Inuvo at 15th Annual LD Micro: AI-Driven Ad Innovation

On Thursday, 10 April 2025, Inuvo (NYSE: INUV) presented at the 15th Annual LD Micro Invitational, highlighting its strategic focus on AI technology and privacy-centric advertising. The discussion revealed optimism about the company's IntentKey platform, but also acknowledged challenges in the rapidly changing advertising landscape.

Key Takeaways

• Inuvo's IntentKey uses AI to target audiences without personal data, addressing privacy concerns.
• The Semicab platform aims to optimize transportation, reducing costs and increasing efficiency.
• Inuvo's acquisition of an Indian subsidiary enhances its position in the transportation sector.
• CEO Richard Howe emphasized the revolutionary nature of Inuvo's technology in advertising.

Inuvo AI Technology (IntentKey)

Inuvo's IntentKey platform is at the forefront of a new advertising paradigm that respects user privacy. By leveraging AI, IntentKey targets audiences without tracking individual data, a significant shift from traditional methods reliant on personal information.

• The technology creates contextual models by analyzing internet content, linking user interests with relevant products and services.
• A demonstration highlighted IntentKey's ability to connect Theranos fraud discussions to Wall Street Journal subscriptions.
• The platform provides demographic and geographic insights, enhancing audience targeting capabilities.

Semicab Platform

The Semicab platform offers solutions to the transportation industry's inefficiencies, specifically targeting the issue of empty miles.

• Utilizing AI, Semicab optimizes routes and matches shippers with bidirectional traffic, potentially increasing truck utilization from 65% to 90%.
• It claims to reduce shippers' total spend by 10-15% and decrease empty miles by 70%, while improving on-time delivery.
• A case study with a major consumer packaged goods company projected a potential $60 million annual savings.
• Inuvo's acquisition of Semicab and its focus on an Indian subsidiary align with the Indian government's NDFE program.

Future Outlook

Inuvo's strategic direction is clear: leverage AI to address privacy challenges in advertising and optimize transportation logistics. CEO Richard Howe's assertion that the technology is "not incremental, it's revolutionary" underscores the company's ambition to disrupt traditional models in both sectors.

Readers are encouraged to refer to the full transcript for a detailed understanding of Inuvo's strategic initiatives and technological advancements.

Full transcript - 15th Annual LD Micro Invitational 2025:

Operator: All right, we will begin our next presentation. So let's please give a warm welcome to the CEO of Inuvo, Richard Howe.

Richard Howe, CEO, Inuvo: All right, well thank you. Maybe I'll just start right away with if you're interested in an investment in artificial intelligence, there's very few of those that are based on real AI. And we are, as far as I can tell, the only company on the planet to a commercialized large language generative artificial intelligence. And our use case is advertising. But just so we're clear, this is technology in Grok in the sense that these are technologies based on scanning and crawling what I call the collective wisdom of humanity as represented by the Internet and then building a model.

And in our case that model was specifically built to solve the advertising problems which I'll describe to you in a second. I may say things that are forward looking. Please treat them as such. First things first, like why did we build this AI for the advertising use case? And the answer for that is because this is one of the largest markets in the world.

And we at Renovo actually focused on two components of that marketplace, the search advertising margin span in The U. S. And the programmatic market. That's about 80,000,000,000. So these are gigantic marketplaces where you can actually build a billion dollar company because there's so much money being invested in it.

For the next slides, and including a demo that I've got here, I'm going to focus on the artificial intelligence, which is the key differentiator in our company. But first, let's talk about the problem. And the problem is these markets are changing. And in fact, it's probably a once in generational change. They do tend to change every generation or so.

But right now, the markets that we serve are in the midst of a quite significant change. And the change is related to the use of your data. For the last thirty years, modern advertising has been based on the exchange of a single commodity. And that commodity is, in fact, your data. People think it's advertising, but it isn't.

We buy and sell identities and the data associated with those identities. That is the commodity. And what's happening in these marketplaces is that the ability to do that is slowly deteriorating. Apple began this change. So it was a technological change that sort of was the catalyst.

But that was reinforced by legislative changes. Countries, states, all coming on board and saying, Hey, it's not okay to track people around the Internet anymore and use their information for ad targeting. In fact, half, easily half of all mobile media transactions today are ignored by the thousands. We have no such constraint with the technology I'm going to show you. And that is in large part why we built the technology the way we did.

