skip to Main Content
Abstract lines. Vector art

“Come on. All this AI stuff, around marketing and sales, that’s all just hype, right? That’s not a real thing that matters. Maybe it’ll matter someday. But it doesn’t yet matter today.”

That’s a question I got recently, and it got Steve Zakur and I wondering where we are on the hype curve. Is it really just all hype? Maybe not — platform companies are snapping up data scientists the way early internet companies snapped up web developers.

The real question might be “is there value in AI?” And that really comes down to what problems you want to solve, and what data you have to solve them. Do you really understand the business problem you want to solve?

Another question I get is how to know which AI tools are really valuable. Salespeople will try to fill in the gaps in your knowledge, but how do you know what brings you value? (Hint: you don’t need to be a data scientist, but someone on your team needs to be to really succeed.)

0:00 Intro

2:10 Is AI all Hype?

12:05 Knowing what AI tools are really valuable

15:30 What business problem are you actually trying to solve?

19:55 Expectations

26:10 Outro

SearchChat is available on

Search Chat is SoloSegment’s podcast dedicated to all things search AI and content marketing related. Who is SoloSegment? We’re a technology company focused on onsite search analytics and AI-driven content discovery to improve search results, increase customer satisfaction and unlock revenue for your company. If you think we might have the answer to your conversion problems, feel free to connect with us.

Tim Peter: Hi, I’m Tim Peter and welcome to SearchChat, SoloSegment’s podcast dedicated to all things search, AI and content marketing related. Who is SoloSegment? Well, we’re a technology company focused on site search analytics and AI driven content discovery to improve search results, increase customer satisfaction, and unlock revenue for your company. SoloSegment. Make your search smarter. You can learn more at

Tim Peter: On this episode of Search Chat, SoloSegment’s CEO Steve Zakur and I discuss where we are on the artificial intelligence hype curve, and how you can make artificial intelligence work as a sales and marketing leader today. We also talk about how to deploy an AI initiative within your organization to increase revenue and unlock value for your company. All that and more on the latest SoloSegment SearchChat coming at you right now.

Tim Peter: Well, good morning Steve. How are you doing today?

Steve Zakur: I am doing really well. It’s a Monday morning here in SoloSegment land and so yeah. We’re catching it early, but yeah, doing great. I had a nice weekend, nice restful weekend, so it was good times.

Tim Peter: That’s fantastic. Well, we sort of reached the dog days of summer. I know it’s-

Steve Zakur: Oh man.

Tim Peter: … hot and crazy and all that other kind of stuff. Things are a little quieter, right?

Steve Zakur: Yeah. Indeed. Indeed.

Tim Peter: So it’s probably a good time to take a step back and ask some bigger questions.

Steve Zakur: Oh boy, big questions.

Tim Peter: Think about where we’re heading in the longer term. And when I say we, I mean we as marketers, we as business folks who care about digital, and about search, and about AI and things like that. And I have a question because it came up in a conversation I was having with someone the other day that I would love to hear you talk about a little bit, which is, you know, come on. This is the conversation I was having with somebody and he said, “Come on. All this AI stuff, when we talk about marketing and all that, and sales, that’s all just hype, right?” Like that’s not a real thing that matters yet. Maybe it’ll matter some someday. This is what the guy’s telling me. Maybe it’ll matter someday, but does it really matter today? And I thought, hmm. Not only is that a fascinating question, but also what a fascinating question for our podcast.

Steve Zakur: I love it. I love it. And it’s kind of interesting because over the weekend I was working on a blog post, which is very similar to this, but maybe we can work that in, but it’s this hype curve. Where are we? Is it real? What does it mean to my business?

Tim Peter: Right. What inning are we in?

Steve Zakur: Yeah, exactly.

Tim Peter: That one, right?

Steve Zakur: Yeah. I mean I definitely think it is early innings. I was struck… You know, I drive an older vehicle, it’s all of five years old, right?

Tim Peter: Right, right.

Steve Zakur: So it doesn’t benefit from all the newer technologies that are embedded. Like Subaru’s these days come with all of the lane change, and the radar, the LiDAR, all that stuff is kind of built in. And so I had the opportunity, I brought my Volvo in for some service and I had the opportunity to drive one of the new models. And it’s one of those that has, again, that they try to suck you into buying a new car, right? So you get the top model with all the bells and whistles, and this thing, of course, you get on the highway and you say, I’m going straight, don’t hit anything, and the car will just go, and it’ll go 65 or it’ll go 30 if the traffic gets tight. And then if you go to a Tesla, right? You see that.

Steve Zakur: And I guess my point is, today if you don’t believe AI is going to matter, just go drive a car because that’s all AI right now. Now the thing isn’t driving itself, but there is a model in there built from machine learning that just basically is doing predictions continuously. Am I going to hit something? Am I going to hit something? Am I going to hit something, right? And it’s just asking that question continuously and using the tools that it has, basically the levers in the machine that it has to not hit stuff. So first and foremost, I mean if you doubt at all that AI machine learning is a thing, just go drive a car. Go drive a modern car. You will know that this is a thing.

Tim Peter: Right.

Steve Zakur: But of course the tougher question is, what does this mean for B2B? So now you’re a digital marketing executive, what does it mean to you? And I think kind of the first indication of it’s not hype is platform companies are snatching up data scientists the way early Internet companies snapped up web developers in 1996, right? Tim Peter: Right. Right.

Steve Zakur: And so we are definitely at the beginning of the next Internet, or the next kind of big thing. And so no longer are full-stack developers going to be a thing unless part of your full stack is I can do R and Python, and pick your AI language or technology, that you can do those things to do the data analytics to build the tools on the platforms in order to do miraculous things.

Steve Zakur: So I think that’s part of the indicator for me that AI, we’ve gotten beyond the hype, because you look at Microsoft, Google especially in their platform business, but also in their application business. You look at AWS in their platform business, you look at IBM, I mean even IBM for God sakes, you know? These are companies that are snatching up data talent like crazy.

Steve Zakur: And so again, it’s 1996 again, and back in… And by ’99 we were like, “Oh. Internet’s a bubble.” And all the sudden there were web developers on the sidewalks looking for jobs. And so we might have that correction, but I also think we’re going to see the 2000’s again with regards to AI skills. So we could be peaking in the way we did in ’99, 2000 at some point, but it’s just to go into the trough of disillusionment so that we can then emerge with the right model and so…

Tim Peter: Right.

Steve Zakur: Yeah, yeah.

Tim Peter: Well, it does feel a little like we’ve sort of hit a little bit of that trough of disillusionment. A little bit, you know what I mean? And maybe it’s just the early days of that, but at least from my perspective, it seems like there was a lot of hype about this a year or two ago, and all of a sudden everybody’s starting to ask these bigger questions, which is an important step, but also it ignores all the stuff that’s going on on the ground, right?

Steve Zakur: Yeah, absolutely.

Tim Peter: It’s a [inaudible 00:06:51] where this actually comes to play. So given that, what are some examples or… We were talking about some things a little earlier. What are some things that people are seeing where this is real, where this actually does make a difference and it’s important for you to focus?

Steve Zakur: Yeah. Well, that’s a really interesting question. So again, I try to talk to a lot of people. It’s kind of part of my job is to talk to a lot of people about digital marketing, marketing, where things are going. And where you see this, there’s a little less hype and a little more reality. I think there’s a lot of technologies in the demand gen space that are looking at all the data, like if you’re doing email campaigns. I was going to say programmatic, but there’s so much fraud in that business. It’s hard to know what’s real and what’s not. But if you’re doing demand gen, there’s a lot of AI type technologies in there that are trying to learn, is this a good lead or is this a bad lead? So in lead scoring, in evaluation, essentially A/B testing of content and that sort of thing. So I think that there’s a lot of work going on there. I think one of the reasons is there’s a lot of money in that space, right? Every business-

Tim Peter: Well sure. Absolutely.

Steve Zakur: So there’s a lot of money to chase, and you can measure results pretty regularly, which is I think one of the important things when it really comes to understanding, is there value in AI, is you’ve got to have some sort of measurable result. And it’s funny. The blog posts I was writing, you really focused on four things, right? If focuses on business problem, do you have a business problem worth solving and that you understand, and can you measure it most importantly? Do you have the domain expertise to understand what’s going on? Do you have some data science skills to understand the data side of it, even if you’re not building it? And then finally, have you set the right expectations with your boss? Steve Zakur: And if you break those down, I think those four insights are really at the heart of any transformation. But I think that with specific regards to AI, you start with that business problem. And so that’s one of the reasons I think like success that I see in some products in the demand gen space are because the business problem is well understood and highly measurable.

Tim Peter: Right. Right.

Steve Zakur: And everybody’s got the pain. And I think that if a company’s evaluating, we all want to have AI in our objectives for the year, because we want to demonstrate that we’re doing something so we don’t miss the next Internet. And so having the ability to demonstrate some measurable progress is key. So that, I think, is one of the reasons why that space is pretty crowded with machine learning/AI type technologies.

Steve Zakur: Now whether or not there’s really AI going on in the back or not, it’s always kind of a question for me because-

Tim Peter: Of course.

Steve Zakur: … I see some of those things and I’m like, “Well I could write a rule for that.” Now maybe the machine can write a better rule and they have the machine writing a better rule for that, but I-

Tim Peter: Or a rule that updates over time, something along those lines.

Steve Zakur: Yeah, sure. Maybe there’s some learning going on there as it evaluates the A’s versus the B’s, and then it… Sure. But I sometimes kind of wonder it’s like, well how is that different from me just getting some off-the-shelf marketing automation platform and just writing a couple of rules? But hey. Maybe. But regardless, I understand the business problem and I understand the measures of success. And so that, for me, is kind of rule one in when you think about AI and separating the hype is do you really understand the business problem it’s going to solve, and if it does solve it, that you’re going to achieve some objective at the end of the year or the end of the cycle or whatnot. So I think that’s kind of like anything. It’s really critical. And I think, again, why you see some of this in, at least I understand it clearly in the demand gen space, is because I can understand the business problem and the measures.

Tim Peter: Well, and it’s interesting that you say that because if you’re the head of marketing, if you’re the head of sales, do you really care whether or not it’s doing AI, or do you really care about getting that result? Obviously, obviously if somebody’s saying they’re doing AI, they ought to be doing AI.

Steve Zakur: And by the way, it should be a lot better. So if I… Tim Peter: It should be a lot better. Right.

Steve Zakur: If I deployed my marketing automation stack last year and I got a 5% bump, and now this year I’m going to augment it with some AI technology, well, now you’re not going to go get better than five, so what am I going to get? And it doesn’t have to be 10 X, right?

Tim Peter: Right.

Steve Zakur: Because I think it was Mike says this all the time, it’s like, “Yeah. 10 X would be great, but what’s 1% more growth worth to your bottom line?”

Tim Peter: Absolutely.

Steve Zakur: And the answer for a lot of our B2B clients is, it’s worth a lot of money.

Tim Peter: Well, the story I always tell, and I admit… You know me, I like to tell jokes and things like that. There’s the old joke about the two guys walking through the woods and all of a sudden they see a bear, and the bear starts moving towards them. And the first guy sits down, pulls out of his backpack a pair of running shoes and slides them on. And the second guy says, “What are you doing? You’re not going to outrun that bear.” And the first guy says, “I don’t have to outrun that bear. I just have to outrun you.” Indeed. Indeed. So I think that’s where Mike is exactly right. Exactly what you’re talking about is, 10 X would be fantastic, but if your competition is doing 10% better and you do 12% better, guess what? You’re gaining market share. You’re doing better.

Steve Zakur: Right, exactly.

Tim Peter: And that’s actually not a terrible place to be. So with regard to that, if I am the head of marketing, if I am the head of sales for some company, and vendors are coming to me and they’re talking about AI in their tools and how cool it is and all this other kind of stuff, what do I need to know to know whether or not that they’re actually selling me something really valuable or whether or not they’re selling me a bill of goods?

Steve Zakur: Yeah. And I think that’s the second area is domain expertise. So that’s the thing that I talk to. Boy, you’re just serving these up for me. Thank you. So…

Tim Peter: That’s why I’m here Steve.

