If you read our series of e-commerce interviews with Xevio co-founder and CEO, Nadim Kuttab, you already know he has a ton of useful advice when it comes to getting your brand out there. In this series, we’re going to be discussing all things lead generation, starting with the ways in which AI can be harnessed to identify high-converting prospects beyond traditional targeting methods. This shift from broad demographic targeting to behavioral-based lead identification is fundamentally changing how lead gen campaigns are being conceived and executed on the open web, so we were keen to get Kuttab’s insights.
How has the landscape for lead generation evolved on the open web, particularly with the rise of AI?
There’s a ton that I think is going to change in the lead generation game in the coming years. You already have call centers becoming more automated, with lead verification or interest verification being handled by AI agents and automated callers. Everybody wants a quality lead, and that becomes easier when you use AI to pre-qualify people. Because that step is essentially being automated, we as advertisers are getting faster signals regarding quality, which allows us to scale and grow our lead gen efforts exponentially faster. The more data we have, the quicker we’re able to optimize, and the better it is for lead quality and scale.
Lead gen is going to be a big spender in the coming years on anything open web-related. We’ve seen it time and time again at Xevio: We’re able to generate 50,000 to 100,000 leads a month on channels like Realize — it’s unbelievably good quality if you do it right. For lead generation on the open web, native advertising and Taboola has always worked well, and it’ll just continue getting better now with AI.
Your prospects. Your conversions. Realize what's yours.
From your experience, what are the biggest misconceptions marketers have about leveraging AI for lead generation?
Everybody always thinks AI means trash, right? Like, they’re going to have a very bad automated call center that tries to sell you something. When you use AI properly at any step in the lead generation funnel — whether it’s in the ads, the content, the call centers, or the lead qualification — it can be extremely beneficial and prove to be a nice experience for both parties. The leads are warmer, but also the person who has demonstrated interest in the service you offer gets connected to the right people and gets a good quote. We all want to be connected to the right people! AI helps us accelerate that process.
Realize’s Predictive Audiences tool identifies users likely to convert, based on first-party data. How can advertisers best prepare their data to maximize the effectiveness of such AI tools?
The challenge for companies like Taboola has always been, how do I get first-party data? Because people don’t have to log into anything. With Meta and Google, when you’re using their apps or their platforms, you’re constantly logged in and identifiable. You don’t have that luxury on the open web, so the fact that Realize can now segment data based on user behavior, contextual settings, device types, location,etc., is extremely promising, because we know that if you show your ad to somebody who is likely interested in the ad, it’s beneficial for everybody.
In terms of advertisers using their own data properly, they need to pixel everything and send whatever signals they get back to the platform. It’s the most efficient way — the best data is the data that people can collect directly on their platform.
We talked about this a little bit last time, the idea that the more you put into it, the more you’re going to get out of it.
Right! We’re moving into a post-media buying era of marketing where you don’t need to micromanage everything: The AIs and softwares and algorithms will do that for you. But, you need to give them the right signals. If you don’t feed Realize the right data, it won’t know what you’re going after. It’ll be slower at identifying the right consumer pockets and targeting them with ads, therefore it’ll drive up your costs, and others will do it more efficiently than you. Feed data into the system across your funnel to more effectively optimize. The more data, the better.
Can you share an example of how a brand successfully used advanced audience prediction to unlock a new, high-converting lead source they hadn’t previously considered?
Well, we see it as incremental. Predictive audiences are always going to be a lot smaller in scale than when you go broad — if you have a broad campaign doing €30,000 or €40,000 a day on Realize, using the Predictive Audience feature might add a couple of thousand euros at a very good CPA, but it’s not going to help us triple our spend. So, we see it as an incremental way to reduce overall cost: If we’re halving our CPA with Predictive Audiences, we can afford to buy cold or broad users that are a little more expensive. Basically, it lets us spend more and increase our reach. Our clients love it, because it’s effective customer acquisition.
What are the critical metrics performance advertisers should focus on when evaluating the success of AI-driven lead generation campaigns?
I think the phrase “AI-driven” suggests that you’re doing it without AI, and if you’re currently advertising without AI in any step of your funnel, you’re probably not doing it right. Most lead generation nowadays is in some way, shape, or form AI supported.
Either way, the critical performance metrics will be the exact same thing as your normal performance metrics: How good are your campaigns performing? If you’re able to insert AI into any step — it can be in the ad creation, in the content piece, in the funnel, in the post-lead process — and you’re able to improve the metrics on that step, then great. I don’t think there’s one thing that you should look at with AI and lead generation — at the end of the day, the question is, for every dollar you’re putting in, how much are you getting out? That is the most important metric.
Looking ahead, how do you see AI continuing to shape the future of lead generation, especially with evolving privacy standards?
There’s AI that you see and AI that you don’t see. It’s the latter where a lot of the magic is going to happen. It’s like when a user types in their data — there’s an AI that’s already scoring that lead and passing that data back on to the traffic source for optimization in real time. Those are things that the people who fill out the forms won’t even notice, but it’s happening. It will be really amazing to see the speed at which we’re able to become more efficient thanks to the use of AI.
With evolving privacy standards, that’s the visible side of AI, right? People won’t want to talk to AI agents when it’s done badly, and that’s not going to change any time soon. Every time I’m in an automated call and I have to press one or two or tell them what I want, my blood level spikes. It’s a terrible user experience. But, if you’re able to integrate it seamlessly to make the call faster and more efficient, everybody wins.
The effective use of AI will speed up processes, then, make it more seamless for us to get from A to B, whether it’s as an advertiser or a user. Obviously we’ll have a lot of inefficient or bad use of AI that’ll drag down the perception of it, but it’s important to really focus on how to use AI to seamlessly improve processes and improve revenue per dollar, without upsetting people by overusing it and plugging it in everywhere just because you can. As a user, AI shouldn’t be something you even think about.
It’s AI as a sort of invisible helper.
Yes. It should be like predictive texting: You’re writing something and then it finishes it, it’s not annoying and you can continue typing if you want and ignore the suggestion. That is the future of AI, it will shape a lot.
You mentioned that all marketers should be using AI in some form these days, but for advertisers new to AI-powered lead generation, what’s one actionable step they can take on a platform like Realize to get started?
Use AI to take the ads you’re running on platforms like Meta or Google to bring them to Realize — that’s the easiest step. One of the challenges that Taboola was facing for years was, how do we bring people from Meta or Google to Taboola? AI makes that step so much easier.
Realize also has the Social Importer tool for doing exactly that.
Exactly! You guys have a million ways to do this, and they all work — that’s the truth. So, that’s the Social Importer is the easiest starting point. Will it be successful? Maybe, maybe not. But, it’ll give you an idea of what you have to work on next.
If you missed the e-commerce series, no problem — you can read them here:
