Insight Meets Sustainability - Steve Phillips

0:00Hey, welcome to the CMO Whisper Show. I'm your host, Steve Olensky. Part marketing practitioner, part ad agency veteran, part journalist. I was a writer for Forbes for 10 years. I've had so many insightful conversations over the years with business leaders, to athletes, to celebrities, to, of course, CMOs. The only difference now is instead of sharing those insights through written form, I'm doing it this way. My guest this week is Steve Phillips, CEO of Zappi and chair of the MRS Sustainability Council. Steve's been at the forefront of transforming the insights industry. Under his leadership, Zappi has pioneered AI-driven tools that help creators worldwide test, validate, and scale ideas long before AI was mainstream.

0:47That doesn't mean he's old, by the way. Today, he's doubling down on new models that unlock deeper value from research data. Now, he's also the co-author of The Consumer Insights Revolution, Transforming Market Research for Competitive Advantage. Written with Ryan Barry, former president of Zappi, and PepsiCo leaders Stefan Gans and Kate Schart. I hope I said those names right. The book gives organizations a blueprint for becoming truly centric by anticipating shifts in behavior. You got it before they happen. Now, beyond innovation, Steve's passion about building a sustainable future for insights, ensuring the industry grows in ways that serve both business and planet.

1:24What a novel idea. His global perspective, shaped by working across four continents, combined with award-winning work in behavioral economics, has earned him recognition from the European Society for Opinion and Marketing Research, the Market Research Society, the Advertising Research Foundation, and way too many more for me to name. Simply put, Steve is an innovator, author, and sustainability champion, redefining what insights can and should be. Steve Phillips, my new found friend, welcome to the show. Thank you, Steve. That's much appreciated. I've got to correct you on a couple of things. Firstly, I am old.

2:00Secondly, I'm no longer CEO. So I founded Zappi and I was CEO for a long time, but I gave up the CEO role about eight months ago in order to focus on innovation. So I'm now chief innovation officer, which means I can spend all of my time thinking about the future, working with AI, trying to focus on data, maximizing the value of data for our clients, all that sort of stuff. So I hate to say it's a lot more fun. Excuse me. Thank you for the correction. And I cannot wait. You and I have had really good conversations off air. We're going to have a lot in common. We already do. We're going to have a lot of fun.

2:34So let me just jump in. And I got to start with AI. Of course I do. Now, Zappy, like I mentioned before, was ahead of the curve in using AI for insights. What problem, I'm going to go to history books here. Let me put the retro on. What problem were you trying to solve when you first launched? And how has that vision evolved? Sure. Yeah. So when I started Zappi, I've always been in Consumer Insights and I've been a serial entrepreneur for four companies now. And I started Zappi very specifically to try and solve a problem that my other businesses hadn't tried to attack. So normally innovation in Consumer Insight was about how you ask the question, how you analyze the results.

3:08And I kept thinking that clients actually were really happy with Consumer Insights or at least the insight they got because it helped make a better decision. Why wouldn't you be happy with that? What were they not happy with? Well, they were not happy with the amount of time it took and how much money it cost. So the idea of Zappi was really, can I get clients 80 % there, the way there, but in 5 % of the time, 5 % of the cost. So that was the start. Obviously, to do that, I needed to automate everything. So at the time, the Apple App Store was very new and cool. And I thought, oh, I know what I do.

3:40I'll create the research app store. So that's what I sort of set out to try and do. Now, then you get to the AIP. So I obviously needed to build it. I needed to find people who could build software. I didn't know anything about software. Some people claim I still don't. But I went through a series of different partners, probably about four or five, before I found a small company, part in South Africa, part in London, who were building some of the software I really needed. We ended up working together, and then we ended up merging. So that company, the CEO of that company became the CTO at Zappi.

