What's the Only Moat Left for SaaS in the AI Era? [Ft Sumner Vanderhoof, Propensity]

The value of our product to our customers is the continued intelligence that that grows as it as it learns. It's not the product. It's the process and the intelligence that gets added to it. To survive as a SaaS company, you have to make the leap from engineering as a competitive advantage to intelligence as a competitive advantage. Fast forward to 2030,

like, what do you really... Where do you think the AI go to market orchestration? Where's that... What is it gonna look like?

If we're selling to humans, think it's a pretty clear future. If we're selling to agents, it becomes a whole different bag.

Welcome to AI twenty thirty. I'm Keith Jensen, president at CADRE AI. We have Sumner Vanderhoof here. Welcome.

2030. That's a long ways out for, you know, just to start Kind of. Not so long. I saw, like, a trailer. I'm a gamer. So I thought... You just reminded me of this. I saw a trailer for a game. Starcraft two. It's like old game from twenty five years ago when I was a kid. My son's 10 now. So we were like into gaming. And now it's like the expansion or the next version of it, and I was so excited I get to share it with him. Release date, 2030. It's 2026. I'm like, what the hell? I gotta wait four years for this game to come out.

There was... I was just watching something where it was looking at, like, it was an older movie, it was looking like a like a dystopian future. Yeah. And it was 2032. And I'll go, get out of here. That's like right around the corner.

That's right. I know. Somebody mentioned 2000 one, A Space Odyssey yesterday. I'm like, oh, that's an old movie that predicted the future when I... You know.

Yes. Which is the past. Way past. So Sumner is a co founder and CEO of Propensity and you're an ABM... Would you consider yourselves an ABM platform still to this day?

Great question.

Let's start... Account based marketing for those that don't know. Yeah.

Yeah. So that's... And it's really b to b targeted marketing is what ABM is in a nutshell. But like the the ABM brand or let's let's call it movement is still strong. Like there's still a a big community around that that wants to do ABM right. So we support that wholeheartedly. And it's core of where we started. It's who we are as a company. At the same time, 2030 is closer than we all think. So we're like, what is the evolution of ABM? So we consider ourselves pushing the envelope of what that is into a broader go to market motion as well. So AI go to market isn't just ABM. There's a whole new category. I don't have a good name for it yet.

Yeah. You know, I would call it like AI go to market orchestration or something like that, but that's a mouthful. So... AIGTMO? Yeah. There you go. That's right. The world will tell me what it is at some point, but we're helping shape that future. And I think it goes beyond the traditional ABM history. And a lot of that is like ABM with like a learning loop built on top of it through AI. That's... So yes, ABM is our core, but we're always, like, looking at what what does, you know, two to four years look like from now, especially with the stuff that's going on in the market.

Well, it's crazy. Right? Because, like, the amount of stuff that you can do now with go to market... I mean, you think about just just me riffing. I'm sure there's a million more you... We could all think of, but, like, you have a... You go, AI can help with SEO and GEO.

Yep.

AI can help, like, what what you guys do of essentially matching the the audience to, you know, device IDs and all that and getting ads directly in front of the people you care about.

That's

It can help you with sourcing your TAM.

Yeah.

We can do TAM waterfalls. You don't have to just go to like one database anymore and subscribe to, you know, an Apollo or seamless, something like that. You can do waterfalls to go drop down and be like, what is the most robust and complete TAM that we can get? Yep. You can then do... I mean, obviously, with the cold outbound email is still a thing. People say cold outbound's dead. It's not. It still, you know, happens and works. All the way through, you know, figuring out who's actually hitting your site Yep. Without them filling out a form

Yeah.

Can help to building the the loops of the workflows and sequences. I mean, it's just like where does it start and end?

