#449 Smarter Ad Spend with AI: Zeke Camusio’s Attribution Playbook for Founders and CMOs

[00:00:00] 

Mehmet: Hello and welcome back to new episode of the CT O Show with meme today. I'm very pleased joining me from the US Zeke Camusio. Zeke, thank you very much for being here with me on the show today. The way I love to do it, as I was explaining to you, is [00:01:00] I keep it to my guests to introduce themselves. So tell us a bit more about your, your background, your journey, and what you're currently up to, and then we can start the discussion from there.

Mehmet: So the floor is used. 

Zeke: Of course, of course. Yeah. Thank you for having me. Um, yes, I've been an entrepreneur my, my whole life, uh, in the tech space. Um, I. I had a few different businesses. Last one was a marketing agency that was acquired in 2015. And you know, my background is in economics and statistics, so I always gravitated towards the, um, you know, the analytics and the data driven decision making aspect of it.

Zeke: Uh, helping my clients look at performance, understand. What's working, what isn't working, what to do about it, how to measure performance, and so on. Um, and yeah, four years ago I started Data Speaks, which is an AI platform that helps marketers [00:02:00] understand what's driving their conversions, revenue so they can invest in the right channels, campaigns, and make the most out of every dollar they invest in ads.

Mehmet: Great. And thank you again, Zeke, for being here with me today. So the first thing you know, which, uh, when, when you were mentioning about your background that, uh, uh, kind of attracted me to ask you, so you come from a background, economics and you know, of course you have the mix of data science and marketing also as well.

Mehmet: So what do you think is the biggest misconceptions that usually marketers have about like, uh. You know, anything which is data driven, of course. Machine learning, ai of course. 

Zeke: Yeah. Um, misconceptions. I, I think that, um, the, the main challenge, there are two main challenges that I see. You know, one, one of them is not being clear on [00:03:00] what to optimize for, you know, so, uh, there's a lot of metrics, you know, and some of them.

Zeke: Are are not really aligned with the objectives of the company. You know, so, so we could have, for example. Somebody writing your social media saying, you know, we doubled the number of likes or views, you know, and, and maybe that translates into conversions. Maybe it doesn't. Um, so I, I think that's one of the most important things that a business needs to do.

Zeke: You know, figure out what's our vision, what are the metrics that help us track? How much progress we're making towards our objectives and, um, and then communicate that to the different teams. So for example, if you are running social media, uh, you know, you know what it is that you should be optimizing for.

Zeke: And, and the reason for that is that I. Often, um, if you optimize for metric A, you're [00:04:00] gonna be, uh, maybe getting worse results for metrics B and C and D, you know, so it needs to be clear what it is that you optimize for. Are you gonna be, uh, trying to maximize your, uh, profit, your margins, your revenue? Um, you know that that's really important.

Zeke: And the, I think the second one has more to do with, uh, the. Building a habit around reviewing data and making data-driven decisions. You know, so you could have all the data in the world, but if you don't have a recurrent meeting, say every week or every month, where you look at the metrics that you already agreed upon, you're gonna be reviewing, um, and the, you know, look at what worked, what didn't work, uh, what, what we learned from it, and, you know, how are we gonna do things different moving forward?

Zeke: Um. You know, it is really important to have that because, uh, other, otherwise you're gonna have all this data flowing your way and, [00:05:00] you know, it's gonna be, uh, you know, basically useless. You know, if you, if you don't have a regular arc cadence to make decisions. Uh, so I think it's, it's more than misconceptions, I think a lot of the time is lack, lack of clarity on what it is that you need to be measuring and, and a lack of process to analyze that data on a regular basis.

Mehmet: So, Zeke, if you, and I know you when I was preparing, uh, for the, you talk a a lot about, you know, the attribution challenge and the gaps that, that exist over there. So from your perspective, uh, you mentioned some of the things now, but, uh, like if you can break down also like into more details, uh, these gaps, like, and do you know what is like the, you know.

Mehmet: One problem that you think, you know, it, it's, it's like, should be the top priority for marketers, uh, when it comes to the attribution. 

Zeke: Yeah. [00:06:00] So attribution is absolutely essential and, uh, what attribution means is that you wanna understand what marketing activities lead to what business outcomes. So let's say for example, you, you, you have a million dollars in your bank account at the end of the month and you.

Zeke: Ran, you know, let's say three channels, you know, LinkedIn ads, email, and um, let's say Google Ads, right? Um. You need to know how much of that million dollars, uh, was contributed to by each of those three players. And that happens all the time, right? Like we, we, we do a number of different things, then we get an outcome and, you know, we don't really know how much influence each one had.

Zeke: Therefore, it's impossible to, for us to determine whether we should be doing more of A or B or C. Um, so that's. A huge issue because, [00:07:00] uh, 54% of all the marketing budgets in the United States at least is towards, uh, online ads. Um, and, and as marketers, you know, we don't really have, uh, great tools to track their return on investment.

Zeke: So we are, we're investing more than half of our budget on something that we can track, and that's obviously a huge problem. Right. What are the options we have? One is we can look at the advertising platforms to see what they tell us, or we could use something like Google Analytics. Neither of those is really a, a, a solution.

Zeke: Why is that? Well, Google Analytics uses a. Last click attribution. So if somebody found you on, you know, say a, an Instagram ad, then they clicked on one of your emails, then they search for your brand on Google. Then they came to your website and purchased the. Google will give the [00:08:00] credit to that last click, right?

Zeke: Ignoring everything that happened prior to that, and it's gonna lead you to overinvest in that last channel. And underinvest in what creates customer awareness, which is, you know, the, the, the first few channels in the customer journey, excuse me, customer journey. Um, and, and it's very similar for, uh, advertising platforms where.

Zeke: Essentially each one is telling you, Hey, you know, somebody saw my ad, therefore I drove the sale. There are a couple of issues with that. In that case, uh, your email platform is gonna claim the sale, Facebook, uh, or Instagram in this case, uh, Google, and you don't, you didn't get three sales, you got one sale.

Zeke: You know, so these different tracking systems don't really talk to each other. Um, so as a result, as as, as, as marketers. Um, we, we, if we make decisions based on that data, [00:09:00] uh, and if. You add to that, the, the fact that, uh, you know, pixels, uh, you know, that, that are tracking conversions are blocked by, you know, a blockers about 40% of the time.

Zeke: Um, the fact that they don't talk to each other, the fact that, uh, they're, they're, they're measuring essentially whether or not somebody saw an ad, but not the incremental number of sales that you're getting from running ads. Um, all these issues make it so. One, you, uh, you know, you just don't trust your data because you know, it's, it's wrong and you just go with your gut.

Zeke: And, you know, that's obviously a, not a very good approach. Or even worse, you know, you have a full sense of confidence in the data that you used to make decisions, and then you don't get the results you expect. Um, so that's, that's in, in just the, the problem, you know, and the, um. The [00:10:00] reason why we exist is because we believe marketers deserve better and they deserve to make good decisions with confidence.

Mehmet: Now, the natural question that would come, Zeke, what's the solution for this? So how? How I, I get it. I'm spending a lot of money, different channels. Each platform is giving me different data and I don't know really, really what's happening, where I'm getting these clients from. So what's the solution? 

Zeke: Yeah, so essentially you have two possible solutions.

Zeke: One is called multi-touch attribution, and the other one is called media mix modeling. Uh, multi-touch attribution is where rather than using third party pixels, like, you know. The Facebook ads pixel, Google Ads Pixel, Google Analytics pixel, you use your own first party tracking. And what that that allows you to do is essentially say, you know, uh, [00:11:00] rather than getting blocked by ad blockers and privacy settings, 40% of the time, you know it's gonna work.

Zeke: You know, every time. Um, and the beauty of that is that, you know, it allows you to track individual, uh, points in the customer journey, you know, so that's really, really useful, uh, to understand, you know, people, for example, like find you on YouTube with this ad, and then they see this other ad, and then they do this, and then this and then this.

Zeke: So it's very good from a qualitative perspective to understand what the customer journey looks like. It's not, uh, useful in terms of understanding the return on investment that you're getting for, for each one of them. And why is that? Well, uh. We talked about Google Analytics and ad platforms using last click attribution.

Zeke: If you have a, an a multi-touch attribution solution, what it's gonna do is essentially it's gonna say, you know, [00:12:00] there's like three channels in this customer journey. Let's just split it, you know, in, in three, you know, 33% credit goes to the first one, same to the second one, same to the third one. Of course, that's, you know.

Zeke: From a, from a scientific perspective, that doesn't make any sense. It's arbitrary and, you know, it can never lead, uh, to understanding the actual impact of each player in the journey. Um, so with, with media mix modeling, I. Uh, you tackle that exact problem and the way it works is essentially it looks at the relationship between your activity and the output.

Zeke: So to give you an example, let's say that you double your, uh, Facebook ads budget and your sales stay exactly the same. Well, that's saying a lot about how little impact. Facebook ads has, if you can, you know, double it or even shut it down and your sales [00:13:00] stay the same. Now let's say that you, you double it and then you notice a substantial increase in the number of conversions.

Zeke: Okay, well, that, that's also saying a lot about the, the impact that has. Now if you look, look at that for, you know, the, the whole world for one day, um, that's really not enough for you to see patterns. But when you look at that data daily for each of your markets, uh, you learn a lot about what's going on.

Zeke: So, for example, if you, you know, double your, your spend for, say, the New York State and your sales for New York State increase, and then the next day you. Decrease your, your budget for California and then your, your sales in California decrease. Uh, by looking at, uh, say for example, your market is the United States.

Zeke: So you're looking at 50 daily observations and you know, over time that becomes, um, you know, tens of thousands of observations. So with machine [00:14:00] learning, you can get very, very good, uh, figuring out what you're gonna get for every dollar that you invest in each of your, uh, channels. And also understand, for example, that.

Zeke: Um, if you, uh, the return investment you get, for example, if you go from a thousand to 2000 a day or 10,000 a day, uh, it's not always gonna be the same. There's gonna be the diminishing return. So you, you wanna essentially figure out what's the optimal allocation of your existing budget, what is the best way to invest your next dollar and.

Zeke: After, uh, you, you invest a dollar, what about the next one and the next one, because different medias are gonna get saturated at different points. Uh, so by, by having that level of insight, you can make sure that every dollar that you put into your ads is producing the optimal return on investment. 

Mehmet: So Zeke, where is the [00:15:00] role of machine learning AI in all this?

Zeke: Yes. So, um.

Zeke: Um, if you were a really, really smart, uh, mathematician and, and had unlimited time, you could take these 50 daily observations for every single day and essentially figure out, uh, the, the math behind it and, and know the calculated, calculate the, the impact. Um, but you have to do that every day, right?

Zeke: Because your performance tomorrow is gonna be different than today, and that's gonna change. Even more a month from now, even more a year from now. So you have to like, spend all this time doing this yourself, right? With, with machine learning, you can automate all that. So, um, first you wanna automate all the data ingestions.

Zeke: So you wanna make sure that, you know, you're already connected to the APIs for Google Analytics, Shopify, uh, you know, all your marketing [00:16:00] tools are already being ingested. Um, these models. Are making predictions, but also they're checking against those predictions. So, for example, if I know that you're gonna get $5 for every, uh, dollar that you invest in Google ads and $2 for every dollar that you invest in Instagram ads and so on, uh, I, if you tell me how much you're gonna spend on each of those tomorrow, I should be able to predict what the revenue is gonna be.

Zeke: Uh, so that's, that's how machine learning works. You know, you. You make a prediction, you look at the, the, the actual results, and then you adjust those predictions moving forward. So it essentially, uh, you know, that's, uh, something that, you know, no human could possibly do. Uh, you know, because we don't have unlimited time and you just, we run the models.

Zeke: So, uh, as a marketer, all you do is sit down behind your computer, look at the results, and. Just know what, what decision. [00:17:00] Decision to make and have the right tools in place to measure results. 

Mehmet: Zeke? Now one question that might come to mind is there are like some. Concerns about, I mean, from marketer's perspective, of course I'm talking here, you know, all these crackdowns on the cookies and you know, the privacy, which is of course, like, don't get me wrong, it's it, it's also for the consumer benefit.

Mehmet: But for me as a marketer, like, you know, and you know, the ad you mentioned also the ad blockers, but everyone knows that the cookies are, you know, the holy grail of. Everything to collect, you know, what the user is doing, all the attributions. So how this is affecting, you know, collecting the data and, you know, how is also affecting the way we need to plan of, uh, on deploying these tools that allows us to collect data from different [00:18:00] places.