4:13As a business, we've ended up being both very strong in domain because of my background, but also very strong in technology because of theirs. And their passion wasn't faster, cheaper. It was to do things better. You can see where we're going. Really, they had AI right at the beginning. So their sort of mantra, their vision was to scale human expertise. It was to take great thinking and put it into software. And obviously, to do that, you needed to use AI. At the time, AI models were pretty simple. So they were sort of, in fact, then this type modeling plus data modeling. You'd have data modeling, output of data modeling, and then we would write the report.

4:53The report was written right from the beginning, like 12 years ago, with an AI. It wasn't the greatest AI. LLNs are much better at writing. But it had embedded in it the understanding of the data that an expert in micro-research had. And so it would spit out these conclusions about your study that you created. So in that way, we could turn around a market research project, not in four weeks, but in four hours, which is really what we're going for. Yeah. So AI, of course, is everywhere now. And you've been using it, when I say you, I mean Zappi, since your inception. But I want to jump forward now to today and get your thoughts on what are some misconceptions brands have about applying AI when it comes to insights?

5:35So I suspect the biggest misconception they have is data protection. And I think understandably, they're very concerned about their data leaking into the general public, but also leaking into the modeling of other companies' models. And at least the enterprise systems that we use are pretty secure. So that's probably the biggest misconception. I don't actually think, though, that most clients have misconceptions. I don't think misconceptions are the problem. I think it's behavior change that's the problem. So you'll have seen the MIT report that came out this week, which said 9 % to 5 % of enterprises have tried but not managed to create genuine productivity gains through AR.

6:20And I don't think that's because of the fault of the AR, because these models are very, very good if you train them on the right way, if you input your data into them, if you find out where they should plug in properly. The problem is workflow. The problem is getting people to change their behavior. So, I mean, it's an illustration. We have now, we have a very large data asset. So if you wanted to create a new beer brand for young men in Texas, our data asset, we know a lot about young men in Texas. We know about what they think about beer. We know what they think about communications. They know what they think about brands.

6:55So we can take all of that insight, integrate it into an LLM and say, right, let's create a new beer brand. The truth is you could sit with our system and create 40 or 50 concepts in a day. That's one person in a day. And in the old-fashioned world, you'd wait three weeks to arrange a workshop, had a workshop for a week with five people and come up with three concepts at the end of it, which didn't have consumer insights embedded in it. And now one of those people can create 40 concepts in a day, all with consumer insight embedded with it. Interestingly, we can now test them overnight. We can not only just test them overnight.

7:32We can then use the consumer insight results that we get and then get an AI to reconfigure the concept in light of those results. So you're cutting down something that took four, six, seven weeks and five, six, seven people to two days and one person. And the outcome is not two or three ideas. It's 10 or 15. That's a huge behavior change. So it's not adopting that is really, it's actually really difficult because interesting, all of those five people liked being in that workshop. They liked creating new ideas. And of course they do. But they need to change the way they work. So I think it's not so much the fault of the AI and misconception of the AI.

8:17I think it's going to take time for people to adapt to the new technology and adapt their ways of working around that technology. I could not agree more. And I didn't want to interrupt, but nodding my head vociferously, because, you know, I talk to CMOs and marketers all the time. And there's that, maybe you're right. It's not the misconception, but it's the, I don't want to say fear that might be too strong or the reticence to, okay, well, how do I use AI for insights? And you're so right about it's a behavioral shift. That's what it is. Yeah, it is. And the interesting thing, there was an ARS or AMA, it may have been AMA, survey done, it's six, nine months ago, about use of AI within the broad marketing remit.

9:03And actually, AI was taken up by researchers much more than marketers. And I think it's because researchers are used to working with data. We're used to looking at multiple types of data. We're used to trying to extract values and data. And really, I think that this way, the insight leaders, quite a few of them, are really taking a charge for their companies, their wider companies outside of insights. But because we understand data and because we're looking to maximize the value, we can really help companies really adopt AI in a much more effective way. So I think it's an amazing time for insights.