This technology is not incremental, it's revolutionary. In much the same ways that media industry went from doing what we know as contextual technology that became the behavioral technology, which is following you around the Internet. The next evolution in this technology is to use artificial intelligence. We knew this seven years ago when we started building the LLM, before anybody was talking about LLMs as a technology. And what we effectively wanted to do with this technology was take the complexity out of process.

It was becoming clear to us that being able to market effectively almost required a PhD simply because the plumbing of the Internet needed to be known, where to get the right data needed to be known, how to do analytics needed to be known. And this was exceeding, for most part, the abilities of people making decisions about what to do in advertising. So we wanted to democratize that in some way and make it easy by providing a simple AI solution to all of those problems. And we boiled the problems into two categories which I'll talk about here in a second. And they were really a category for measurement meaning how do I actually measure effectively the performance of my campaigns across all the channels.

If I can't track people to those channels anymore, how am I going to do that? And the second was, if I can't track people around the internet anymore and follow them, then how am I going to discover audiences or target audiences? Those were the two problems that we trained our AI to be able to do. The first of these is the measurement problem. And it's equally as important as discovery of audiences and targeting audiences problem.

There's a growing number of channels. And up until the onset of privacy, the method that was used to be able to figure out which of the channels you're using is contributing to the performance that you're getting was as a consequence of being able to track you. So I know that you saw a display ad on some web page. And from there, maybe you saw a connected television ad. And then you went to your social media page and you were looking at something there.

And then you ended up it's proven itself to be exceptionally accurate. So we're pretty pleased with this technology. And most CMOs this can become their dashboard. If you want to look at it this way. So how am I spending money?

Where should I spend more money? Where should I spend less money? Turning the knobs, if you will, which is something they all want but rarely get. The second component is is the audience discovery and the audience targeting component. And again here, remember what I said, we were building a new paradigm for our tech.

We were saying we don't want to know who the person is in front of a screen and we don't want to track this person around the Internet. So how do we solve this problem? The LLM we built was the answer to this problem. And in simple terms, the way to think about it is represented on this slide. This example is a real one.

And it's one most of you, since you're all mostly investors, would know. But I'm sure you all remember the Theranos fraud. You know, pretty widely publicized, one of the biggest frauds in American history. The company that broke that, the news outlet that broke that story, was the Wall Street Journal. John Kerry Hugh was the reporter.

And in fact, the individual that was a whistleblower in that case was a guy by the name of Tyler Schultz who is George Schultz's former Secretary of State's grandson. George Schultz was actually on the board of Theranos. Now, because our AI has read everything ever written about the Wall Street Journal, and it has read everything ever written about every product, service, brand that you can think of, It knows all of the reasons why people are interested in the Wall Street Journal. Now that I told you that the Wall Street Journal was the one that broke the Theranos case, it's not hard to imagine that one of the reasons why an audience might be interested in in a Wall Street Journal subscription would be because of the Theranos case, which by the way our LLM told us for this particular use for this case. And you see that represented contextually here with connections between it because in our AI's mind's eye these connections exist.

They all have probabilities on them which I haven't shown here but it has the probability between these connections. But if you look at the Wall Street Journal product on the side thinking the product there is the Wall Street Journal subscription, It's saying, hey, Theranos is highly correlated to the Wall Street Journal. But it knows more than just that. It knows that the Theranos case involved Sonny Belwani and Elizabeth Holmes and the Edison machine and George Shultz. There's by the way there's hundreds more connections just to Theron's this, right?

And that's the product side. The right side of the model is the media side. And there's two iPad screens there. The top one is a connected television. It's actually a movie.

It's called The Dropout. And the bottom one is just a web page on the Internet. It's a biopic about George Shultz. So I'll go to the bottom one for a second and tell you how the AI works here. That page is one of hundreds of billions of pages on the Internet.

And it just happens to be a biography of George Shultz. Now there's nothing on that website that indicates that George Shultz is affiliated with Theranos in any way. But what our AI knows is it knows what the probability is for why someone's in front of the screen for that page. And in this particular case, it would say, Oh, George Shultz, there is a really high probability that the only reason someone came to this page was not because of of the former Secretary of State's storied, you know, political career. It's because of his affiliation with Theranos.

And that's why the person's actually here. And it would match that. It would say, Oh, because this person is interested who I don't know and I don't care who it is, because this person is interested in Theranos, what customer do we have, our company, that also has a similar relationship with Theranos? And the answer would be, oh look the Wall Street Journal is there and it has a high correlation with Theranos. And it would say we should put our ad here.