Zeke: Well, that's, that's, uh, yeah, that's, that's a really good point. Right. And, um, what's, what's interesting about this, right, is. I saw this coming five years ago, and that's, that's why we started Data Speaks because I, I was realizing that, uh, you know, before the internet, you know, we didn't have the, the tools to track individual customers, you know, so media mix modeling was, uh, the, the, the main way to do it.

Zeke: You know, so you had radio, you had tv, you had print, um, you know, with the internet that, you know, we. Started believing that by tracking individual users we will be able to reach the social truth. Um, even when there before ad blockers or before the, there was anything that could prevent us from tracking individual users.

Zeke: Um, I. We still have the issue that even with perfect tracking, multiple platforms are taking credit for the same [00:19:00] sales. Uh, it's based on correlation, not causation. And what I mean by that is that if you get a hundred customers a day and then you hire a promoter to stand outside your shop, you know, just with a sign, uh, and now you get 120 customers a day, the the, the person will tell you that.

Zeke: 120 people saw the sign in reality that got you 20 incremental customers, you know? And so that tracking is not based on incrementality, is based on exposure to, to ads. And the the, um, we can't infer that just because somebody saw an ad, that's what drove him to, to buy and maybe they were gonna buy anyway.

Zeke: Uh, so because, um. That has made it, uh, you know, very, very clear, um, OO over the years that we can't, uh, it's becoming more [00:20:00] and more difficult to track individual users. I, uh, that's what happened like five years ago. I, I, I saw that on one one hand. You had users demanding more privacy, more, uh, tools in place preventing this.

Zeke: Technologies to work. And then on the other hand, you had all these solutions that were based on this ability to track individual users across the web, across devices, and so on. Um, it, it, it, it was very clear to me at the, uh, at that moment that, um, you couldn't get to the, to your truth. Uh, when you're fighting against the world and users are saying, you know, you, I don't want you to track me indepen, uh, individually now.

Zeke: That's the beauty of media mixed modeling is that you don't really need to know, uh, the, the, the conversion for each individual user to get to the truth. Essentially, what you, what you, what you are measuring is dollars spent, uh, on different, uh, [00:21:00] advertising campaigns and so on. Comes. And so if you invest more and your sales increase, you don't need to know what customers purchase because you, uh, you spend more, you just, you just need to know that you did this and that this happened.

Zeke: Um, and, and, and that's why, uh, in some cases, if you, what you wanna know is the ROI in, in Impact for every channel. Media mix modeling is, you know, without a doubt, the first, the best, uh, approach. Now, if you do wanna understand individual customers, MTA is very good for that, and in some cases we deploy both, you know, so, uh, MTA makes or multi multitouch attribution makes sense when, for example, if you're B2B.

Zeke: Uh, and you wanna know, you know, let's say you're about to call a prospect and you wanna know, for example, that they came to your website, they downloaded this white paper, they saw the pricing page. In that case, I wanna know that it's you who did [00:22:00] that. Um, and you know, you might have that data for some users, not for other users.

Zeke: But, you know, it's, it's important qualitative data, but it doesn't help measure impact. Uh, so that's, that's why it, you know, we say. Um, for sure you need medium mix modeling. Uh, if you're investing $25,000 or more, um, per month on advertising, it's a no brainer. You need to have it. Um, if you also wanna know individual customers, also you need MTA.

Zeke: Uh, but it's, it is, it is good to know that medium mix modeling works, even if pixels don't. 

Mehmet: Zeke. Now the question, because you're mentioning about the budgets and you know the money, so maybe it's like two questions in one I'm gonna ask you. Um, especially if they are like kind of a small business, maybe still a startup.

Mehmet: I. I gotta say first, like we don't have data scientists to do all this and you know, is it going [00:23:00] to cost us, as they say in English, uh, arm and leg to do this? Like, is it something which is very expensive to, to have such system in place? What you can answer maybe business owners or founders who might be thinking, uh, this way.

Zeke: Yeah, so you, you really have, uh, two paths that you can take. Uh, you can build your own solution or hire somebody to do it for you. Uh, if you're gonna build your own solution, you have to budget at least half a million dollars per year. You're gonna need, uh, one data scientist, one machine learning engineer.

