Executive Interview

Generative AI for Marketing: A CMO's Perspective

Gary Tubridy, principal at the Alexander Group, spoke with Shonodeep Modak, chief marketing officer at Schneider Electric about generative AI and how it is changing the game for Marketing.

Shonodeep shared his experience with AI, how his team has leveraged tools to reduce time and how AI has allowed the company to gain better outcomes. “You have to jump in. You’re going in 100% and you either see savings, see the pain, see what is going to happen. But if you don’t go all in at the low level, none of this will start to evolve into a great value-added interaction.”

Gary Tubridy: Hello. Gary Tubridy here, principal at the Alexander Group. I’m here with Shonodeep Modak, who is the chief marketing officer at Schneider Electric. And we’re talking about his keynote, which is going to be “Generative AI for Marketing: A CMO’s Perspective. So it’s a hot topic. Shonodeep, great to have you here.

Shonodeep Modak: Thank you for having me.

Gary Tubridy: Thank you for joining us to talk about AI. And as I was mentioning before, it’s coming up all over the place unprovoked. And here we are talking about it specifically. So we’re going to have a great keynote. Thank you for sitting down for this interview.

Shonodeep Modak: You’re most welcome.

Gary Tubridy: Tell me a little bit about the impact of AI and marketing at Schneider.

Shonodeep Modak: So maybe a quick plug for Schneider. Schneider is an energy management and automation company involved in essentially the digital revolution of energy and helping digitize both our partners and customers, both been around.

Gary Tubridy: For a long time.

Shonodeep Modak: Long time, you know, a couple of hundred years overall. And then in the last 100 years or so in the electrical distribution space and now getting more and more into automation. And so our our whole AI adventure is not only just using AI, but also creating AI,  generative AI and machine learning for our customers in the areas of IoT and, say, a manufacturing plant or AI and how you’re interpreting and using buying energy and managing it in your building or your facility. We create products that have AI built into it now.

Gary Tubridy: Yeah. AI is so important and we’re talking about energy management, the list goes on for where it can provide value. How would you recommend leaders build their personal knowledge and skills in AI so they can take a more assertive role in figuring out how to leverage AI for their business?

Shonodeep Modak: I mean, certainly, there’s a tremendous number of consultants and researchers who have published papers on AI. But I really believe that to be truly knowledgeable, you have to be a practitioner. And it’s not that hard. It’s not like, for example, I need to use CAD software and I need to go attend 10 different courses in a university to go do it. This is like you can actually go pick it up, just like, say how Microsoft Office works or whatever.

Gary Tubridy: Don’t be intimidated.

Shonodeep Modak: Don’t be intimidated and just start using it. And you may not get the ideal output on it, but it’s most of these tools are readily available, and you just have to start jumping in and using it. And that’s the best way to be able to help lead your teams and be able to help figure out how to best transform and transition to it.

Gary Tubridy: Yeah, it’s not all academic. So we’re talking about AI in Schneider. You’re a CMO and so AI resides in Marketing. It can also reside in Sales or Service. So where does AI reside in the business entity? Does it have a home as a department on its own or is it inside of Marketing and Sales and Service? How does that work?

Shonodeep Modak: I really believe it’s the latter. It sits inside every different entity and function, and what we found is there are two ways to go about it. One is you bespoke, you have to go and build it yourself, or the other way is you go buy the platform and you go implement it. And so for us in Marketing, there are what we call five pillars of AI. And this is not Schneider saying this. This is just what’s out there right now. You’ve got AI for text generation, AI for image AI, which is design, AI for video, AI for audio and speech, and then AI for basically coding.

Gary Tubridy: Tailor-made for the marketing function.

Shonodeep Modak: Yeah, exactly. And the back to your point, those things from marketing actually spread into like your point customer service into sales, into field services, a lot of different places that the call it the area of marketing has and I can give you some more examples we can go into. But this is a lot of this generates benefits for other functions in a new way for marketing.

Gary Tubridy: Does this often start in marketing because it’s so tailor-made for the function?

Shonodeep Modak: Well, it’s usually the first use case. It’s the it’s the easiest place to go. I would say the lowest hanging fruit to go start with. So for example, Marketing is all about generating content. Well, generative AI is all about making that go faster. And so the huge low hanging fruit is today I would have to write a media brief and that brief would say, I want to write a blog about Alexander Group and, for example, this event. So I can now take all of the content that was being prepared, put it into AI, publish it out as a blog in 20 minutes. Versus say several days. So the lowest hanging fruit I think is in Marketing.

Gary Tubridy: Are there things that companies do if it starts in Marketing, but it could also go into Sales and Service? Are there things companies do to help that migration from Marketing and out to other customer-facing functions?

Shonodeep Modak: Yes, I’ve seen it happen a lot with just this aggregating and going through and sifting through this vast amount of data. Perfect example. Customer service. Customer service, we get thousands of troubleshooting calls, issues with the product or how to use a product from to our customer service board. So they’re answering all these questions all the time. It takes up massive amounts of bandwidth of people to go support that. But if I could sift through all that data, I could then go in and say, AI read to me, what are all the commonalities and the issues and troubleshooting? And what would you recommend are the types of content that I should build? And then all of a sudden now we can tell our content team, our technical communications team, hey, we need a how-to guide for this particular control that’s always giving people problems. Or for example, they just don’t know how to use this part and it’s consuming all of our team’s time.

Gary Tubridy: So the application isn’t always obvious until you get into it. But then if you work the if you work the process, you can see where the cross-organization effort is worth it.  Makes sense.

Shonodeep Modak: Exactly. And we’re just at the very tip of the iceberg, as you probably, as everybody is experimenting and trying it out. But imagine, a few years from now where it’s really promulgated all into the commercial, activities with the, with their customers. And then you can actually go and use AI to go and figure out the interactions that the customers have with their networks and be able to give additional value, that that’s not to be seen yet.

Gary Tubridy: Yes. So the downstream of possibilities are pretty interesting, aren’t they? Yeah. Hey, are there new jobs that come out of this that we have to think about and build because you need them to take advantage of it?

Shonodeep Modak: Yeah. The biggest ones that we are seeing now, the need for is you have what the previous work was where you would go and say for Marketing, I’ll speak for marketing’s sake. You go and orchestrate a brief or you go in requisition a company that does design work for your images or takes photos or does a video shoot. And in fact, even in this room you’re sitting in, you have, for example, this crew that’s with us. Well, in the future, the way it would work is you go to one prompt and the prompt then goes and generates all your content, your design, your videos, etc. Now, of course, some of that’s not fully mature yet, but the skill of writing the prompt is what we call prompt engineering is going to be this new world skill that everybody’s going to want to go hire for. Some of it you can train, but some of that prompt engineering is how to actually get exactly what you need out of the AI prompting requires actual education and call it unique skill sets for.

Gary Tubridy: A little bit like getting the right words early, early on in the search engine that started to pop up 20 years ago.

Shonodeep Modak: Exactly, exactly. And just like how, for example, you have people that were trained like I talked about CAD drawings and everything. Now, what it does is it makes it easy for anybody to use. But the prompt engineering is going to be very helpful to accelerate it inside your company.

Gary Tubridy: Yeah. Is there an AI, a group or department that would work underneath you in the marketing department?

Shonodeep Modak: So we’re trying to figure we’re trying to figure this out now at this point is there is the bespoke you have to go develop it yourself, in which case you’d have to create your own support for or do you buy it off the shelf and then have as far as platforms, then you have to have support inside for those platforms. The platform, call it managers and experts. What we haven’t figured out yet is if we keep it, for example, I always gave the analogy about Microsoft Office, you know, back when we were using manual spreadsheets on paper or Lotus Notes and we went to Excel. Did we have to have a whole department created to get onboarded with that? In most cases, not really. And if the tools become usable enough, possibly, we don’t have to have a whole nother department created.

Gary Tubridy: Yeah, yeah. So it might be a tool that feels to us like Microsoft Office, which we use in Marketing, we use in Sales and we use in Service.

Shonodeep Modak: Exactly. And then here’s the funny stuff. Imagine when I actually starts to create AI. So you ask AI to help. Hey, write me a prompt that I want to do this, and then it writes you the prompt. Or for example, AI create me an AI service that can do this.

Gary Tubridy: Interesting. Are there skills that folks in your marketing department need to acquire to leverage AI? And to the extent that’s true, how does a Schneider help them achieve that or get those skills?

Shonodeep Modak: So we’re going through and evaluating the landscape of training that’s available right now. So from different, call it, sources of both vendors as well as, call it, independent training areas. That’s where we’re kind of looking at, and hopefully we can find the ones that are good enough that can get everybody upskilled. And then, of course, where you want to go to the next level, for example, design AI or video AI, that requires even more, it’s not coding, but it’s that prompt engineering that might be something we have to go in and really buy.

Gary Tubridy: There might be there’s some make or buy decisions there. But the acquisition of skills is going to be something that everybody needs to think about within Marketing, Sales and Service and the Revenue function.

Does AI offer an opportunity for functions like Marketing, Sales, Service? We think about those as the Revenue functions. They’re separate functions, but they kind of have to team together. Does AI offer an opportunity for teaming to occur between these functions in a smoother fashion?

