going to do a more objective, more thorough job than the agency you used to outsource to.
And by the way, you always had to worry about whether Publicist was giving kickbacks to that agency. We weren't, but. And you, you trust the robot to be less. Less nefarious and persuadable, then.
Yeah, you're just, you know, there's already a precedent that you like to outsource these hard decisions, and so now you've got a new, better place to outsource them to. Potentially.
Adam Davidson:Potentially.
Jason Goldberg:And by the way, is that we could do a whole nother podcast about if you. If the robots are more or less trustworthy and all of that.
Adam Davidson:But yeah, and I'm already seeing attention. That's kind of like the SEO Google tension, which is. So you, as publicists, want. Every time someone says what's the best global.
I mean, presumably someone asking that already knows, but a CMO is like, hey, I hate our ad agency. You know, who else should we go to? You want them to.
Jason Goldberg:Actual answer would be whoever you're not using is the best. Right?
Adam Davidson:That is the best. Yes, exactly.
Jason Goldberg:But.
Adam Davidson:So you want them to say publicis, but every time you do a search, you want them to be completely objective and not at all swayed by the players involved. So, and, and I guess our members are in that same place. They're going to be on both sides.
Jason Goldberg:And that's what I'm saying. Like, I want Publicist to be a love brand. Right? Like, I, like, I don't actually want you to Go ask what agency you, I, you should hire.
I, I want you to say which pubis agency should I hire? Right. Like it's a foregone conclusion. There's all these endemic advantages to working in the publis ecosystem.
So like, of course we're not going to ask the robot to pick a holding company. We're just going to ask the robot whether Saatchi or Leo Burnett is a better creative agency for my particular needs.
That would be the best outcome for us. And in the same way, you know, the best outcome for Coke is you know, you know which flavor of our electrolyte energy drink is going to be the best.
Not, you know, which, which alternative brand is going to be best.
Adam Davidson:Right. I just want to kind of recap because now I want to move to like a, a real, like here are some things you can do.
You can learn, you know, with a huge uncertainty umbrella over everything we say. But where we are now is not a ton you can do to game these systems.
o have been really popular in: Jason Goldberg:The best time to plant a tree was 20 years ago. And the second best time is right now.
Adam Davidson:Right? Yeah.
I mean the one thing we can say with some degree of confidence is agentic will be more and that in our early part of the conversation we were thinking of a very human in the loop way, I'm going to a chatbot, I'm typing things in.
Whereas we can assume certainly in B2B but probably in B2C as well, there's going to be a lot of agent to agent conversation that like I had to trigger this camera decision. It's not hard for me to imagine.
I mean, I already use like Hermes agent that two years from now it'll say to me, hey, by the way, I noticed you've been taking a lot of pictures and you last bought a new camera 10 years ago. Maybe you want to try this new one. It's a pretty good deal. So one thing our members should do is waste a ton of money on AEO and geo.
And you didn't say a ton of money.
Jason Goldberg:Ideally with Publicis.
Adam Davidson:Ideally with Publicis, yes. And we should say if anyone wants to talk to Jason Goldberg, just reach out to us. We are not endorsing we just like that.
You seem to be less full of it than others and you seem very clear and feed forward aligned in the way you approach it.
Jason Goldberg:I'm not sure my family would agree with You. But thank you.
Adam Davidson:Okay, good. But, yeah, we are happy to connect you with Jason if you want to talk to him. The.
Jason Goldberg:Yeah, you were about to say not a ton of money.
Adam Davidson:Not a ton of money. Spend some money. There should be some group of people.
If you're the CMO or the CEO or the cfo, you want to have some group of people in your building who you can call and say what is working, what isn't working. You know, something Ethan talks a lot about is you want to be working on things that don't quite work yet. Like, because, you know, like when.
When fable or mythos 6.0 or whatever comes out you. The only way to know how advanced it is is to have stuff you can't yet do.
Jason Goldberg:There's this budget allocation process people think about that we talk about a lot. And then the actual ratio can vary, but we'll call it 99. 1. Right.
Like spend 90% of your budget on stuff that you objectively know yields a return and that you've historically done and you have evidence works. Spend 9% of your budget on stuff that you have evidence works but isn't as broadly adopted yet.
And spend 1% of your budget on stuff that you have absolutely no idea if it's ever going to work or not. Right. The orders of magnitude of your investment based on your evidentiary knowledge should be different.
But you absolutely want to be spending some stuff on new things that you know are harebrained because some of those will turn out to be the most important skills of the future. Right. And so even and probably sounds like.
Adam Davidson:Yeah. And what you said before is it's not just will this work or is the scheme going to work, it's the language.
It's understanding that there are, like, things I have learned. For example, you want to not just know what training data is, you want to really have a. Have a feel for what the training data is.
You want to know what the training process is like, how you know you'll never know with precision. Nobody does about what's inside these models.
But at least you want people in the building who know that stuff doesn't mean every executive needs to know it. Yeah. You want to have a sense of how post training impacts choices.
You want to have a sense of how context, like, and I don't know a better way than trying stuff, having it fail and, and then being like, wait, why did that fail? And. And learning some more. The problem is I keep learning the same lesson over and over again, but that's a separate issue.
You Know, we, we at FeedForward talk about leadership lab and crowd that you need, you know, broadly for AI. You need your leadership broadly aligned on. This is a thing and we got to really take it very seriously and we're using it as a strong thing.
We advocate. You need a lab that is on the frontier, both building actual usable tools today, but also having a sense of what's coming.
And then you need to figure out how to distribute that, how to. And the learnings move around like the crowd, you know, you're kind of line level workers, whatever there might be learnings from them that.
So I think we're quite aligned there. What does that specifically look like? Like the one thing I, I want to push back on is you talked about hiring an AEO geo firm.
Something I worry about with this AI moment because we're in this period of really defining basic terminology. I mean it's like every few weeks there's, you know, it was prompt engineering, then context engineering, then harness engineering.
Now today everything's about loops. I don't know, you know, we weren't hearing about loops, you know, six months ago. Now it's all we hear about.
I don't know, like three years from now, whatever people are saying AI is will be some new thing.
And so I fear getting locked into some system where I'm using proprietary terminology for one company without actually being closer to the bare metal AI tools.
Jason Goldberg:So yeah, I think you're right. I probably spoke imprecisely when I say like hire aeo. Like that could mean give people a new job to be in charge of aeo.
It could mean hire new people that are full time employees. It could mean renting tools. Is what I actually had in my mind when I said that. Right. That your employees use to help inform their aeo. You're right.
Like because we're trying to learn like black box outsourcing. This is, is certainly not the way to do it. Right. Because you get the, the least of any learnings from, from that, that black box outsourcing. Right.
It's about building muscles. Building muscles, hiring a guy that has muscles.
Adam Davidson:Right. Although it could be renting someone who has muscles to get you often.
Jason Goldberg:That's the. Yes. Best way to get the heavy thing in the truck.
Adam Davidson:Yeah. Yeah, great. All right. What other things? I mean it sounds like people are going to need a lot more data about their products and services.
That's in kind of machine readable way. Luckily, machine readable is a broader category. It doesn't mean like super well structured CSV files or whatever or JSON files.
Jason Goldberg:I think those are all things like change organization management, thinking about how we're going to get our organization to behave differently, the data governance and thinking about how much more data we need and, and that we have to explicitly decide what data we're sharing with the frenemy robots versus the robots we own and all of these things. AI governance is a big thing.
Like even just saying like what the policies are, what tools people in my organization have access to, what if any, I don't. All those things are important. But the first thing I always say is it starts with leadership. It starts with the top down.
The companies that are most successful here, the, the CEO gave the company permission and encouraged the company to think different and behave differently than they had before. So my, my favorite example, I, I work in retail. Best retailer in the world is Walmart.
That their, their CEO just annoyingly retired Doug McMillan because he's, he's a little younger than me and he's wildly successful and retired. But, but Doug like, you know, literally said, huh, this new thing is happening. Agentic commerce is going to have a huge impact on Walmart.
We got 1.7 million associates at Walmart, none of which were hired because they know how to use AI, right? So first thing I got to do is get those 1.7 million associates thinking about AI. So what he started doing is he went to every meeting.
I was literally in a meeting with him two and a half years ago. Big deal to be in a meeting with the CEO of Walmart. Everybody prepares, they're super nervous. Doug walks into the room, sits down.
Thank you all for coming. Quick question before we get started. How'd you use AI to prepare for this meeting? And you could feel that air suck out of the room, right?
Because nobody two and a half years ago used AI to prepare for a meeting with Doug McMillan. And they all felt like suddenly like they just failed the CEO.
And so there's this awkward silence and eventually one brave soul chimes in and goes, doug, I'm really, I'm really sorry we didn't use AI to prepare for this meeting. Which is not a surprising answer. But then they dropped the bomb. AI is illegal at Walmart. And everyone's like, what are you talking about?
And he's like, oh, well, we get an email from it every week. AI hallucinates and don't trust the answers and you don't have a license. Don't use this stuff and it leaks our data. Don't use this thing.