Zeke: Probably a data engineer, uh, infrastructure servers, ETL tools, um, you know, a number of different things. Now, this is, this is a, a really good option if you, uh, if you are enterprise and, um, uh. Wanna do this in-house, you have maybe some concerns about sharing your data with third parties. Uh, so this is a really, really valuable [00:24:00] option.

Zeke: And, uh, if you're looking for someone who is an expert in patient statistics that has worked with, uh, you know, MCMC. Models, for example. That's, that's a really good profile to find. Uh, but yeah, of course it's, you have to decide whether or not that's your core competency and, and you wanna have that as, uh, you know, one more thing on your plate.

Zeke: Um, I. Or you could, you could essentially hire a, a team that does this for you. Um, you know, of course we do this. You know, that's our, our company Data speaks. Uh, that's every, you know, all we do, right? We are, we're a team of data scientists from Stanford and M-A-M-A-T that focus on attribution, uh, to help marketers make these sort of decisions.

Zeke: But we're not the only ones, you know. So, uh, recast is another really good company that I, you know, I. Uh, I can recommend, um, north Beam is, does, does a really good [00:25:00] job. There's another one called Measure that you know, is kind of newer to. Media mix modeling, but they've been around for a while doing incrementality testing, so you have options.

Zeke: Um, and what we, what I noticed about all these companies in, uh, before we got in the market is that, um, you need to have at least a hundred thousand dollars a year, um, to invest. You know, that that's what, what, what's gonna cost you? And that's a, a really good option for, for a lot of companies. But I also realized that, um, you know, this could be done in a much more cost effective way, making it.

Zeke: Uh, available to smaller businesses, you know, which is why we, uh, entered the market in the, you know, this like fi four, four or five years ago. Um, you know, so we wanted to bring a solution that was, that in the past was only available to the, you know, the big [00:26:00] guys to, to the rest of us, and. What's interesting is that, you know, we are now working with a lot of the, a lot of, you know, big companies and, and, you know, servicing both smaller businesses and medium businesses alike.

Mehmet: This is really interesting. Now, one question which I started to ask recently, especially if I'm talking about, uh, marketing, right? So one. Thing, I'm not sure how Zig can be tracked. So nowadays we are starting to see kind of a trend, let's call it, of people instead of going to the traditional places to find things like Google or like any other search engine.

Mehmet: So they are going to the LLMs, right? So they're going to charge GPT, they're going to cloud. Claude maybe doesn't have the internet connectivity yet at the time of recording. Um. So how [00:27:00] can we track that? You know, people are coming to my, to my website because they found about me, uh, maybe using perplexity or maybe they are using, uh, chat, GPT, that and now growth of course.

Mehmet: So. Is there a way to track this? 

Zeke: Yeah. So we, what we do is we, we model, uh, the, the first aspect of our modeling is we, we wanna differentiate between what's advertising and everything else, right? So everything else could include, uh, AI searches, search engine optimization. Maybe you are going to trade shows, maybe you have, uh, you're doing content.

Zeke: All that is, um, you know, doesn't have a, a cost associated with that. Um, we call that baseline. And essentially what that means is if you didn't do any ads. What, what would that be? Right? Like, what, what would that look like? What, what is the [00:28:00] number of conversions a day that you, you have without doing any ads?

Zeke: Uh, and then we can measure the, you know, what's the incremental, uh, number of conversions or the amount of revenue that you get, uh, as you increase, you know, the, your, your investment for all these different campaigns that, that you're running now. Uh, though we do it that way because. Um, it's. The really, the, the, the, the, the, the most important consideration you need to make as a marketer is, you know, I have for example, like a hundred dollars.

Zeke: You know, what's the optimal way to invest that? You know, I can put 50 here, 20 here, you know, like, I wanna make sure that I'm getting the, the, the, the most optimal, uh, outcome. From, from that investment. Um, but at the same time, we, you know, we wanna, uh, model all the other things that are, are organic, you know, so I'll tell you that.

Zeke: Um, one thing [00:29:00] that you, you, you don't wanna do, or, I mean, maybe you wanna do it, but, you know, just take it with a grain of salt is, you know, you can't ask people, right? Like, where do you find us? Because, um, they might just tell you, you know, if they went through, you know, five different touch points in their journey, they might tell you maybe they're, you know, number two, um, you know, maybe, yeah, they saw an ad, but then, you know, they, they listen to a podcast or whatever, right?