Steve Zakur: Yeah, so it’s domain expertise. I think where a lot of people get in trouble, especially with regards to technology deployments, is when they are trying to deploy a technology in an area that they don’t understand, forget the technology, right? Do you really understand how your business runs? And that’s part of what kind of knocks through the hype. Because let’s go back to my demand gen story. If you are an expert in that area, right? So you are the MQL expert, you can stuff the pipeline like crazy with whatever you have today, you really understand the processes, the leavers, the resources that you have, and the data that you have at hand, and that helps you in your valuation.

Steve Zakur: So if you don’t have that domain expertise, you need to get it. You need to find the person in your organization, because they don’t have to understand the technology. They don’t have to understand machine learning. They don’t have to understand any of that stuff. What they have to understand is, how is this going to help run my business, effect the levers that I have? And if they’re a new lever-

Tim Peter: Right. How do we know better is better?

Steve Zakur: Right. Exactly. And so I think that’s another area that requires… And by the way, when I worked at a large enterprise, that was one of the biggest challenges that I had as an IT executive was talking to line-of-business folks and knowing clearly that the person I was talking to didn’t understand their fulfillment processes.

Tim Peter: Oh, sure. Oh, yeah.

Steve Zakur: Okay, so I’m trying to help them improve time to market, all that other stuff. And we might be talking about an SAP implementation or some other vendors implementation, and trying to optimize, and they have no clue. They just have opinions. And so I know that as the executive you want to be the smartest one in the room, but at the same time, the smartest people often realize that they have gaps in their knowledge. And I think that’s-

Tim Peter: Of course.

Steve Zakur: … the error, right? That is the error is that you have gaps in your knowledge, and you’re allowing the salesperson to fill in the gaps versus allowing your domain experts. So it’s got to be a partnership. You have to make sure that your domain experts are at the table helping you to evaluate the technologies that you’re looking at.

Tim Peter: Got it. No, that makes a lot of sense. So you talked a little bit about make sure you have a well-defined business problem. I mean, we need to avoid this scenario of go ahead and sprinkle a little AI dust on that, right? Like what is the actual business problem you’re trying to solve? You talked about making sure you have people who have domain expertise. You talked about making sure we have measurable results. You said there were four elements. What was the fourth hype?

Steve Zakur: Actually, I’m only on two right now. I’m coming up on three.

Tim Peter: Okay, good.

Steve Zakur: Because measurable results was part of the business problem-

Tim Peter: For one A, one B.

Steve Zakur: Well, you know, I’m just rambling here. You know how I get when I start to ramble. So three. Number three, or as my pal Bill says, C. Data science skills. Now you don’t have to be a data scientist, but you do need somebody on your team that has those skills, and more importantly not because they are going to be able to build what the vendor has or evaluate the vendor system architecture. They have to understand your data, because at the end of the day it has to be your data that feeds the beast.

Steve Zakur: Now that’s not to say that there are probably some models out there that cross industry, that might cross companies. But what we have found in doing this work is, and maybe it’s a B2B thing, I don’t know, because we haven’t done a lot if any B2C work. But what we found is these large enterprises have very unique data, and sometimes they can source some third-party data that helps the model be smarter, but by and large, the models are driven by what’s going on because their processes are unique, their systems are unique, their customers are unique, they’re product, everything about them is unique, right?

Steve Zakur: Large enterprise B2B, these billion dollar plus companies, they are unicorns, right? They’re just a field of unicorns. So they really are unique animals. And so I think the mistake that a lot of large companies are making today is they’re trying to become and build their own data science capabilities that allow them then to do AI application development. I’ve seen one company that does it well so far. I was talking to a large scientific products company, and it’s a pretty impressive kind of platform that they’ve built, an application layer they’ve built. Very impressive organization. The only one I’ve seen do it well so far. And so, yeah.

Tim Peter: Which would be fair. I mean this is a company that does science for a living.

Steve Zakur: Yes. Yes.

Tim Peter: So I mean realistically it’s part of their core competencies, so I think it makes a lot of sense.

Steve Zakur: Yeah, I think you’re right. They are definitely a very… And when I say technical, they’re a technical company, but not in the IT sense, but in the… Yeah, they’re scientists. They are definitely scientists, so it helps.