We've all been given the fear of God about using AI, Right. And so Doug's like, oh, my God, we got to fix this, right? I need these guys using it.
So he left the meeting and he bought ChatGPT for 1.7 million associates. And then a week later, he said, oh, that's just one tool. We need to buy all the tools. And then he's like, oh, we better build our own tools.
And he started building tools. And while all this is happening, were they. Were they instantly driving new economic value for Walmart?
Probably not, but they were giving the public stock a halo and Walmart's stock past a trillion dollar market cap for the first time ever, largely based on this AI halo of all this AI change that Doug was driving. And to me, the most telling thing he did is he hired a new person to sit on the C suite.
He hired a new SVP of AI, Dan Denker, who is a former Amazon guy that ran product at Instacart and said, hey, you are our new AI czar. Your job is to help me make these 1.7 million associates AI native instead of AI fearful.
And, you know, Doug would hate me for pointing this out publicly, but it's all public data. Doug's one of the most successful CEOs of a public company in history. He paid himself $29 million in his last year on the job.
He paid Dan Danker $44 million.
Adam Davidson:Wow.
Jason Goldberg:And so you look at these huge companies that are having good outcomes by being fast followers at least, or first movers with AI, and they started with Doug McMillan waking up and saying, you know what? The last thing I'm going to do before I retire at Walmart is set them up for success in this new environment. And that means put new leaders in place.
That means put new processes in place. And most importantly, give everyone permission to do their job differently tomorrow than they did yesterday. Right.
Adam Davidson:And so eat the risk, own the risk.
I mean, that is something we're learning from a lot of our member companies is when it's something I want to ask is, you know, what happened to the last person to really screw up with AI? Because my argument would be there's a good chance that person should be promoted or celebrated. Yeah.
Jason Goldberg:But the guy I work with who runs retail at Amazon, Doug Harrington, like, launch the fire phone.
Adam Davidson:Wow. Yeah. What are the messages? I mean, we had one major American company tell us that the CEO was like, why isn't this working?
And they realized that a part of just using software there was, like, someone three levels deep who was only Incentivized to avoid risk. And so you needed more senior people to say, you can give us analysis, but you don't make the decision.
That's a business decision, not a risk avoidance decision.
You said at the beginning of this conversation that tam, the total addressable market for agentic commerce is essentially a rounding error negligible.
Jason Goldberg: In: Adam Davidson: In: Jason Goldberg:So there's still one in Beaverton, Oregon. Yeah.
Adam Davidson:Oh, that's right. I did hear about that. Yeah. So what, what, what is our. How urgent should we feel? How scared, how excited?
Jason Goldberg:I really do think we need like the, the charts that scare the bejesus out of me are like, how long it took all these new consumer behavior. I don't care about the tech, I care about the consumer behavior. Like, what new tools are we using to shop tomorrow that we didn't use yesterday?
And how, how fast were those tools adopted? Right? And most changes were super slow and you know, took, took.
If I looked at the total history of commerce, there's like six big disruptions in commerce over 6,000 years, and most of them took hundreds of years to stick. So digital is the new one. AI is to me, a subset of digital in that, in that broad 6,000 year historical context.
The thing is, they're getting adopted exponentially faster than anything before it. Right. And so that this is more urgent because of how much faster this is all going to play out. Right?
And that, that to me, I actually don't have the answer. But like, I talk to CIOs all the time that are like, oh, I need to buy new AI tools. I'm going to go, you know, do a vendor selection, right?
We're going to do a shootout. I'll put someone in charge.
We'll go interview all the vendors, find out what their tool does, how much it costs, put a committee together, and in six months we'll know what tool to buy. Those tool companies will all be out of business in six months, and whatever is the best today will not be the best then. Right?
And so, so that CIO saying, oh my God, I need like my whole process for making these significant investments of my company's resources in tools and capabilities has to be wildly different than everything I became good at to earn the job as cio. That's a bad day.
Adam Davidson:I mean, that magical word exponential means like the last seven months have been six Months have been crazy. Like if we think about what like December 4.8 and 5.4 or 5.5 launched to today. The wrong conclusion is, oh, the next six months will be crazy.
The right conclusion is, oh, the next two months will have as much change as those six months. And then it's going to take a month after that and then a week after that and then. And all those numbers are.
Jason Goldberg:Yeah, I think somebody much smarter than me came up with that quote. Like, the pace of change has never been this fast before and it will never be this low again. Like, that's the.