Zeke: So you don't know all of that. Um, so even as a consumer, you don't remember, you know, like what influenced your purchase decisions, you know? Um, so. There are a couple things that you could, you could do. You know, you could, you could, you, you could ask that question. You could use Google Analytics for, uh, last click tracking.

Zeke: And by the way, data speaks, takes all the data in, right? Google Analytics data, uh, platform data, all that comes into the, into the platform. Uh, but then we work from your source of truth and we say, you know, this is the [00:30:00] total number of conversions that you got. Uh, rather than letting each platform that doesn't have access to my data, figure it out.

Zeke: We do it the other way around. We start from the source of truth. Um, so, um. I think that there, there's value in anecdotal evidence and asking your, your, your, your clients, especially if you sell, you know, say a hundred thousand dollars projects rather than a $50, uh, gadget, right? So it makes sense to ask people like where they found you, but just know that they even, they might be unaware of that, you know?

Zeke: So the, um, the, the world is very, uh, complex and. No model can track a hundred percent of all the things that could positively influence a purchase decision. Uh, the, the one problem that we solve is helping you make sure that you get the most from your advertising. 

Mehmet: Great. Now if I want to ask you, because you said like how you predicted five years ago.

Mehmet: So if [00:31:00] I want to ask you to predict the next five years in this domain, Zeke, like five years is a long time in my opinion nowadays, like a lot of things can happen, but I mean, in the future, what trends, uh, are you seeing and which ones are like maybe the most, uh, excited about? 

Zeke: Yeah, yeah, of course. Uh, so I think that, you know, with.

Zeke: LMS being so mainstream these days, um, it, it is really gonna change. My prediction is that that's gonna change the world more than the internet did. And what I mean by that is if you have, uh, agents that. Have access to all your data, you know, every email you've ever sent, your company's internal, uh, documents, your, your blog, uh, any client, uh, emails that you've ever sent, access to your calendar and so on.

Zeke: Uh, it's essentially a replication of your brain. And not only that, but you know, you can fine tune [00:32:00] agents to think the way you do, you know, so, um. That's, that's gonna be, um, that's gonna be in my view, how, you know, AI becomes mainstream. You know, you have this, this foundation, which is the LLMs and then, uh, individual, uh, agents that are fine tuned, uh, for the specific things that you wanna accomplish with access to your own particular data.

Zeke: So not only are are, are they trained, you know, with all the data in the world, uh, but you say. You know, your job is to do this. Um, you have access to everything you need. You, you know how I, I work and every time there's an output, you can fine tune that. So you could say, you know, next time somebody asks about this, don't answer this way, answer this way.

Zeke: So I. Very much like, like, like training an employee, uh, but much, much more effective because you can replicate that multiple times. Uh, it will logs also have access to [00:33:00] a number of different tools. So you can send emails, uh, to, to people, you know. So one of the things we're building right now is, uh, if, uh, the, the performance.

Zeke: For a given metric is below a certain threshold, you can send a Slack message to, to your agency, letting them know, right. So, um, they'll be able to, you know, send emails to searches, use the web. Um, so I think that the vast majority of things that you can write into a document, uh, for an employee and say, do this, do this, do this, do this, and AI will be able to do, you know, so that's, that's really exciting for me.

Zeke: I think that the. What, what I'm most excited about is that, um, right now, um, you, you know, all the knowledge that we have in our brains as data scientists will be able to be downloaded into a system that will be able to interpret data, answer, answer questions, make recommendations, um, [00:34:00] and, um. Essentially replace user interfaces with, with the chat bot.

Zeke: Right. So, so I, I see a couple different scenarios. You know, one is, um, if something goes wrong, let me know, right? Like, like, you know, have the. Um, have the knowledge to identify what, what things are high priority, low priority, medium priority. Based on that, you know, if something really bad happens, maybe give me a call.

Zeke: If something, you know, low priority happens, maybe put it in my project management and, you know, assign it to somebody else. Uh, so the ability to evaluate, uh, issues and take the appropriate action is, is, is huge. Of course, you know, the, the one we all know is the, you know, the, the chat bot where you can ask questions and get answers.

Zeke: Um, uh, but yeah, I think that, um, I don't think we've seen any [00:35:00] technology, uh, in human history that will change the way we do things as much as ai. 

