Shauli Zacks
Published on: October 15, 2025
SafetyDetectives recently interviewed Kyle Smith, the Founder and CEO of Escalated, to discuss how his company is tackling one of the digital advertising industry’s biggest challenges—ad fraud. With over 30 years of experience in ad tech, Kyle has worked on every side of the ecosystem, from the buy side to the sell side, building the infrastructure that powers online advertising.
Kyle explains what inspired him to launch Escalated, how the company’s platform helps eliminate invalid traffic (IVT), and why its flexible technology can be applied across the entire ad supply chain. He also shares how his team uses data, AI, and constant innovation to stay ahead of evolving fraud tactics while maintaining transparency and trust in a rapidly changing industry.
Can you introduce yourself and talk about what inspired you to
found Escalated?
Hi. My name is Kyle Smith and I have been in Ad Tech for 30 years now. During that time, I have worked in every aspect of the industry from the buy-side to the sell-side. I have built and operated the infrastructure that enables advertising technology, such as an ad server, a content delivery network, auctions servers, and more.
I founded Escalated in order to solve the very specific issue of ad fraud, which is when bad actors use bots and other malicious tactics to generate fake traffic in an attempt to trick brands into thinking they are reaching a real human audience, thus siphoning off their ad budgets.
By the time I started Escalated, ad fraud had hit a boiling point in the industry and actually, there were already solutions out there, but the solutions were mostly being applied to the buy-side, which was really just scratching the surface of the problem. I knew we could create something that could easily be adapted for any side of the picture, whether it be buy or sell-side, or even an intermediary in the middle.
Applying a solution to the full spectrum also solves another industry problem, duplicate bid requests. When a bot is used to generate an artificial ad impression, a bid request goes out to an auction server. That server then holds a real-time auction and calls its partners, who then may also hold their own auction and call their partners, and so on. By shutting down fraud at every step, we immediately free up the resources being committed to process fake requests.
Escalated offers solutions in IVT mitigation, proxy detection, and ad viewability. Which of these do you see as your “core” differentiator today, and why?
It’s actually a bit funny because while the IVT mitigation and proxy detection do go together, we set out for this to be our main distinguishing feature. The fact that we developed a platform that can easily be used by anyone in the industry no matter where you are in the ecosystem, but just as long as you are simply touching media.
We always looked at our Ad Viewability service as being an add-on or a convenience feature. However, over the past year it’s emerged as a differentiator as well. Not only because we offer some unique customization in the ways in which it can be used, but also because the industry has fragmented in that many players once simply relied on numbers and statistics provided to them by their partners, but are now instead trying to create their own internal numbers and statistics rather than relying on others. The fact that we are a neutral company that exist in the middle and offers Viewability, has quite literally driven some clients directly to our doorstep.
Fraudsters adapt fast — how does your R&D / detection team stay ahead of new invalid traffic techniques and evolving bot behaviors?
Adapting fast they do and they always follow the money. So anytime there is a shift in the industry that may redistribute the balance or a new method of transacting, you can be sure the bad actors will come out in full force.
Discovery is definitely a task within itself for sure. Thankfully though, we’ve developed a very sophisticated platform that can alert us to possible emerging trends and threats. From there, it’s a matter of reviewing and verifying. We’ll develop a test that we can then apply to a suspect set of data. If the test allows us to verify the issue with accuracy, then we know that we’ve spotted a new type of fraud or scheme that needs our attention.
What role does machine learning / AI play in Escalated’s detection stack? Are there boundaries where human oversight or rules still dominate?
With the current state of things, human oversight is very much necessary in the detection and flagging of fraud because there is so much money on the line in advertising budgets that a human signature is essentially required on the technology itself. However, AI is going to play a huge role in rapidly speeding up workflows for clients combating ad fraud.
In the past, even after ad fraud was uncovered, how to handle it was a major pain point on the client end. One thing we’ve said very early on at Escalated is to be proactive. Just block it out before it becomes a problem and we do provide the tools for that, but depending on where you are in the stack, you may not always have that option and instead need to rely on reporting generated after the fact.
In this case, it may take finely combing over multiple reports in order to spot problem areas. When you find the one-off issue, then it’s easy enough to add it to a blocklist a move on. The real problem arises when you keep seeing similar issues popping up on a regular basis that might seemingly all tie back to one source or partner. Once this conclusion is reached, it requires going back over all the reporting data to verify and then potentially reaching out to the partner’s team, conducting an audit with them, and then deciding on an outcome whether that’s a makegood, clawback, or something else. But whatever it is, it has already been a terribly time consuming process that has had a real human capital expense.
AI will be able to automate much of this allowing companies to efficiently invest their human capital into growing their future lines of business rather than protecting prior ones.
What are the key metrics you track to measure success (beyond revenue)—for example, detection accuracy, customer retention, reduction in fraud losses?
The number one here is retention. In many types of businesses, if you have a high turnover rate, then there’s something wrong with your sales process or product itself. In SaaS, you’re either reaching the wrong customer base or your software is adding little or no value.
Many of our clients have been with us for years and that speaks to our other successes from the accuracy of our detection, the ease-of-use of our platform, to the prevention of monetary losses. We’re also very customer service oriented and will work with our clients on any issues that may arise if they request our assistance. All things considered, we have been able to earn trust over time and that is why retention is the top indicator of success.
What are your short-term (next 12 months) priorities and long-term (3–5 years) aspirations for Escalated?
Our main focus at the moment is to acquire industry accreditation, which ensures larger player that our software is constantly updated and adheres to industry standards and guidelines, that we work with other industry organizations, and ultimately our software does what it claims.
In the long run we want to be able to expand and introduce our technology at scale by diving into different segments and facets of the industry and even other industries altogether as other markets exist in the adjacent.
However, these goals will all coincide will our own advancement and integration of AI within out platform. As I touched on earlier, we’re very much interested in creating better workflows and outcomes for our clients. Ultimately, our focus is to not only save you money (by preventing fraud), but to also make it seamless by saving you time as well.