The terrifying I had a bucket list moment. I got to do a keynote speech at the National Retail Federation on ajantic Commerce in December, in January.
And I, I like a big message in my thing was, Chad, GBT is weeks away from launching this new thing, instant commerce, where you can buy the product right in Chat gbt and you'll never even pick a retailer. You'll never go to Walmart or Amazon. You'll just buy it straight from ChatGPT and you guys better be ready. This is a huge thing.
And now I'm sitting here in July and ChatGPT did launch that and they gave up on and. And Google is about to launch it. Right. And so, you know, like that's a much more nuanced conversation.
Like again, you know, I don't think they actually got a minimum viable product out there and decided it didn't work. I think they actually ran out of. Of development dollars and decided they needed to protect their core assets from Anthropic and Microsoft.
But nevertheless.
Adam Davidson:Yeah, I mean, but that does mean. Because that's the other piece.
read a computer magazine from: e most right in history. Like:For me, I think if you randomly picked anytime I predicted the future, it would be pretty darn embarrassing.
Jason Goldberg:Oh, yeah, yeah, yeah. I don't like the track record. No.
Adam Davidson:Yeah, yeah.
Jason Goldberg:All right.
Adam Davidson:Phenomenal. Quick just cleanup thing as we wrap up. You've been very generous with your time.
There's a few things I do like to point out that are like less like Theoretical and less challenging one is just don't shoot yourself in the foot.
Like you should have a website probably that if you want it to be scannable by robots, it should be in a format that the robots can easily read and easily see. Like that, that feels like just basic, you know, that's.
Jason Goldberg:It does. Although the robots are getting better at reading anything. So like that, that is a potentially short term problem.
The bigger thing I would say there is decide what you want. If you want the robots to read you, there's pros and cons, right? The robots are your frenemy.
But in most companies, in most of your member organizations, there is not an intentional strategy right now.
Like, like I could go across all the properties that any one of your members own digital properties and the robots are blocked from some and they're invited in others and it, it's nonsensical. So like you should have an intentional strategy and execute that strategy.
And if the strategy is we want our data in the robots, then you ought to be unlocking some of your, your locked gates.
Adam Davidson:So that's one big area. I guess that area plus like get in there and fail feels like the right, like I read that.
I mean that was the main message that I wanted to have you on to talk about that just because we don't know yet isn't an excuse to sit around and do nothing.
Jason Goldberg:And your point? You know, Jeff Bezos was annoyingly smart when he started Amazon. He, he had this leadership philosophy, one way doors versus two way doors.
Are you like.
So the idea is if you can walk through a door and there's, there's no big ramification or cost to find out that it's the wrong door and then you can walk back through it the other way. I want you walking through those doors as fast, as often as possible with as little friction as possible. Right. So take risks, do things.
Every leader is empowered to walk through two way doors all day long.
If you're walking through a door that walks behind you and you can never go back through it and there's a huge cost to it being the wrong door, then we ought to have some governance in place before we walk through that door, right?
Adam Davidson:Like don't fire 40% of your staff and loudly proclaim AI can do what they did.
Jason Goldberg:Great. Hypothetical.
Adam Davidson:Yeah, yeah, yeah.
Jason Goldberg:So the, he's, he's in a midlife crisis and not that involved anymore. So. Yeah, but the, the, the, the thing here is to know the difference which of the things we're facing are one way doors versus two way doors.
And we ought to be reducing friction from those two way doors and we ought to be building some safeguards from one way doors.
Adam Davidson:Yeah, I really like that. All right, Jason, this has been a joy. I hope to have you on more often and have you in front of our members, so thanks so much, Jason Goldberg.
Really long title that has something to do with digital commerce. Big Brain at Publicis. That's not.
Jason Goldberg:That's my. Yes, that would be my aspirational title is Big Brain. Very few the people that know me use that one, but yeah.
Adam Davidson:Okay, great. Very happy to have chats with folks. Just reach out to Jessica or Mattie or me and we'll put you in touch.
Jason Goldberg:Great.
Adam Davidson:Jason, thank you so much.
Jason Goldberg:Yeah. Happy commercing, everyone.
Adam Davidson:I hope you enjoyed that conversation with Jason. The thing that's sticking with me is his idea that we're in batting practice right now.
The game of agenic commerce hasn't really started yet, but that doesn't mean you sit in the dugout. You build muscles now, so you're ready when it does. If you'd like to talk with Jason, reach out to me or Jessica and we'll make that happen.
And let us know in the discord what you thought or what you'd like to hear more of. As always, I'm Adam Davidson, and this is the FeedForward podcast.