The important thing to keep in mind here is no data whatsoever has been looked up, accessed to generate this information. It's truly artificial intelligence that didn't exist. It was trained, trained on the collective wisdom of humanity. But it is generating this knowledge. This is a paradigm shift because in the modern way of doing things, you'd look up an ID and you'd go in a database and you'd find all the information about a person.

We don't do any of that. We don't require any data in our modeling. So what I'd like to do now that I've sort of teed up this Theranos thing is I'll show you just how easy it is within our platform to actually build a model like this. And I've got a little two minute it's really a demo. Right?

So just hang with me here and I'll I'll play it.

Unidentified speaker: What you're looking at here is the intent key platform interface. At this point, all a marketer needs to do is define their audience. For example, in this case, we're asking the AI to build an audience interested in the Theranos fraud. Based on that information, the AI generates a set of seed concepts. These are the foundational ideas it identifies as most relevant to the prompt.

Marketers can refine these by adding or removing concepts before the platform finalizes the targeting model. Once satisfied, the platform builds a full contextual model. Here we see concepts like mail fraud, Elizabeth Holmes, Sunny Balwani, and George Shultz, all closely tied to the Theranos story. Each concept has an associated importance score. For instance, Elizabeth Holmes ranks 85 out of 100, which makes sense given her role as CEO and central figure in the fraud.

Marketers can also explore how these concepts connect. Clicking on Elizabeth Holmes, for example, reveals strong ties to Sonny Balwani and wire fraud, reflecting their shared convictions. You'll also notice signals like George Shultz, a board member, and David Boyse, Holmes' attorney. What makes this so powerful is how it aligns with real world interest in the product. In this case, a Wall Street Journal subscription.

The Wall Street Journal originally broke this story, and sure enough, John Kerry Rue, the reporter who exposed the fraud appears in the model. Beyond targeting, IntentKey provides rich demographic and geographic insights. You'll see the total addressable audience, what percentage is highly interested, how interest has trended over time, and overall sentiment. We also break down gender, age, education, and income. This audience skews older, more male, highly educated, and higher income.

And the map shows geographic concentration. Dark green areas are highly engaged. Dark red areas are not. Now building an audience model is one thing, but execution is what really matters. Marketers need to be able to act on this immediately.

With just a few keystrokes, they can select their campaign system of choice all in a matter of minutes.

Richard Howe, CEO, Inuvo: So what you've just seen has never existed before, not even the out of it.

Unidentified speaker: Attendee verification, dial in numbers, and q and a features Test. For the day and time, and you're ready to go. This is just Share the specially generating event link with ANSYS, then use our analytics to review who attended. If you'd like a more individualized experience, you can even schedule one on one meetings with key stakeholders. Sequire Audience offers a complete solution for all

Unidentified speaker: But they are not actually trying to solve the problem of empty miles that we're setting out to solve. So basically, what did you got Procter and Gamble, PEP's load from point A and delivering it to point B? Obviously, that's a huge part of the equation, and that's obviously what we're doing. But we're also spending a lot of focusing on where is that truck going the minute it's dropped off its load. And so the way we're doing that, which I'll get to that in the next slide, but think about it as a giant network.

Right? The more shippers that we have on the network, the more transportation providers we have on the network, the better we're able to make that network more efficient. And this is what we're doing that's completely different than the way that the traditional digital freight brokers are handling it. So what we've been able to prove through some of the pilot programs that we've been doing so far, like I mentioned earlier, Traditional model, only 65% utilization of a truck. That same truck running through the semi cap platform is able to get 90% utilization.

So another way of looking at that is instead of one out of every three miles it's empty, that same truck now is only one out of every 10 miles empty. So huge, huge, huge efficiency gains. We're able to show our shippers that we can cut their total spend by 10 to 15%. We're able to cut empty miles by about 70%. And at the same time, we're able to still improve on time delivery.

And so I wanna show you guys just a visualization of how we're thinking about this and how our approach is different. So this is these are screenshots that are coming directly out of our semi cab transportation platform. The top left hand corner, you're seeing all of this all of the trips that have been successfully completed through our platform. Again, think about it as a network. And what we're doing is we're looking for high volume, bidirectional traffic that is very, very, very predictable.

That's why we love working with these Fortune 500, these enterprise level customers. So I'll give you a good example. Let's say you've got Pepsi out of Texas. Let's say Pepsi is shipping a lot of product from Texas to Chicago or from Texas to New York. Typically those trips, they're always one way trips.

