(888) 815-0802The Intersection of Technology, Creativity, and Consumer Behavior, with Ricardo Olivo [Episode 1155]
Our latest podcast episode features Ricardo Olivo, Vice President of Technology at VML, along with host Alastair Woolcock for an engaging discussion around the integration of AI in understanding and shaping consumer behavior. The episode explores the shift from traditional marketing to a more data-driven, personalized approach, stressing the importance of aligning creative content with emerging trends and legal considerations. The conversation emphasizes the necessity of maintaining a human-centric approach amidst rapid technological advancements, offering a glimpse into the future of marketing where creativity and technology converge to enhance consumer engagement.
Podcast Transcript:
0:00-0:29
Speaker
Welcome back everybody to this week’s sales strategy podcast. I’m Alastair Woolcock CSRO, and I’m excited to have with me today, a special guest, Ricardo Olivo, the Vice President of Technology from VML. Ricardo, how are you? I’m doing well. Thank you. Thank you for inviting me to your podcast. I’m really excited. A lot of things to talk about. So yeah, thank you.
0:30-1:14
Speaker
The one. The elephant in the room, Ricardo, right out of the box is the blockbuster acquisition of VML of Wunderman Thompson. And I just want to give some context to the audience here of who you are and VML as a company. That acquisition now means that VML consists of approximately 30,000 employees working in 64 global markets. John Cook, the CEO has brought this all together and has now created the biggest creative commerce and customer experience agency in the world. It is a big one and you’ve been living it front and center. The last say it’s been about 90 days so far, roughly.
1:15-1:39
Speaker
Yes. Yes. It has been quite a change. The general announcement was recently but we’ve been working on this probably a good chunk of last year as well just getting to know all the teams getting to be fully integrated on things because we were overlapping a lot in our portfolios and our capabilities So it was it has been and it still is quite a journey, but it’s It has been great.
1:40-4:40
Speaker
And for all those fans of Mad Men TV show out there, while there may not be cigars and debauchery and all of those things that exist in the world of marketing agencies anymore, it is the world, you just happen to be Vice President of Technology, working with firms like Coca Cola, U. S. Marine Corps, AT&T, Turner, Porsche like big global brands.And you’re helping them with the emerging technology because it’s changing. And it’s changing a lot in terms of how consumer behavior is there. And Ricardo, I just want to start, we’d love to start with some news and beyond the excitement of that merger. And now what you guys are creating a VML, but I just want to get your perspective on emerging tech in light of three big news events and trends that we’re seeing in 2024.
The first is the CMOs are actually for the first time this year expected to preside or prioritize what we call practical aspects of the market instead of, these ultimate unique brand experiences. There’s a lot more prioritization again on, Hey, I’m a Coca Cola. I’m a Porsche. The product quality, the cost effectiveness, the value pieces, not just building like Hermes into a big contentious billionaire brand, right?
So very practical stuff. That’s number one, which is different than what we’ve seen in the last few years. Two, consumer behavior has shifted really heavily on social justice and ESG programs. We’re now seeing that 10 percent of U. S. consumers indicate the social policies of companies is actually what drives their behavior.
And only 12.9 percent say that the environmental policies impact their buying decisions. As a result, nine of 10 mainstream brands are now saying we’re de-prioritizing social justice and ESG and how we communicate and how we work with our consumers anymore. Now those big changes are being disrupted by exactly what you’re doing at VML, generative AI front and center. 20 percent of CMOs now are required when you go read job descriptions worldwide. 20 percent are now required to have skills with generative AI technologies. We know it can write what else are you seeing it can do. And this year is going to be the year of Martechs finally consolidating with the broader GTM, the data first and third party data sets coming together, the engagement side coming together.
Obviously, Wunderman Thompson was one of the leaders in that side, and that data is now colliding very front and center with the creative world on there. So with all that news in mind, Ricardo, you’re on the front end of this. You’re in the driver’s seat, you’re the biggest company in the world, helping these brands figure out how to connect with consumers next.
What are you seeing in the emerging tech? What do you think of those trends?
4:41-7:42
Speaker
Yeah. None of that is new to us to tell you the truth. This is the news that are coming out today are things that have been benefiting us for quite a bit of time. When you are in a, in the world of emerging tech specifically, you gotta read the trends.