Tim Peter: Right.

Steve Zakur: And so they’ve done a nice job with it. But by and large, what companies should be focusing on, and you see this as well, I don’t want to say that you don’t see this, is focusing on the data. And a lot of it, you got to go back a couple of years, it was data links and blah, blah, blah, right? But it almost doesn’t matter what the structure is because there are a lot of modern technologies that can access the data regardless of its structure. It’s that you understand it, that you understand what the data means. It’s sourced well, that it’s aggregated and processed well, so you really have kind of a good robust ETL sort of capability, that when it’s used to predict outcomes, you can then understand the Genesis of all that.

Steve Zakur: And part of that’s because you want to understand it and trust it. And part of that’s also I think there’s a real hunger, especially amongst the non-technical, non-data science types, is to have things explained to them in terms that they understand. And so if you can say, “Oh well, remember how we used to do marketing campaigns, or email campaigns and we’d A/B test against the population, and then we’d get some stats back and then we’d make modification? Well that’s how AI works.” It just does it a lot faster, right? It does it a lot faster. And so that A/B test, you can do instead of one a week, you can do 50 a week, or 20 a day, right?

Tim Peter: Right. Right.

Steve Zakur: And so you can really increase the iterations. And so if you understand your domain and you understand your data well enough, you can explain the context of the business problem. Boy, there are the three points I’ve made so far. You can understand the context of the business problem in ways that kind of non-geeks will understand. And that’s important because those are the people who at the end of the day are going to allocate the resources, and define the targets, and do all the things that are going to make your life miserable if they don’t understand it, and you could be turning them into allies if you do.

Tim Peter: Yeah. Cool. So we got to have the right business problem. We got to have the right measurable result. We got to have the right data in place to make that work. What’s number four?

Steve Zakur: Expectations.

Tim Peter: Oh, interesting.

Steve Zakur: Yeah. Yeah. And this is really, I mean there’s a big umbrella around this. I could have called it change management. But there is an expectation with a lot of things that when IT gets deployed, then magic happens, and you see this. A couple of our clients have recently gone through website redeployments, and by redeployment I mean not only new design but new platforms.

Tim Peter: New platforms and all. Yeah, yeah, yeah.

Steve Zakur: Yeah. And I’ve been through a number of these, so I know that on June 30th when they deploy it doesn’t mean on July 1st it’s roses and sunshine, and you know. No.

Tim Peter: What? Surely not. Surely you…

Steve Zakur: Right. It’s just then like, “Oh damn. What did we do?” Right?

Tim Peter: Right.

Steve Zakur: It’s all the mess that comes from any large scale IT deployment. If by day 30 you have it running, you’re perfect, right?

Tim Peter: Yep. Yep.

Steve Zakur: That’s something I always had to coach business executives is like yeah. If we’re great on day 30, then we’re great. It was a perfect deployment, and it stretches out beyond day 30, then you might have a problem that you have to go pay attention to. And I think that the same thing is true of AI, that we are still in learning mode, and your vendors, right? Your vendors are still in learning mode. Everybody is in learning mode about how AI is going to ultimately help the business. And I think that any vendor product out there, including ours by the way, is an iteration on the final thing.

Tim Peter: Yep. Oh yeah.

Steve Zakur: And so that’s one point is that this is a journey much like, quite frankly, in 1997 we were all trying to figure out what is this Internet thing make, or mean for our business, means for our [crosstalk]? And so I think recognizing that it’s a journey and making sure you’re setting those expectations with the leadership team I think is really important. I also think that setting expectations about, what does happen on day one? Because on day one it’s really you begin the final test, right? In your A/B testing you’re thinking about what’s plan B, what’s plan C as you think about how to improve the technology over time. But the other expectations you have to set, not only about what business results are we going to get, and what the timeframe is for those, but what processes are going to be affected? Because that’s one of the things about all this data coming into your business. It’s going to help you discover things that you didn’t know about.

Tim Peter: Oh, yeah.