They ship out, they're full, they come back, they're empty. What we're doing and what we're doing differently is that we're looking for corresponding customers, and you can kinda see it here in the bottom left hand corner where it's highlighted in yellow, those are the lanes that we know we can be the most efficient. And the reason we do that is because we're looking for customers that are operating in a bidirectional manner. So my example earlier of Pepsi shipping one way, let's say, from from Texas to Chicago, we're looking for corresponding customers, could be Coca Cola, that's shipping the opposite direction. So they're looking for things shipping from Chicago or New York backwards.

What our AI optimization tool is able to do is we're able to look at all of that load data in real time, 20 fourseven, and it's constantly improving and making those routes more efficient, so that when that truck driver drops off that load to Chicago, that truck knows exactly where it's going right after it drops off, and it's going to the next most efficient point to pick up differently. That's how the software platform works. You can see the same thing visualized on the right hand side. This is now broken out into geographic density. Nice thing about that is it's showing us the lanes, the shipping lanes.

If they're shipping on those lanes that are highlighted in in bold blue, we know we can price that customer really, really aggressively because we know we've got so much traffic pick up, and we're just keeping that truck running constantly back and forth, back and forth. And I should also point out, so we are asset light. We're a technology company. We do not own any of these trucks at all. We don't want to own the trucks.

We want to go very much like an Uber model where we're subcontracting and sorry, subcontracting those trucks out, but we do not want to own them. So I just want to further illustrate a good case study here that we did. So we were working with a large and $80,000,000,000 CPG company last year. And they came to us and they said, hey, if we were to run all of our freight through you guys, what would that amount to in savings per year? So they were nice enough, they gave us about six months worth of real actual shipping data.

We ran it into our platform, we had our AI optimizing tool that analyzed all the data and based on transportation. We ran it through the tool. It came back, and it said, given this level of volume, we believe we could have optimized 77% of all of that traffic. It could have cut down almost 12,000,000 miles of wasted transportation movement, which would have saved that particular customer about $30,000,000 in just a six month period. So you're talking about almost a $60,000,000 savings over an annualized period while still executing on on time delivery.

So pickups and drop offs on time, where this customer is now an existing customer of ours. So I just want to talk a little bit about the founders of the business. This is important. So the the the two founders, Ajesh Kapoor, Vivek Sagal, both career long supply chain logistics developers. Both of them came out of big tech, so coming out of places like Google, out of GT Nexus.

And this is key because we've we've retained both of them. They're both part of the team. We've both we signed them both up to long term employment contracts, where it's key to us that these two individuals are very highly motivated to stay with us and to continue to see the success and the growth of the business. So out is because this software is not vaporware. This is not software that is still in development.

This has already been it's been built, it's been invested in, it's already been commercially deployed. We're already transacting with many, many, many large, major customers, and it's already out there deployed and transacting. We're doing millions of loads. So the software is working. I wanted to point that out.

Now, right now, Semicab serves two main, The US market and the Indian market. When we acquired Semicab late last summer, we acquired their US entity. We're now in the process of closing on their Indian subsidiary. The reason why we're so excited about India is we're seeing most of the growth, the early stage growth opportunities, are all coming out of India right now. And I'll explain the reason for that shortly.

So the Indian government has this quasi sponsored program. It's called the NDFE, the National Digital Freight Exchange. And what that is, it's essentially 35 major consumer product companies in India, all highlighted here on the screen. And what the Indian government has done is it's basically gone to a lot of these companies and said, hey, there's a big problem in India right now. Congestion, infrastructure, these are all priorities within the Indian government that they're trying to solve.

And so they're working with these companies, and they've appointed us, Semicab, to be the exclusive vendor to work with all of these companies to try to solve some of these these very, very large problems, particularly in India where you see just massive congestion. I mean, the average truck in India only moves around 3,000 miles a month. And we've already seen that that same truck through our platform is going from 3,000 miles a month upwards of 6,000 miles a month just by being more optimized and continually moving. So the carriers love it because their asset's always moving. The customers love it because they're getting cost savings.

And we love it because we're creating value, and that's how we're generating, our margin and our profit. So that's it on the semi cap side. I'm gonna switch gears now. I'm gonna go to something very, different. I'm gonna go to, the Singing Machine.

So Singing Machine is a forty year what we do is we design and distribute, karaoke products to retailers like Walmart, Target, Amazon, Costco, Sam's Club. And we also handle, our our international

This article was generated with the support of AI and reviewed by an editor. For more information see our T&C.