It's crazy. And I... No. It's... All of those... Everything you mentioned is is is also like a data point for learning that you can apply to the system. So I think of those as hypotheses. Like you said a hypothesis. This is what we think our TAM is. This is what we think the definition of in market within that TAM is. So we're going to set up those waterfalls to find the people and the accounts that meet that hypothesis. And then we're going to run it through a cohort, an engine, a go to market engine. And every moment is a data capture point. And if you can capture the data so I'll give you a streamlined example here. So if you take a market for us, like everybody that's researching account based marketing or that has an open job role for an account based marketer, just whatever that signal is we're looking for, then we market to the buying committees.

Not the company. We market to the discrete buying committee people at those organizations in an omnichannel way. We get into every conceivable delivery point. We measure who's engaging, what signals have changed, and we validate whether or not there is an active initiative there that's worth sales pursuit. And then we don't push the data to sales. We ask them to claim the leads. Tell us what or reject them equally important, right? Like hey, you sent us something that's junk. Kick it back. We're not asking so that we can manually figure out what's going on. We're asking for them to declare, claim, or reject so that the engine can learn. So now every component is it.

So you go through that whole loop, hopefully results in some ops. And then the next, you get rid of the false positives, you shake the hypothesis a little bit tighter, you go and you tighten up the channels and the messaging, you tighten up what it means to validate those leads, and you pass off a new crop to sales. And now they're claiming more than they're rejecting. Yeah. And they're building more ops out of it. And it just keeps getting... And now the intelligence is built around every single one of those like those traditional data points that we used to capture. So if you set that engine up and then you put AI on top of that to learn, man, is it like it's a whole new world of of, like, intelligent go to market that we didn't have before.

And so you guys are a SaaS platform. Yeah. Right? And we keep hearing about SaaSpocalypse Yeah. And, like, you know, oh, it's so easy to code now and like, all you need to do is take screenshots and drop it in and like, Claude will rebuild it for you. Clearly, you guys are still doing well and and thriving. Like, what are you doing differently, you know, with all of this kind of hype around SaaS going away?

Anxiety is a powerful motivator, right, like around this stuff. So it's like, yeah, I'm always looking over my shoulder and it's like, what what is the next crop of like Y Combinator graduates gonna come disruptive. It's a great measuring stick to say, can you keep up? Can you outpace what they're doing? So we look at what's possible and what's the moat that we can defend against going forward. So traditional SaaS tooling is not enough. Like six months ago, would have said, hey, we have deeply entrenched partnerships with LinkedIn and Meta and all of the ad networks that we have. Those are hard to replicate. Sure. But all the engineering around it is table stakes now. Somebody can go build this stuff. So to survive as a SaaS company, you have to make the leap from engineering as a competitive advantage to intelligence as a competitive advantage.

So when you start for us, we're building that learning loop I just mentioned where we go through all the data points, and the system learns and loops back in and gets smarter. The value of our product to our customers is the continued intelligence that grows as it learns. It's not the product. It's the process and the intelligence that gets added to it. That's a gift that's only been around for even three to five months. So the SaaS companies of the future that figure that out I think are going to be the ones that hold on stronger. So again, six months ago if you had asked me the same question, would have said, oh yeah, become the execution layer, the headless execution layer that AI can lean into.

But if it's just a tool, there's no longevity there. You're going to have churn like crazy. But if every time your execution layer is used, it gets smarter, over time it's going to be impossible to replace because the value is in the intelligence that it's gathered, not in the process that it facilitates. So we've made that move head first. So we're building the intelligence engine for our customers' go to market motions and helping them be smarter every single time. And the more they use it, the longer and more valuable we become within their business.

And I've used your platform before. So my background is is as a CMO. And when I came across the platform, was like, wow. This is really... It's interesting because I had only used his... Like, the legacy kind of account based marketing platforms out there. Would take three to six months minimum to set it up right. And then one of the things that annoyed me so much back then was I'd be like, oh, okay. I wanna market to these types of businesses and they'd like, okay. Here's all the businesses. I'm like, yeah. Yes, no, yes, no, yes, no. And then I wanna market to these types of people and it'd be like director and above in operations or marketing, let's say. And then they'd be like, great. Here... Are you ready to launch? I'm like, wait a minute. But you're not gonna tell me who those people are? And it's like, no.