It’s all about the human experience. Right now it’s all about learning more about your consumers beyond just the click or the tap on your mobile device. It’s beyond that. It’s to understand what triggers them. What kind of messaging they are accepting. It’s a lot about measurement, right?
And, a lot of things that happened in the last few years to the world that has, have actually shifted those behaviors and we have to adjust. Things that you were, that you planned, let’s say, three or four years ago that would be successful today might not be. And it, it might have to die, right?
A lot of this emerging technology areas, a lot of the startups, it started with a mindset of a long-term goals that had to be adjusted because of all these things that have happened. At the same time, there’s a lot of new legislation and policies that are coming in place. Countries are looking at emerging technologies as as enablers to larger communications, larger communities. So legislation comes on top of that and everything gets into a different place. So right now for us, as we serve as big clients like Coke or we do things with U.S. Marine Corps and all this, big companies that kind of like disrupt areas where they live.
We have to be careful about how do we deliver messaging to consumers? We have to make sure that we are aligned with their likes and their needs. If not, we just, we don’t hit the numbers. So it’s a lot of, in the world of CX. In emerging technologies within CX, we lean heavily into machine learning, the machine learning aspect of analytics and analyzing data to make sure that we’re creating these decisions from a knowledgeable standpoint.
So when you talk about generative AI, to me, it’s a little bit more of their keywords that become popular based on the foundational work that has been done in the past. When you talk about generative AI, my mind thinks about all the steps throughout machine learning that we had to go to actually create this type of outputs, right?
But all the knowledge we’ve acquired. From that process is the one that comes out now as tangible outputs to consumers like generative AI, exactly like that, but there’s a lot of foundational knowledge that comes out of it that allows us to forecast what’s going to happen next based on the change and the shifts of all of this.
So legislation, knowledge, evolution of markets new ways, even new shifts in culture that happen all the time that allows us to do that. And. Yet again, we need to use AI and machine learning to understand that in a more automatic way and then make decisions on top of that. Basically, I’ve said a lot about it, but that’s pretty much my point of view.
7:43-8:42
Speaker
I am curious when you think of let’s pick on a little bit of the machine learning side, like it’s very interesting to me to bring on machine learning quite a bit and the legislative side, right? Generative AI is amazing because of course it can suddenly create content and it can speed up the whole velocity and all of that.
But how are you dealing with the desire for your output, which a lot of companies have right now versus what I hear you saying is the customer experience, the CX side is really important. So how are you helping a firm like Coca Cola be different from Pepsi? Because generative AI can really close the gap very quickly between the two. I get there’s the creatives that are going to build the next look and then the great content that might be neat, but how are you using this to inform that customer experience in a unique way that’s going to make Coke stand out from Pepsi?
8:43-11:38
Speaker
Yeah. Yeah, definitely. I do see the use of generative AI as a way of, and probably I’m butchering the whole world of generative AI, but the way we use it as a baseline of inspiration, right? Right now, when you talk about legislation, it’s what makes us look into generative AI as inspiration, not as a decision making factory. This is inspiration for our creative approach. For example, we have brands that have done a cultural analysis of their consumers, like Sprite or Fanta, and they have seen that there is a clear connection between culture to the brand, right? Now we use generative AI to enhance the baseline benchmark on how do we start being creatives, right? So we go into large language models or we go into even visual gen, as a generation to provide a baseline.
To the creative work of CX. CX is beyond visuals, right? CX is more about how do I connect to the user, with the consumer. And in that journey, we use generative AI to query the right parameters based on historical knowledge to get us to this point. So if we say today, there’s a campaign that talks about uniqueness in culture and what is this thing that they’re talking about now? I am unique in the world, but I go against the rules. For example, there’s a brand that is going around that and we link those two points and motivations, right? I’m unique, but then I don’t like the rules. And then we use generative AI to come up with a handful of examples, visual and multiple formats to give us a baseline.
Okay. What does this mean? And then we retrain some of those models, for example, adjusting from using human input, right? What the brand wants to do, how do they want to achieve that? And we adjust. And then we also go within the guardrails of legislation to say, okay, can this be done? Is this doable?
This approach is right. There, there was this fashion brand that just went I don’t want to say the name, but it went raw on. Let’s use our imagination to go against the status quo, but, and they failed and it got hit and there were it took a few years to do it just because they look into the culture, they wanted to make decisions, they did use generative AI and then fail because they did not readjust, right?
They just went as risky as possible and it actually hit the brand. That’s what we have to be careful with this big brand because the statement of generative AI, if you let it loose, it could be pretty powerful, but it could also be powerful in the wrong ways. It could give you the wrong messaging, it could be, give you all of that, and that’s how you use it as inspirational baseline today, at least that’s the way we do it right now.
11:39-13:18
Speaker
I am a big fan of the approach because as you think about artificial intelligence technologies this concept of human in the loop. It’s really important, right? We all are, it’s easy to put out a news headline. Oh, it’s going to replace this job and that job and so forth. But if history is taught us one thing, most technological revolutions and emerging technologies tend to augment a function and then the function improves.
It can change and down the road, but it is that augmentation and improvement. Then the displacement happens after that. And we’re in the, we’re the early stages of this. And I love the multiplication effect of how VML’s thinking about generative in the sense of you’re absolutely using it, but it’s fast testing and creating, I would be curious on the amount.
Is it four times what you did before? 10 times the amount you did before, which is a big piece of the creative process, right? The more examples you get, the more you’re like, I hadn’t thought of it that way. That’s an interesting piece but you’re not letting it go completely all the way the human in the loop is saying.
I asked, but I happen to know this about the customer experience side of the brand loyalty of Fanta. And while that is really provocative and neat, I’m not going to, we’re not going to go all the way there yet. We’re going to, we’re going to move it. So it sounds like you’re really using this to, be that, call it the 10 X multiplier internally. You don’t quite see it yet as the. It’s the Nevada that is going to be, Dall-E is suddenly going to be creating all imagery on billboards going forwards, for instance.
13:19-16:21
Speaker
Yeah, actually, I can tell you that there are some executives from actually the, some of our clients that are all in for that, and we are the ones that are telling them, Oh, hold your horses for a second. Cause you know, we can’t get there. There’s a lot of people, very enthusiastic and very, influential that want to get there, which is great and exciting because that’s what you get your funding to actually get things like that. But we got to be cautious about how we do this in getting the human in the loop is very important for multiple ways, right?
I don’t want to get into the politics of it, but I can give you an example. As we go into government entities that we service, we have. We have the foundation, foundational beliefs and within those foundational beliefs, we have this human in the loop as one of the core beliefs, right? We always have to make sure that humans are the drivers behind the wheel all the time, right?
When you want to create an effective campaign and you want to use generative AI, you need the human input and the human knowledge to actually be more effective, right? you could definitely develop Machine learning algorithms to replace, my point of view, that should not be the motivator, it should be to enhance, right?
It’s how can we put these people on, instead of driving a little, 80 horsepower, little car, how we put these people in, 500 horsepower. Big machines who actually go as fast and have better performance and have more efficiency. That’s what we’re trying to do with these technologies today.
Yet again, we also take this into outside of the government and we go into commercial. The motivations are a little bit different, but we have to still keep the human factor in the loop because there is some sort of moderation that needs to happen. Because if you, if we let it lose again, it could be, it could affect it in a negative way.
So I think that it’s more of the, I have an example here that I always use, it’s funny. It’s in all the shades of gray, how do you know what’s the right shade? And that’s, am I going to use a little bit of the internal terms that we have here? We have the influential biases.
The non conclusive biases of taste, for example, you say, when something is good enough to be good, right? You can train things to get you to a place, but what is the criteria behind what good is? And that’s where the human factor comes in and says, I, in my own bias of my environment of what I know and how, what I learn, what I teach.
I decide that this is good enough and that’s what we present to the client, right? So it’s not about a machine doing it and putting it in front. It’s some, it’s somebody making a decision that has their own cultural in circumstances and professional circumstances that decides what is the right shade of gray? You know what I mean? That’s basically where we sit on that point.
16:22-18:18
Speaker
Do you see as you’re in the emerging tech side of this? When I think of these trainable machine learning models, trainable AI models now and self learning ones and if I’m sitting out with a big global brand or the government, a lot of it is, I need to take what are the, to your example of the shades of gray, what are the, what’s the shade of gray formula that you have already as part of your company ethos, your thinking, your understanding.