9:37And what I would say to CMOs is really hug your insights department at this time in a way that you probably haven't before because they really can help you. They have a lot of skills in understanding data, understanding different types of data, integrating that data. So expand on a little bit because you mentioned CMOs and brands. And is there one or two or how many, what's the most important shift they need to make in how they approach consumer understanding and research? And you started to talk about it, but elaborate a little. Yes, I wouldn't focus in on consumer insight. I would focus in on the whole thing.

10:13I think we're moving into an agentic world, a world where agents are continuously, you'll have different agents doing different jobs and doing different things, including agents sitting on consumer insight, data agents trying to create communications, agents trying to create concepts. if you try and just focus on insights on its own, I think you're missing the trick. I think it's about thinking through entire workflows. So instead of thinking about insights, I would think about innovation, for instance, or communications and say, okay, in an ideal world, how would the new world of communications look?

10:54We're already seeing, you know, so at the moment, most large companies, when they create a campaign, will create, say, one major piece of video content. And they'll test that with consumers. And it costs a bit of money, and they're interested in the results, and they take the results on board and optimize their ad. But then when it comes to all of their digital ads, they're not going to test those. It's just too expensive. They're not going to test outdoor. They're not going to test radio or whatever. But in a world where, for instance, there's synthetic data, which is what we're creating now with synthetic respondents, actually the price point for an individual test the price point and the time changes completely to you know marginal cost to zero and timing under a second to do a test well in that world in that world you can start saying okay let's embed consumer insight throughout the process not just at one stage but throughout and if i'm embedding consumer insight throughout the process then i'm creating much better briefs strategies ideas but also final executions of everything I'm creating.

11:59So I think you can, if you think about the macro place, and then you start adding agents on, and it makes the whole process, again, much quicker, much more connected, but much more coherent. So we're now allowing companies to go, okay, I've got my main piece of communications, I've got a poster over here and a digital ad over here. Are they all doing the same thing for my brand? Are they moving us in the same direction? Or has the digital agency gone off at least? It's something weird that doesn't say the same thing. It may be good, but it may not say what I want it to say about my brand. And now we can say, no, don't worry.

12:37We can make sure all of your communications is coherent. We can make sure it's all doing what you want it to do. We can make sure it all works together. And we can do that because we plan up front to use data in a much more, with AI, data plus AI, in a much more coherent, much more integrated manner throughout the website. So I think it's such a change. AI is such a change, and particularly agentic AI, that you need to stop thinking about existing silos and start thinking much more macro in terms of what are the things that we're creating as an organization and what do companies do? They create products and they sell products.

13:14I mean, that's effectively what any company does. So let's rethink all of those workflows, see where data, see where insight can play a much more sophisticated role. and obviously AI direct process. Without question. Without question. Correct me if I'm wrong, but I think you said either in print or in person about the role research is shifting from validation to inspiration. What did you mean by that? So research, you'll know the old adage that it's always being used a bit for arse protection, right? Oh my God, I can't release this ad or this product unless I've done some research, just in case someone blames me afterwards.

13:52And then I can go, no, no, no, research. well. So the old adage of research is used like a drunk uses a lamppost, more for support than illumination. I love that line. I've always loved that line. And I think when you're in a world where you're using research to test or help with an individual business decision, and that's how you use it, then that's true. So you come up with an idea or a product or a piece of communications or a brand strategy and you go, oh, let's test it with consumers. Will they want us to go in that direction or create that thing? And you get some nice feedback and it's wonderful.

14:29You make a decision and you move on. And what do you do with that research? What do you do with that insight? You just throw it away. I mean, that's the traditional, you've made the decision, let's throw it away. But actually buried in that research is huge amounts of really fascinating information about how consumers think about brands and products and communications and features and all of it. So what we now say, let's stop just testing. Continue testing, but let's stop just thinking about it as a test. Let's think about it as a data asset. So you build up that data asset over time. And then you can start asking different questions.

15:02So instead of saying, do young Hispanic females like this ad? You can say, what type of ads do young Hispanic females like? So then that's inspiring me as a creative to create something that I know is likely to work with them. So then you're going right to the beginning of the process to inspiring people because you've got that data asset. Yes, sure, you test, but you test and you use it to help with that decision. But you also put it back into your data asset. So your data asset gets bigger. It gets smarter. You get bigger and smarter over time. And you use that to just make the business generally much better, make better decisions right at the beginning of the process as well as right at the end.

15:48Yeah. It's almost like not to reword what you said, but the validation will always be inherent in any research. But now, in addition to validation, it's inspiration. Exactly. Exactly. Exactly. And the interesting thing is, I think, you know, we've always said we want research to do. So there's always the validation and optimization were always the two things that people used insights for. So you'd validate whether it's a good idea, but also get some idea for feedback about how to make that idea better. So optimization and validation were always there. I think now we can, now because you've got the data asset, you can also create inspiration.

16:25But also when you align it with an LLM, you can start doing the creation process itself. So you can inspire, you can create, you can optimize and you can validate. all of those things can have insight buried at the heart of the process itself instead of coming in much later. Gotcha. Okay. Sustainability. I know you're the chair of the MRS Sustainability Council. Before I get into a question about that, tell my listeners what that is, what the council is. So this is a part of the market search society in the UK. And what we wanted to do is play our part alongside the advertising industry, which has something similar about trying to create much more sustainable advertising.

17:03So instead of flying 37 people to Australia to film on a beach, maybe you try and only send three and go to Croatia instead. So they're focused on trying to create more sustainable advertising. And we thought it was important that the insight industry played its part. So we are trying to get all the different insight companies in the UK to sign up and say they will move towards net zero. They will use sustainable processes in terms of offices, in terms of research processes. So we're trying to push sustainability within the UK with MRS. We're also doing the same thing with SMR for the rest of the world and have worked with some other industry associations around the world, like in Germany and looking at the States.

17:50We're also trying to create a sustainability library. One of the things Insights has is a lot of data about consumers around the world and how they think about sustainability, how they think about climate change. And we're trying to have a central library through SMR of all of the most interesting data that can then be accessible to people like governments, NGOs, academics, students. And that really is being spearheaded by Kantar and Ipsos, who do a lot of work in that space. So it's wonderful that they've come on board and they're helping with the process. And we're really trying to get the insights community to act together to just play our part in trying to make things as sustainable in terms of our future as possible.

18:35I love that you have that sense of responsibility, which we all need, quite honestly. But the question I want to ask you and on behalf of everybody listening. So how can a business balance growth while still being social, environmentally responsible? Yeah, sure. So we're we at Zapier, a B Corp. And V Corp is a certification you go through. It's actually really painful to go through. But we balance people, planet and profit. So we are VC and PE back, right? So we absolutely have a growth and profit incentive. We absolutely do. But we also want to be really good to our people. Our people, our employees are really important.

19:14So is our local community. So we do, people do volunteering. We, our office isn't always full nowadays post-COVID. So we let some charities use some of our office space. And being a part of that community, I think, is really important. I don't think that negates profit at all. I think it makes people feel better about the company and want to work for the company and want to work hard for the company. So I don't think it's a negative. But we're also thinking about that. And then also planet. So we're trying to – it doesn't mean I don't lie to the states to go to a meeting. I do. But I think about it.

19:46So every time we do something that will have a negative impact on the planet, we try and factor that into the decision. And sometimes you just go, look, it's a board meeting. That's more important that you're there. And then we can offset the CO2 emissions. But it should be part of the thought process. So the beauty of the B Corp thing of people, planet and profit isn't that you focus on one at the detriment of the other. It's that you're thinking about all three simultaneously. And of course, commercial is probably going to come first in a lot of situations, but the others will always factor in.

20:22It's almost like, I don't know if you get this granular, but it sounds like it's the decision literally to travel and do events or it's a case by case. Yeah, yeah. But they're case by case on a commercial basis, right? So if you're thinking about travel, yes, it impacts the planet. It also impacts your pocket because it costs a lot of money to travel. way. So it's always true. We're just bringing other factors into it as well. And so if I think about people and I'm going over to a board meeting in Boston, I better make sure I meet up with as many of the employees as possible and find out what's going on with them and checking in with them and seeing if they're okay.

20:57So those are just different lenses to think about in that decision. Yeah. It's like maximize your time to the fullest kind of thing, which sounds so obvious, but oftentimes obvious gets overlooked as we both know. Yeah. I think it's the discipline of just having it on the balance sheet of having that thought process is the important thing. Quite often it doesn't change for decision, but sometimes it does. I want to shift to get a global because you've worked across four continents, right? So you've experienced a lot of different cultures. What are some of the differences you can say people ask you about, you know, when it comes to just insights globally or research or how it varies by different areas across the world?

21:38I suppose I come from it. I worked out in Asia for about six years. I've done a lot of work in different markets in Asia. I worked in New York for six years and been back in Europe. I think the interesting thing is you can always find differences. You can find differences within countries. You can find differences between cultures and between countries. You also find, frankly, trends within business. So 20 years ago, you had to have a China strategy. And now as you have a China strategy, you look at it being weird because we're almost separating. You know, it's very, very different. But there's more that unites than divides.

22:14I mean, that's, I think, the core message. If you look at what sort of products people like, what sort of advertising people like, what messages and brands people like, saw there will be differences, but everyone has family, everyone has friends, everyone likes sports. There may be different sports, but they like them. They like consuming. So there are so many more things fundamentally that align us and divide us. And I think it's very easy to focus on the differences. But quite often, if you can create something universal, create a universal campaign, it can absolutely appeal to everyone. And you see that with iconic brands around the world.

22:53you know love everywhere literally everywhere whether it's dove whether it's pepsi whether it's you know whatever it might be there are there are brands brands that uh people all across the world i'm really interested to get your take on this next question or topic and it's about it's about back to ai but you're very uniquely qualified i think to answer this question to those skeptics or cynics who say, my fear is we're going to lose the quote unquote human touch, right? No longer will my gut instinct matter or intuitions, right? You see where I'm going. What do you say to that? So that's not my experience so far.

23:36AI is incredible at doing lots of different things and it absolutely can be creative. But what we're finding very much is that having a human in the loop makes things significantly better. So for instance, we had a competition within Zappi when we first created our agentic system to create new product ideas, new product concepts. I got 150 people in Zappi to create a new idea. It was for a chocolate bar. And we got people from Mars who have been clients of ours to judge which were the best ideas. And we got the top 10 ideas. And then we tested those with consumers and we had a winner and they had a prize.

24:11But the point of the competition was to get everyone using this new agentic system and playing with different agents and seeing how they would work and overwhelmingly within the top 10 the top 10 ideas came from people who spent the most time with the ai talking and interacting with the ai people who were really lazy like me who spent five minutes came up with the concept and then got on something else our ideas did not do well the people who spent 40 minutes an hour creating a product going oh no i I don't like that bit of the packaging. Or can you change some of those ingredients? I'm not sure people are like this.

24:46What do you think? And conversing with the AI and therefore effectively conversing with the data underneath the AI, they came up with them by far the best ideas. So I think it's not going to be one or the other. I think at least for the next five years, it's going to be a really nice symbiosis. Does it mean that you can do things a lot quicker? A lot quicker, yes. But also it means you can create a lot more things. So as I was saying before, you don't come up with three concepts, you come up with 30 concepts, 50 concepts, you know. So it changes the dynamic. All of those concepts should be better than the three that you'd come up with on your own.

25:24And all of them have some legs that you can use to create the one or two things you really want to take to market. So it just massively improves the standard of output with human beings at the heart. And those human beings have, I mean, this is one of the things I've always said about consumer insight is that we at Zappi do quantitative research. My background is a lot of qual. And I used to think that by far the most valuable research you could do was things like ethnography, where you as a human really went to spend time with your customer. And it might be, I remember doing a study with a client that did bathrooms.