No. We're not gonna tell you. We're just gonna market to them. I'm like, this this is so weird. And with your platform, I could actually see the people that we were gonna do the device ID matching and everything to to get in front of them, was really unique. But I guess my my point here is that with it being a while since I've used the platform, like, does that that loop that you were describing, like, change? Like, is it is it making my media buy better? Like, if I have a list of 300,000 people that I'm trying to get connected TV in front of and display ads and all of that, is the loop actually working for marketers? Like, what's the value?

Yeah. So first of all, if you're a marketer that's listening to this and you're doing... Like, let's call it black box marketing, which is basically like we think we're generally targeting the right people by job title ish maybe, you know, using broad match criteria, then you're you're doing yourselves a disservice in your company. You're wasting media spend. You have no way... There's no measurable outcome from those impressions and clicks unless they directly convert into a form fill that's gonna give you any data to make the engine smarter. You have no idea who's seeing the ads and who's clicking. So it is worth the effort to go and evaluate a contact based marketing strategy that lets you go one level deeper than what you're doing right now. So definitely take, I'll say it's Keith's advice, and look at contact level because it's gonna give you much more granular information around what the market's doing with your information than than otherwise.

And then to enter the kind of the second part like, the evolution of that is now that you have the contacts and you know 300,000 people, we can measure every impression, every click. Know, some of my favorite channels right now are surprisingly direct mail. Like sending a six by nine postcard. Getting out of the digital and getting physical and sending a six by nine postcard to the home address of a remote worker or the office address of somebody that you know works at a factory or in the office. And then LinkedIn, but don't sleep on Meta or TikTok or Reddit. Those are all channels. Since we talked last, we activated a wonderful partnership with TikTok. And it's maybe not surprising that businesses employ people and they are on TikTok, and as they're doing their doom scrolling, they're more likely to stop and go do something adjacent, business related, now than ever before. We get more traction out of Instagram, Reels, and TikTok than almost any other channel as a B2B company. So I highly recommend looking at contact level, looking at those channels, and don't think that you can't advertise there just because they're B2B.

CTV, we chatted about a minute ago, that's a great channel. And then to your point, because we're at the contact level, because it's all resolved to an identity graph, we can track who saw the ad while they were on their Roku, who saw the ad while they were pretty much every peacock. Everything except Netflix is the only one that we don't really serve on. But we can see where the ads hit and what that resulted in. Was it engagement that produced measurable outcomes for the business?

How do do that

on CTV, though? Because we bid out on the device ID. We know Keith bought a 60 inches smart TV, and when you turn the TV on for the first time, you put your email address in there or scanned a QR code to register the TV in your name. And now that TV has your identity associated with it. And we resolve against it. Keith is presumably in our database, and we have your identity graph basically. So we don't de anonymize is kind of a dirty word. Don't de anonymize, we resolve identities. And when your TV gets targeted, we are serving an ad to that device ID because we resolved that device ID against your contact record in our database.

Speaking of the de anonymize, think this is an interesting one. So it was a handful of years ago that a company came out. I'm not gonna name them here, but the company came out and said, hey, and as... Think there's a few of them now. They said, you can just put this pixel on your website Yeah. And of the 99 whatever percent of people that are gonna hit your website and not fill out a form, normally you wouldn't know who they are. Yeah. But the fact is that your CRM actually does in most cases. Right? Because what happens is once somebody... It's tracking their IP address and all that and once they fill out a form, you actually know all the pages they hit before. Right. But they're not... The CRM isn't gonna tell you who that person is until they take that action to fill in out a form, but we all know Yeah. It is very unlikely for someone to fill out a form.

That's right. Yeah.