Steve Zakur: And it may change things in ways that you don’t understand. So for example, our Guidebox product, it allows any site manager to deploy it, and to deliver basically a modal kind of an overlay on the screen that helps guide visitors to an objective. Oh, by the way, some web McGuy can do that without asking all the product people. Okay. That might be a problem. So, that’s a business process change because now content is getting delivered to people without being managed in our visitor journey and going into some grid about persona. None of that happens because the machines doesn’t care about that.

Tim Peter: Right. Right. Nor should it.

Steve Zakur: No, it shouldn’t. And by the way, humans should get over that, right? The business model, how the organization works is going to have to adjust overtime.

Tim Peter: Well, this is something I talk about all the time. I mean no disrespect to the very fine work people have done on personas over the years, or the very fine work people have done on buyer’s journeys over the years. But the reason those exist is because we don’t actually know what the customer wants, and when we deploy something like GuideBox or our competitors or other tools that think about how you do this, the reality is it’s bypassing those because it’s actually getting to what does the customer want? Not how do we come up with a proxy that helps us understand what it is our customer might want.

Steve Zakur: Yeah, absolutely. Absolutely. And then so finally, so now you’ve got the stakeholders who own the business processes on board. You also want to know what resources need to be committed both during and after the deployment, because again, if it’s magic, well we don’t need any more budget for writers, right? No. You still need content. So you might need more content, right?

Tim Peter: Right. Yeah. Steve Zakur: Because you might find gaps, like for GuideBox, one of the interesting things and we’re just enabling this. It should be ready next couple of weeks with some wood to knock on. But it’s a dashboard, because one of the things that I’m actually most interested about GuideBox is not where it makes predictions, because that’s certainly important. I’m going to be really interested to see where it doesn’t make predictions. Where are the gaps and often that’s going to point to content gaps. So that’s one of those things that like, why aren’t I getting a recommendation on this page? It’s like, well because there’s no content to recommend. Oh. There’s a resource that has to be committed after the deployment to go find content to stick in that hole.

Steve Zakur: So there’s going to be a lot of interesting things I think people are going to discover as they measure the results. And that is kind of the final one is around expectations setting is does the business get 10% better tomorrow? It’s like, well no. A, our lead generation process takes 90 days, or the qualification, right? So…

Tim Peter: Right. Right.

Steve Zakur: It’s definitely not going to get better any earlier than 90 days. Maybe it’ll get better in 70 days, but that’s when you’re going to start to see the result. And, oh by the way, you’re not going to see 100% of the advantage. You’re going to see it kind of now begin to build, right? It’s like anything. It’s the expectations around measurements.

Steve Zakur: So to recap, I’ll just go through them because I got them here. So it’s the business problem, something worth solving. And this is where the measurements come in. Is it something you can measure, a measurable result? Two is domain expertise, because you want to make sure you’re able to evaluate what the vendor is telling you with some knowledge about how the world works. Three is the data science skills and it’s mostly focused on the data piece. Do you understand the inputs so that you can understand the outputs. And then the final piece is expectations. Number four is expectations. Are you managing those well within your organization so there’s reasonable expectations for when and how results are going to appear?

Tim Peter: Words of wisdom as always Steve. Good stuff.

Steve Zakur: There you go.

Tim Peter: Good stuff. Well, all right. Well we are at time, so as ever, really great chatting with you.

Steve Zakur: Yeah, you bet. Happy Monday.

Tim Peter: Happy Monday. Hope you have a great rest of the week, and look forward to catching you next time.

Steve Zakur: Thanks very much, Tim.

Tim Peter: All right, take care.

Tim Peter: Search Chat is brought to you by SoloSegment. SoloSegment is a technology company focused on site search analytics and AI driven content discovery to improve search results, increase customer satisfaction, and unlock revenue for your company. SoloSegment. Make your search smarter, and learn more at If you like what you’ve heard today, click on the subscribe links. You can find the on iTunes, Google Podcasts, Stitcher Radio, Spotify, or wherever fine podcasts can be found. On Twitter using the Twitter hand bill @SoloSegment, Or you can drop us an email at Again, that’s

For Search Chat, I’m Tim Peter. I hope you have a great rest of the week. Thanks so much for joining us, and we’ll look forward to chatting with you next time here on Search Chat. Until then, take care everybody.

This Post Has One Comment

Comments are closed.

Back To Top