Shonodeep Modak: The obvious parts are there’s so much data that both of our departments have. Marketing sits on all this contact data. Did the digital, call it, cyberspace? Of what all our customers and partners are using from our website and how they’re engaging with us. So there are all these triggers and behavioral things that Marketing has. But it’s so much information that today no even small team can really go and digest it. I gives that much at a faster rate. And then on the sales side, you have the CRM systems, you have all the sales interactions that are happening with customers. Sales, for example, in one department, I’ll take our company, for example, has no idea that another country is dealing with the same customer, similar problems and they can bring that together. So now you have all of this AI reading through all of this, and then imagine now when you bring it together and you can create either support content or you bring it together as far as sales enablement for the commercial team or you bring it together as far as campaigns, outbound campaigns, it’s just an untapped environment.

Gary Tubridy: Where do you think this initiative starts? Does it start in Marketing and let’s team with Sales? Do we come together in an organic sort of way? Or is this a chief operating officer thing that says, let’s bring this thinking together? Let’s create a new process.

Shonodeep Modak: So there’s there’s some great studies about AI achievers and how do you become really successful in being an AI achiever versus just the low maturity area? And I’m a true believer that you cannot dabble. You have to jump in. And it’s not the operator, chief operating officer saying, you got to go do it. We’re we are trying to now move into pilots where it’s not, hey, take one person from your team and just experiment with it. It’s no, you’re going 100% over and we’re going to see hopefully you either see savings, see the pain, see whatever else is going to happen. But if you don’t go all in and it has to happen at the low level, none of this will start to evolve into this sort of great value-added interactions.

Gary Tubridy: Yeah. So sometimes you got to kind of force the issue. With so many opportunities to consider Marketing, Sales, Service and upstream and downstream, there’s a matrix, maybe a Rubik’s cube of possibilities. How do you decide where to focus the time and attention and maybe even decide where to start?

Shonodeep Modak: Yeah, yeah. So I would recommend…I mean, I’ll speak to the marketing department and then maybe I’ll hypothesize other areas. There are call it the lowest hanging fruit, like I mentioned with text, that has implications in terms of cost, in terms of what you spend with agencies and what you spend with external media buying, etc. That’s where I would go first is where can I help save money, redeploy it into more advertising, into more events, into more outbound stuff. So I would start where you can see real hard savings first. That’s like quick, quick cost savings, redeployment into productivity. Then you can start building on like, okay, now that my team is getting upskilled in it, I want to do campaigns that go in search for behaviors of all these customers in cyberspace and brings it back and create a targeted. So building on that and then outside of the outside of our function, beyond those hard savings, I think the biggest benefits is sifting through this mass, all this massive amounts of data that nobody’s ever had time to look for. Now it can do that and create reports and create even PowerPoint presentations for your boss to go and present the data to you at the same time.

Gary Tubridy: But it’s interesting what you say because it almost got to my next question, which is you can see where you can save time. But you also started getting into and with that time savings, I can now redeploy the time and perhaps the money to actually improve results. So there’s efficiency but there’s also effectiveness possibilities here as well.

Shonodeep Modak: Yeah. Exactly. And so it’s yeah it’s efficiency, it’s productivity. Not always going to be like hey I’ve got to save people or save costs. It’s just with that same person I can do maybe, maybe 5-10 times more.

Gary Tubridy: Yeah. And again there’s so much there. You’re not going to know what goodness will accrue until you get into it. Any watch-outs for leaders to be mindful of as they approach the AI challenge?

Shonodeep Modak: That’s a great question. I would say, if you go and look at the landscape of players in the AI space that provide services, the off-the-shelf stuff, there is thousands and thousands. It’s almost like, remember when you would see a picture of logos of everybody that was in IoT, or it was in bitcoin or cryptocurrency and all the players there. It’s a massive landscape right now. I would say don’t get stuck into one particular player and get and commit forever because the landscape is going to change just like it did for those areas. So, it’s not a one-way door. You have opportunities to go and make other pivots as you’re going to have to. Your tech stack now. And I’m talking mainly about using it as an internal function. You also can create AI products. Many companies here in this event are tech companies. But don’t get stuck in one spot.

Gary Tubridy: Be fluid. Be dynamic. Yes. Yeah yeah. Be and and and I think very importantly you said don’t get yourself committed to something which downstream is going to be a hindrance.

Shonodeep Modak: Exactly. Yeah. Exactly.

Gary Tubridy: Any predictions for AI as we are on the cusp of entering 2024?

Shonodeep Modak: I bet you by the end of 2024, people are like AI for creating visuals and videos and deepfakes. It’s always it’s always a little fake-looking. And there’s somebody with six fingers on one hand. That’s going to be resolved. I guarantee you that’s going to be resolved. Speech and audio will be resolved. And maybe if you invite me back next year, it can be a deepfake of me and a deepfake you.

Gary Tubridy: We’re going to have a then and now. Let’s plan on doing that.

Shonodeep Modak: We’ll make ourselves look younger and thinner.

Gary Tubridy: I’m all for that. Shonodeep, thank you so much for joining us at the forum.

Shonodeep Modak: I very much appreciate it. This was an amazing time here. Thank you.

Learn more about the Alexander Group Leadership events

Back to Top