And these pixels came out. Said drop them on your site And we're just gonna tell you who hit the site. Yeah. And I remember trying it a few years ago and I think my wife set up our phone plan. Yeah. So like, when I came... Went on a site, was like... Said that like She did. You know, she went You know? And I'm like, oh, this thing's garbage. Yeah. But and I was like, I'm not putting this on. This is gonna be this is gonna be illegal, and it's gonna be gone within months.

Right.

And it's been years now, and it's still around. And your point of the whole, like, deanonymizing is a... Is the word that, like, really is what it means.

Yeah.

But what they're doing is is saying, hey, we have this database already of people that have opted in from other sites and we're just gonna... If the person who's on your site matches one of those, we'll tell you.

Yeah.

And it's maybe what? 60% in terms of the amount of traffic.

Yeah. It ranges depending on geographic area and market. How likely are they to have accepted cookies on another site that gives them, follows them throughout the. So, but no, it's anonymization is okay at the account level. That word is fine when you're de anonymizing companies. But people are not de anonymized. That's a legal threshold you don't want to cross. You resolve identities for people. And in order to do that, you have to have them already in an opt in database that exists. So there's But not you.

The company who's doing it for you does. You don't have to have the Oh right. The end user of this.

As the person that bought the product. Yeah. You don't have to have them. It's the responsibility of a service like... So we do the same thing. As a service like us, we have hundreds of millions of contacts that are part of our identity graph. And not just the contacts, but the devices that they own. So now it's billions of devices. And as they interact, that's how we target them with ads. And that's how we interact with them when they visit the site. We resolve their identities, those device identities, against the identity graph that we maintain. And it makes it a resolution process, not a de anonymization action, which is a totally different and above board type of way to approach this. If you're doing...

A few years ago, it was unclear what that meant, and I was with you. I was like, dude, this doesn't seem right at all. It's like, oh, this is how... If you have the data already, you're resolving identity against it. Clean. Yep. If you don't have the data and you're, like, looking at their browsing history as they visit your site, that's a no no. Just can't can't touch that.

Well, it's... And it's really beneficial to have this because not only... I mean, I don't think it's beneficial in the way that people think it is. Like, I think a lot of companies go, oh, I know who's on my site. I should just feed those leads to the sellers and have them call them. I I haven't found that that works No. That's not what I it's a data point. Right? Like, if you're sending email, even if it's cold email Failed it. You may not wanna put the pixel on the email Yep. That tracks open rate

Yeah.

Because it's gonna... It's more likely that it's gonna go to junk. Yep. So send it without that... Send it without a link And if they go... If they read that email and they don't reply because the reply rate's still gonna be very low Yeah. They might go check out your site. They hit your site, you know you sent them an email, boom, count that as like... You can consider it similar to like an open. Right? And same thing with connected TV. Like, they're they're not clicking on anything when they see an ad on their TV, but they might go to your site. And another perfect example is I was driving. I I just got here to Boston yesterday and like I was driving by the airport and I saw a billboard for a tool that I was like... Was an AI tool. I like, hadn't heard of that. Yeah. And I was like, how is it possible that it's it's it's big enough... The company's big enough that they have billboards. It was at the, you know, one of those, like, bus station type ones. I was like, oh. And I went and looked at I went to their site. Yeah. And if they were, you know, tracking that and they they maybe had me on a list of, like, potential buyers for this, they would see that, like, oh, that might be... That's gonna raise the interest That's right. You're doing like one through five, one through 100 chili peppers, it is growing their interest over time because at some point you wanna pass it to the sales team.

That's

right. Right? But you don't wanna do it just because they hit their site.

That's man, you're circling around. Maybe the the the biggest faux pas that marketers make is data dumping to sales. Like that's the death sentence for any ABM program. It's really any GTM program. You... Hey, they visited the site. Go ahead sales, figure it out. Like, I will say it for everybody, sales is not demand gen. They should not be doing like validating that the lead exists in the market. If sales is treated like... A lot of the times they're incentivized like demand gen, it's a numbers game and it's like just, you know, they're they're they're make the dials, do the thing, figure out who's in market, but then they're commissioned by demos booked. So it's like there's a misalignment in sales incentive structure and marketing is sometimes guilty of enhancing that just by dumping data in their labs. If marketing becomes like a validation engine where the job is to not think of these even as scores zero to a 100, I like to think of them as confidence intervals to use like more of an AI modern term. We're

using like multifaceted modeling to understand somebody's engagement. Did we send them an email? Did they visit the website? Did they click on the ad? Did they recently hire somebody in the position? Like every conceivable signal gets aggregated into a model about that person, that company, and the buying committee around them. And then the further upmarket you go, if you're targeting Microsoft, Microsoft is never like a hot account. But the Seattle office for a particular initiative might have five people that are hot for that initiative. So what we're trying to find are the five, 10, 15 people that are the direct buyers or influencers around that decision for that one initiative. So now you're even going a step further for sales. You're not saying that Keith visited the site, you're saying Keith from Cadre visited the site, this is what they do and this is what the potential outcome could be going forward. And this is so why this account, why now, and what's next is what we're giving sales with the confidence intervals to say how to proceed with it.

That's all... Building that engine is a nightmare again a year ago. Yeah. Now it's becoming like everybody should have confidence interval modeling that they're doing instead of scoring. And go way beyond just like a... And it's rehashed every day. Oh Keith was a hot lead yesterday, today we have more signal and more information about him. We're gonna you know, devalue that lead because he's a partner and not a direct potential customer. It changes dynamically and now you're pulling leads back from sales

Yeah.

And then passing them new opportunities. And they're, again, claiming and rejecting him, and they're telling you, dude, Keith is not a customer. Get him out of here. Like Yeah. He's doing competitive research. Like, I don't know what's going on here. So just... And that'll that'll inform the model for the next Keith Right. That, you know, visits the ad.

So you mentioned something that I wanna double click on, which is you mentioned signals and and, like, hiring or jobs. I have found this to be tricky, at least the engineers that that I've talked to have said it's tricky to to flag when, like... So you can go scrape a company's website to see if they have it on there. But a lot of times what what you want... What companies wanna do is they wanna say, hey, is is the is the person that I'm... People that are in market or or the people that are in my TAM, are they posting about a certain thing on LinkedIn? I wanna know because I wanna go engage with that. Yeah. Or are they... Is that company... Did they post a job Yep. Like this? Or are they saying certain keywords in their posts? Yeah.

And I've heard that LinkedIn is really specific about being able to get in and and scrape and and get those kind of signals.

Are you guy...

Have you guys been able to crack that or no?

They used to call it social listening. Yeah. Maybe they still call it that, but it's like become a lot harder to do at scale. Social listening, I think, has evolved from we'll listen through the fire hose and hear everything that's happening around any keyword. And now it's moved into we can only listen when our brand is mentioned. So listening, and this is meta all of, I mean as soon as the, what was it? 2021? When they started shutting off API access across the Twitter first and then Reddit, they did it to create... Because they saw what was coming. It was the data wars were starting. And they're trying to create silos of data that can't... That's where they make their money so they need to keep it there. So they've moved from an open model that way you could do social listening at scale to a closed model where you can only do social listening around your brand. So we've moved the way we think of this from social listening to contextual advertising. So what we're doing is we're saying anytime somebody uses a specific hashtag or likes a post with a specific hashtag on it, so like ABM.

We'll find them on TikTok or LinkedIn or Instagram, and then we'll target the people that have interacted with that hashtag. So it allows us to have contextual interactions. It's not as direct as social listening, but we have customers that'll be like, hey just for TikTok Mexico, if somebody mentions our brand name, we want to advertise to anybody that hashtag our brand inside of it. And now they have a very targeted way to interact with like previously something that would have been off limits to them and they couldn't listen for anyway because it's a hashtag and not a brand mention. So that's the... It's a high intent signal that's until the last couple of years been buried behind advertising capabilities. Now you can do contextual targeting like across all the major media platforms.