And actually what I want to do is take that information. Those visual assets, those brand identities. I want to take both the first and third party data from the consumer buying behaviors and what they’re doing. I want to use all of those things and more. I want to pull that in. Into essentially a singular model, but actually what I would frame is actually create the master model, not the apprentice model, because that master model is what I can then use to create good foundational elements that I think can augment with my human in the loop where I think a lot of people are currently going wrong is they’re not actually trying to build a master model, they’re actually trying to build an apprentice model that’s very wide.
So by doing, wide you actually get a lot more generic, you get a lot more risk. If I actually put tighter tightrope rails on it and they say, no, I want a master model. I want to help Porsche connect in this way against these attributes, this thing, I’m going to make a really master model. It’s going to be very specific, hyper specific brand and marketing customer experience, output suggestions. Then I’m going to put my creative team on, my VML team on and expand from. Is that how you’re seeing the clients think about this? Or are they doing much more kind of what a lot of folks are right now? Is that a bit of the more apprentice model, very wide based usage?
18:19-20:57
Speaker
Yeah, that’s a good question. And I can tell you, I don’t really have a final conclusion of either or As I see the master model being effective towards, an out, getting to an output faster, I might say, right? It’s helping you understand what the guardrails are in, in a more performant way.
I also believe that as a creative agency, for example, we’re servicing moving targets, we’re servicing cultures that shift, we’re servicing consumers that change with Times and with the influence of other things. So master model, sometimes, I do understand why people go and build the apprentice models, because you gotta be ready to react and pivot. And with a master model, the amount of investment that you put there sometimes might not justify the output because in five minutes. You got to retrain the master model to understand new rules and new ways of working. I can already tell you something like when you are servicing globally, you’re talking about different regions that have different cultural biases and also have different ways of living, and with that, if you go with a master model, what’s effective in one region might not be any other. Now, if you want to create a master model that encompasses all of it, it’s more about the amount of investment that needs to happen to go with it. But now let’s go on the tactical side of things.
When we talk about a company that does something, carbonated beverages, they, their amount of investment on their own product. It’s what takes prevalence over investments of other things. So how much money would you spend to create a master model that understands a global culture, right? For example, versus taking that money, investing it in R&D for your product. And that’s like how it goes, right? Our budgets of MarTech are never compared with the budgets to, improve a product or release a new taste or something like that. That ephemeral investment that we get here and there, it pushes us to just serve as the moment in the right moment.
But what I envision, I guess is more my hope is that one day with all this data and all this apprentice models, we can assemble something that then we can tap into the apprentice side of it, but also elevate it into a master model when you need it. See, but still that’s a work in progress.
20:58-21:53
Speaker
I think it’s, I think it’s prudent advice for any CMO listening in around the world today on there, in the sense of, I’ll attempt to paraphrase what your point of view is here, right? Focus on the speed over the large investment costs right now as the technology lands is shifting so quickly.
That you’re better off fast fail, move forwards, keep the human in the loop to augment the speed of change until we can get to enough point proof points where it makes more and more sense to build bigger and bigger master models and consolidation of the VARTEX stacks into the GTM stacks, et cetera. I want to one final question, because you did bring it up a few times and we haven’t touched on it yet, legal. You brought this up a few times. What should we be looking for? And what are you seeing as it presents, legal or compliance challenges as it pertains to emerging tech and AI?
21:54-24:35
Speaker
Yeah I’m right in the middle of that, as we are servicing a lot of government initiatives and non government initiatives that are regulated in different parts of the world. The main obstruction that we find right now is misalignment. Misalignment of. How far you can take the abilities of AI in what means, right and wrong from different perspectives. Again, yet this is falling again into the same thing. Every region is different.
Every culture is different. Every legislation is different and seizes in a different way. So we’re taking advantage of legislations that are a little bit Better flushed out, right? We are able to deploy better programs using generative AI, for example, in regions that have better understanding of the technology and what it can do. North America being one of the leaders there. And what I would say is that I’m looking forward to a place where legislation does not interfere with creativity. One of the biggest things that we have right now is that we have this amazing, access to open source and non open source wealth around AI, but we have to limit ourselves because of that is, what, how do you control things from IP or not stealing one thing from the other, right?