Mehmet: Absolutely. And at that fast pace, I would say like, uh, I was, you know, maybe I'm repeating myself, but the other day I was telling someone like, like I remember when I was still kind of kid, let's say when the internet came.

Mehmet: So, but yeah, like things we used to see another breakthrough maybe after six, seven months, maybe one year, sometimes even when the smartphones and you know, like the apps and all this. So like, yeah, there's a new phone every year, right. Now with ai, it's for the first time. Like it's sometimes every day like I sleep, I wake up and I see some something new.

Mehmet: Right. Which is fascinating indeed. Um, Dick, as we are almost coming to an end final, I would say advice maybe you want to give to, uh, entrepreneurs and maybe also for. People who are leaders in, in, in marketing and of course like where people can get in touch and know [00:36:00] more about you and of course about data speaks.

Zeke: Yeah, of course. Uh, so I, I'll say this, you know, um, data is power. Uh, the, the, the, the more data you have potentially the more power you have to make better decisions. And the problem that we have is that we have way more data than we can process, right? So, uh. For us, you know, our, on average, our clients are, are processing like 30 million data points every day.

Zeke: There's no way a human can go through that spot problems, you know, like, there's just not enough, uh, hours in the day. So just know that, uh, their machines are your best friends when, when they come to that, you know, so if you already are the kind of. Organization that looks at some data, uh, on a regular basis, could be financial data, could be marketing data.

Zeke: Uh, you already believe in the value of making informed decisions and having the tools in place to track. [00:37:00] Whether your decisions get the results you expect or not, um, but you are overwhelmed with the amount of data you have. Um, and, you know, you're just not making enough data-driven decisions just because it's, you know, it's overwhelming.

Zeke: There's so much of it, and, and it's impossible. Uh, just know that. You know, there's, there's a better way to approach that and, um, uh, more specifically when it comes to making sure that you get the, the, the highest return on investment. Um, the way to think about what we do at Data Speaks is if you have a hundred dollars to invest, you know, right now maybe you have 40 here, 20 here, 25 here, and so on.

Zeke: That's never the most optimal way to do that. Uh, there's, there's, there's, uh, we can reconfigure that, the budget reallocated in a way that. We don't change the total, it's still a hundred dollars, but we just allocate in a, in a more optimal way. And what we found is that usually, [00:38:00] uh, leads to 20% or more increased, uh, efficiency.

Zeke: So the, the way most most companies approach this problem when they wanna grow is they throw more money. You know, like, like that works, right? Like you can double your budget. You something good is gonna happen, but it's gonna cost you a lot, right? So we wanna make sure that, you know, you can grow, but you can also be very efficient with every dollar that you put into it.

Zeke: So that's the problem if have, if you're investing in 25 KA month or more on ads, uh, you can learn more about at Data Speaks ai or find me on LinkedIn. Um, I'm sure we can put the link in the show notes. 

Mehmet: Absolutely, Zeke. And this is what I always tell my guests, so, uh, the audience, they don't have to look around.

Mehmet: So if you are listening on your favorite podcasting app, you'll find the links in the show notes. If you're watching this on YouTube, you'll find this of course, in the description. Zeke, I can't thank you enough for all these insights. I think [00:39:00] this is something very interesting for any business, whether they are B2C, B2B, uh.

Mehmet: To understand to and what you said, that that is very, very important. Like we have a lot of data that we can act on, but we don't have the manpower to do that. So, uh, I, I would say that this is the most important takeaway today. And of course, people should reach out and speak to you and to your team about, you know, how you can help them.

Mehmet: And thank you all for all the insightful, uh, also information that you shared with us today. So. Really, I appreciate that. And you know, this is for the audience. This is how I end my episodes usually. If you are new to the show, thank you for passing by. I hope you enjoyed it. If you did, so please give us a T thumb up, subscribe and share to your friends and colleagues.

Mehmet: And if you are one of the people who keeps coming again and again. Thank you for being loyal to us, and thank you for making us making it to the top 200 charts in multiple countries this year for the first time since I started the show. [00:40:00] And also thank you for putting the CTO show as one of the best ranked, most listened business podcast in Dubai.

Mehmet: It's in the top 40. Thank you for making us, making it made it. And as I say, always stay tuned for a new episode very soon. Thank you. Bye-bye. 

Zeke: Thank you.