So you're saying that if I had a 100,000 people or whatever in my TAM Yeah. And I wanted to know whether that... One of those people put a... Like put a post up with a hashtag Yeah. I can do that.

You can do... As long as they hashtag a keyword, you can target them. They're not gonna let you tell you exactly who because you can't scrape LinkedIn. Right. It's the violation of the terms. But at least you can retarget them. You can retarget them as though that was a website visit that you're targeting off of. So it's like more of a retargeting motion. Hey, they mentioned us.

Is it have to be us or can it be any hashtag? Any hashtag. Okay.

Yeah. So we have our contextual targeting motions, keywords, hashtags, pretty much anything like if you wanna... Like if somebody's visiting a blog that has a keyword about... That you care about in it. So if somebody's reading a blog on account based marketing Yeah. We'll target contextually like we'll be on the leaderboard ad on the side or

something like that, the skyscraper ad. Yep. Cool. And... But but just to be clear, it is... You can't... If if someone in that list posts on LinkedIn and use it as a hashtag, you won't know that they did it on LinkedIn because you just can't. Right? Because it's just... LinkedIn is blocking it. Yeah. Yeah. And and again, that... You're saying that's because if they allow that data to be fed out to others, they are essentially democratizing it and it's no longer as valuable to them to sell.

They need your eyeballs in their platform to sell ads to other people. So if they're giving you a way to get that data out of their platform, they're losing your eyes. So their their value is your eyeball inventory. So they have put... They're protecting that like in in an... At an insane clip and they'll keep doing it which makes us as a product valuable because we can find ways to find out which eyeballs we want to look at the ads, right, in an intelligent way. But I think from a, like, man... Another example of that is, like, with AI right now. I would love to know everybody that says, give me a list of the top ABM platforms. Yeah. Like, if only ChatGPT would just feed me that data, I would just be, you know, super happy.

There's no way they're ever gonna do that. It's like Google is never gonna tell me who searched

Yeah.

But I can buy keyword ads. All that's swirling around the same topic. Like that's our proprietary... We can't violate the trust of the user and give you that data. Right. But you can, you know, pony up some dough and pay for an ad when that occurrence happens. Yeah. So that's that's the model that these guys are all adopting now is like, we'll give you more insight into the the actions and then let you advertise around those hashtags or searches or now with chat GPT, you know, profiles or or topics that they're searching about.

I don't know about you, but I... My Instagram or LinkedIn feed is just total nonsense about, oh, Claude got me 80 leads this month and all that, and it's just... It's... I just don't I just don't believe it. Yeah. We have not seen that actually work of like AI just going and figuring out how to go and get the leads and everything. There's just so much stuff they're not telling you.

Yeah.

Right? Of like, oh, well, we already had a list and we already did this and we then... All it did was like help us write the emails. Like it just everything makes it seem like like... I think if I was if I was not in AI and I was on the outside, as a marketer, I'd be like, oh my... I'm so behind. Yeah. Because look at what these people are doing and like now... And I'm on the inside, I'm like, they're not really doing that. Yeah. That's that's ********. Yeah. So so I guess my question to you is, like, given the the the hype of what it's... We're hearing is possible, but knowing that, like, we're not quite... It's not as as dramatic as everyone's making it seem.

Yeah.

Fast forward to 2030, like, what do you really... Where do you think the AI go to market orchestration... Where's that... What is it gonna look like?

Oh, man. That's... So from a... If we're selling to humans, I think it's a pretty clear future. If we're selling to agents, it becomes a whole different bag. So that's like... That's interesting. Yeah. Let's let's leave that one on the side because that unpacks like who...