We just have to arrive into a place where AI can be embraced and not Yeah. Look at it as the bad cop kind of thing that the bad agent because when you use it for inspiration it’s amazing. When you use it to have superpowers, AI is amazing. But then when you put it through the funnel of politics, all the things that the amazing side of it get diminished to probably zero, right?
You’re not able to execute your campaigns. You’re not able to output your creations because of this. I understand very little about the legal intricacies of, how you legislate in multiple areas and multiple countries and things like that, but I do understand the impact of that. And some of the amazing development that we’ve done internally on the use of generative AI or the use of machine learning on its own have had to shift quite a bit for us to just abide to these rules. Rules are good, but you also have to evolve them with times. And also you have to understand in which environments those rules are getting applied, right?
If you’re crippling creativity, that should not be allowed. You see what I mean? Pretty much that’s what it is. I dream about a time where at least talking about North America, where we all understand that the use of AI could be an amazing source of knowledge and inspiration. And we should have rules around it to embrace it and not ban it.
24:36-26:58
Speaker
I will wrap us up with this thought on the legal side, and it’s we can talk more about this potentially in the future, Ricardo, but I do remember myself and many colleagues when I was at Gartner, we were looking at what we called the rise of citizen data for sale and techno nationalism. And there, what we were looking at is, you basically, what’s the driver between accessibility of consumer information and data that drives the efficacy of say, artificial intelligence and other data driven models versus the privacy pieces.
And what actually was interesting is you, if you, we looked at, we actually removed country names that looked more in terms of economic zones around the world. And based upon a zone’s ability to deliver above a three and a half percent GDP rate, You tend to see much looser rules and legislation towards access of consumer data because the consumers by giving away and having more corporations, more access to the data, but they’re benefiting from a GDP standpoint.
So they’re prepared to make the trade off essential, right? Their society is improving. Via growth, therefore, access to data is there. I’m simplifying the research, but that’s bucket one. Bucket two is where there’s a lower wealth creation and lower GDP environment. Slower growth environment. So you actually see higher privacy rules come in, because in there you have to have protectionism of the citizen’s data, which is a valuable asset, and then you force trade agreements of the companies coming in to access that data.
So it actually protects the corporate environment. And it gives you a taxable base, a higher taxable base to, to work with that provides the social service fabric that you need in the region when you have a lower GDP output. So it really is this, it’s never as simple as, do you want to give access or don’t you want to give access to things? It’s much more down to, are citizens benefiting and how are they benefiting from their governments? Under what economic proviso. And that’s a, that’s one of the biggest leading indicators where the legislation will land, especially as it pertains to AI.
26:59-27:57
Speaker
I completely agree with you. Yeah. And then you have what trickles into I go more into the human experience, access to technology that, that speaks exactly to it. How do you elevate the use of AI if access to technology is limited? It all goes through the same thing and it all suffers the same in different levels.
Ricardo, our team is bugging me. I’m going way over today on the extended long episode, but this has been so fun to dive into this and just get the perspective from yourself. VML is, as I say, is the biggest agency in the world dealing with the biggest brands in 64 countries. It’s really interesting to hear your perspective and what you’re thinking about in the world of artificial intelligence here in emerging technologies. So thank you for that. We love to wrap up with a little bit of levity and some light trivia just to dive into. So I’m going to ask you, Ricardo, a question. I’m going to give you four choices to the answer and let’s see how you do.
27:58-28:36
Speaker
You ready? I’m ready. Let’s do this. Okay. By 2025, just 12 months. Not even 11 months away at this point by 2025, what percentage of the global population is expected to be frequent users of either virtual reality or augmented reality technology? A) 20%. B) 35% C) 50% D) 75%. I gotta say B. B 35%?
28:37-29:07
Speaker
Yes. Alright, buckle up because according to Snap Consumer Global AR report by 2025, the working prediction is around 75% of the global population is expected to use AR/VR in growing importance. In their business and in their personal lives, like it is phenomenal. I would love to dive actually into that a little bit more with you on the IOT side and AR/VR so let’s come back, get your opinion and go, are they right? Are they wrong? Sound good?
29:08-29:25
Speaker
Sounds good. Sounds good. Great. Ricardo, Vice President of Technology at VML. Thank you for joining us today. For everybody listening, don’t forget to like, and subscribe, send your questions into future episodes, and we’ll see you all soon. Have a great day.