Do we agents who are literally doing procurement and picking the tools and sign... Yeah.

That's... There will be... What do they used to say in the crypto world? Democratized... There's like DAOs. Right? They used to call them... It was like an organization that was strictly run based off of blockchain contracts. And that was the vision of it. AI is actually going to realize that vision, where it's like organizations that are independently run by agents. And they're going to be purchasing. There are going to be some SaaS companies. Now they're selling access to agents. In the future, you will have some companies, maybe a bunch, will have models where they're selling licensing models for the number of agents that can access their tool. So you have five... It's the new user, like a licensing model. So instead of a human user, you'll have an agent licensing model. And you'll be able to have like five agents, up to five agents access it with its price point or 10 with you. So this...

The... There's an abundance that happens there because there's only a certain amount of people on the planet, but there's gonna be billions more agents very, very quickly. So that changes the whole pricing game of the future.

It changes the UI game. Yeah. Do you need a UI? Like, if an agent is accessing propensity Yep. And and doing the... You know, going in and figuring out like, oh, we have a new product. We need to market to a new audience. Yeah. Feed that in propensity. Parenthood goes and builds the audience and starts running ads. Do you even need a UI? Like, is everything gonna be MCP or API?

So I I should I should be clear for the the humans listening to this that There are any. Yeah. There are any still. That I am... I do not see a future where AI, it it, like, disrupts everything and takes all of our jobs. And, like, I think there's a great... Mean these are trillion dollar companies that have the best PR teams in the world that are hyping up stuff for their own business interests. So I would take the SaaSpocalypse or the doom and gloom of the future that AI is bringing as a job market with a grain of salt. Ideally what it does is it gives us four day work weeks or three day work weeks, but it doesn't put us out of jobs. There still have to be human managers to manage the business that we're running, and there has to be humans to purchase the products on the other side of it.

That's not going anywhere. So any idea of universal basic income as a outlet for all of us, I would not count on that. It would be great if we had more regulation for societal benefit through this stuff, but that's probably a different topic, different conversation. So I'll snap back to the... Let's say the humans, which I truly believe are still gonna be involved in every aspect of business in the future. The future of ABM and how we go to market is going to be just, again, smarter intelligent loops to target the right people. I appreciate relevant, meaningful ads that educate me.

When I use Google and I search for something and there's a new product that it told me about through the ad, the keywords that I would never have known about otherwise, that's a benefit to my life. As long as it's relevant to the search that I make, if it's totally out there, it doesn't have any meaning. And I think that's the problem... Or if

we start seeing sponsored in there Yeah. It has less meaning.

That's right. Yeah. So if it's an overt ad that's meaningful to my life and has value, like I will buy products and interact with them if there's relevancy to it. It's when it's just ads for ad sakes that it really sucks. So I actually think the future of marketing and ads is more relevance, better targeting, more meaningful product placement. I think when you walk through the dystopian future of everything's Las Vegas with the billboard, Times Square and the billboards everywhere, they're just trying to capture as many eyeballs as they can on broad reaching But if every billboard is tuned to you or every ad is meaningful to something that's value added to your life, and it's not even just a sales pitch, it's an educational offering. It's hey, here's how to learn more to empower your life by eating healthy because we know that's meaningful to you.

Yeah. Like all of a sudden you're being gifted pathways to better yourself, your company, the people around you. That's the promise of I think relevant ads that AI can help us bring of the future with better targeting. But there's a distaste because advertising companies historically haven't exactly earned the trust of people do that. So you have to be careful. There are gonna be those that use their power for good and use their power for evil, and you gotta make sure that you pick a side. And we are trying to be like, let's do relevant targeted ads because people actually care. And like you said earlier, email's not dead.

Advertisements can still lead to form fills if they're meaningful ads to the right people through the right channels. And you just gotta you gotta penetrate through the noise with that. And eventually, I think some of the market will move that... Not the entire market, but some of it will move there.

The hyper personalization of of ads in real physical environment Yeah. I think is one of the most exciting Opportunities for advertising as we we get more advanced with this stuff. I had a couple folks on the podcast recently that were talking about just like the one being like there's a VC firm that is investing in the companies that are building out the pipeline to be able to... Knowing that the amount of data that we're gonna have to flow through these, like, archaic infrastructures Yeah. In order for for, you know, robots to work in a physical environment or, the... You know, another company that connects directly to a camera and just feeds a JSON out Yeah. That you can use AI for like any kind of camera. Like, all these things need...

They're they're coming. It's happening. I almost wonder if the the whole idea of the glasses is going to come back really hard where when you had these billboards and stuff in public places, maybe they have a standard ad that's up there. Yeah. But if you have your glasses on and it knows you and your personalization as you look, those things are tailored directly to you. Yeah. And to your point earlier, I... I've always been a kind of strange marketer where like, maybe maybe not, maybe because I know what... What's happening, but like I will reject every cookie. Yeah. I will spend the time to go into the preferences of every website to be like, nope. Nope. Nope. Strictly necessary only. Having said that, I'm now... If if Claude or Anthropica came up with a pin that I could put on my shirt that just recorded everything in my day, I would wear that thing. Yeah. Because the value that I would get out of AI having access to everything I said from a personal perspective of like, my wife said you didn't say that, and I'm like, well, let's check. You know, let's black mirror this thing and like, let's go find out.

To being at work and just being like, I don't wanna have to upload that

fire

flies That's right. Conversation. Like, I just want you to know everything I said because the value I'd get out of it would so so high. So I'm open to that Yeah. Quote unquote invasion of privacy because of the value it adds to my life.

Wait. Wait. Wait. The the word I've been using for that is context. The more context AI has, the more valuable it becomes. Yeah. So when we first started adding AI to propensity, I called it dumb AI. It was very much like, what do you want me to do? Okay. And put the blinders on and it would do the thing. So we added like execution layer first and then we added a conversational layer. It's like okay, now I'm not just gonna do the thing. I'm gonna ask you more about like what did you really want me to do? And the third layer we just added recently is the contextual layer. So now it's the learning layer. It's gonna follow your ads. It's gonna follow your audiences. It's gonna follow your CRM data. It's like the pin for your business. So now propensity follows your go to market motion everywhere it goes and it records every interaction and it and it informs that constantly. And so it it automatically pulls from Fireflies.

Says, oh, is what's this is what's going on in the sales world and this is what's meaningful and your messaging for marketing is over here. Try pulling it back this direction. It's the equivalent of that except it's you know most ads and most interactions are digital these days so we have the opportunity to put that digital pin inside of your business and track everything, centralize it, and that's where the intelligence and the context really makes things like super interesting for companies that couldn't... Like you can do that, but what a pain in the butt. Like you said, I don't wanna upload this to every single Now it's just, oh, just connect Salesforce or HubSpot or Snowflake or Clickhouse or whatever it is and it'll just... It'll rip through and add the context not just for your business, not just for Microsoft, but for Steve in the Seattle office for this product initiative. Here's the context and here's how it fits in the greater whole of things. Here's how that conversation with your wife fits in the greater whole of your relationship with her and why you probably shouldn't say, see, I told you so at this moment.

Or I'm gonna let Claude do it.

Exactly. You know? Send my wife an email letting her know she was wrong. That's right.

Alright. Well, thank you so much for coming. This is such a fun conversation. I love talking about this go to market stuff. Honestly, like, the the... I could say, like, having worked with with Propensity in the past, it's it's... It was a pleasure to to how fast it got onboarded and the the data that I was seeing. That was pre AI. So... I know. It's gotten faster and better. Yeah. We gotta we gotta check this out. But thank you for coming in. Appreciate your I

love these conversations. It's a lot of fun. Thanks. Thanks, Keith.