How to Prove The ROI of AI with Elizabeth Maxson Martinet, CMO of Contentful | Ep. 400

0:00Today's episode is brought to you by Customer.io. You have customer data, but turning it into personalized marketing at scale? That's the hard part. Customer.io connects the dots. Feed it your data, start sending messages that actually feel personal. Learn more at Customer.io slash TMM. Welcome to the Marketing Millennials, the no BS marketing podcast. I'm Daniel Murray and join me for unfiltered conversations with the brains behind marketing's coolest companies. The one request I tell our guests, stories or it didn't happen. Get ready to turn the f*** up. Welcome back to another episode of the Market Millennials Podcast.

0:51I have Emacs on the podcast, CMO of Contentful. I am excited for this conversation and I'll let you get into how you got into marketing and that journey and then we'll get into this conversation today. Yeah, that sounds great. I went to school for marketing, so that's pretty simple. But I have had a long journey to get here as CMO. I think a lot of CMOs tend to come up through marketing routes, through product marketing or demand gen. I'm unique and I come from an events background. But that also has given me an opportunity to be very passionate about brands and creativity and creating physical experiences and being able to translate now into this world of digital experiences.

1:38and a big part of my career journey has been spending over a decade at Salesforce where I was head of marketing for Quip as well as the CMO of Tableau and now been in seat here at Contentful almost 18 months. This is not exactly where the conversation is going to go but since you were in events, what is the state of events right now in your opinion in 2026? Should people be doubling down on it. What does it look like? And then we can go into more of the conversation. Yeah. I mean, let's think about COVID, right? During the pandemic, that was the biggest event disruption anyone ever saw. Huge opportunity for companies to have to pivot hardcore to virtual.

2:24And what did that mean? So what's interesting about that is part of that world has not changed. There still needs to be virtual options. I think that gives more flexibility and availability to content experiences in a virtual and digital environment. However, I do believe in-person events are making a huge comeback. I think people crave it. They just, I mean, we just had our own company kickoff recently here in Berlin at Contentful just a few weeks ago. And I can't tell you how much our employees just really are coming out of that experience more energized, inspired, ready to go to work and think about that across every capacity.

3:01So I actually do think it is a place to double down. We see a lot of success through our events. But a big part of that event is making it personal. And what can you do to really drive experiences that are going to separate from other experiences, right? Like every dinner is a dinner. What's going to make your dinner more special to really get the right accounts, whether it's prospects or your current customers at an event. So I do think there's a lot of opportunity there. And now consumers are expecting that. They want those experiences, they crave them, they crave that human connection. So the more that you can really double down on that, I think is definitely going to be something that all companies should consider.

3:40I totally agree. I just wanted to hear your take on that. I wanted to also go, since you said there was the one disruptor of COVID, now we're in another era of another disruptor coming into every space, not just marketing, which is AI. I know you've talked about how entry-level marketers are becoming more full stack and thinking more like engineers. So what three technical skills will every marketer need in the next few years? So on the technical side, first, it's being more analytical and looking at data. So having access to data, knowing how to read it, knowing the story that it's going to tell, that trend.

4:17I think there is, I come from the data space. There's obviously data analyst as its own role of someone you can lean into on more of that technical aspect, but the more you can actually get your hands dirty and have access to data and analyze on your own, I think is really going to set marketers apart. Obviously, prompt engineering and really looking at how you can be spending more time on writing the correct prompts and really understanding the output. One of the things that is really interesting is not all results that come back are true or accurate or pulling from the right data sources or pulling from, you know, places that are still accurate.

4:55So being able to be curious about your results is also, I guess, more of a soft skill of being able to really focus on that from a propped engineering standpoint. And then ultimately it is softer skills. It's being able to operate in ambiguity. It is being more curious. It is taking more risks. It's being welcoming experimentation. So I would say it's more like two technical and one soft skill that I think is really going to help, especially with all this new technology that's consistently changing at such a rapid pace. So you talked about those skills that entry-level marketers need, or actually probably most marketers.

5:32All marketers. Yeah. But entry-level marketers, you have a playbook of what you should be doing now in the age of AI if you want to be successful. Now, marketers listening that has been there for a long time, start developing those technical skills as well as thinking about those stop scales. But I know also a lot of CMOs are dealing with this agility paradox, needing to move faster with AI and automation while still delivering personal experiences and human experiences. So what systems or operating models have you seen work best with balancing speed with quality? So honestly, I think a big part of this is a mindset shift.

6:18And where do you want your team to be going to support that strategy? So you're correct. We did a report with The Atlantic and held some really great data and found that 40 % of marketers say that fast and efficient execution is important to them. And the ability to deliver quality content was 42%. So delivering 42 % of high quality content is also helping with that efficiency. So what I think is happening right now though, is what we're calling a great content collapse. If you think about all this pressure and all these tools that you now have as a marketer, more is expected to you than ever before.

6:57Of course, we've always been asked to do more with less. That is not new, but we are getting a lot more pressures from our boards to be really having AI show ROI and how are you doing that? And how does that come from your culture? So that's the first part of the collapse. The second part of the collapse is not producing the right content. More content doesn't equal better content. Better content equals better content. And getting that high quality really involves needing to have content to be personalized. So all of that leads to AI slop and content not resonating. And then ultimately, the third thing that is breaking is your funnel.

7:34And I don't even know what a funnel is anymore these days. I mean, you know, we joke that, you know, back in the day, you would think of a funnel as a three-course meal, right? It's like top of funnel, middle funnel, bottom of funnel. Now it's a buffet, right? And you really need to understand your audiences at a deeper level. You need to be meeting them where they are on the channels that they spend the most time. And you have to break through this noise. And so while we're in this agility paradox of, you know, trying to have that speed and efficiency, I think it starts with core fundamentals of marketing, which is understanding your audiences and what are the channels that really resonate and then apply that to your messaging and how can you personalize that to make that matter.

8:15So what are some things you have done to implement this mindset shift at Contentful? So first, I think it really starts with empowering that curiosity and testing and being not afraid to fail. If we are going to try something, let's fully try it. And if it doesn't work, great, then let's pivot and learn from it. So that starts with a culture. But I also look at one of the best parts about my job and being the CMO of Contentful is my team uses Contentful. So we call it customer zero. We are our best use case, right? So in the case of personalization, we like to have fun with it. I think a lot of CMOs are very protective of their homepage and they are afraid to make changes there and experiment.

9:04And they look at that as, you know, this is this thing that I cannot touch. And I would challenge all CMOs out there that this is something that you should be personalizing. I'll give you a great example. We found that 26 % of visitors to our website were clicking log in. Well, obviously they're logging into the Contentful platform. They're already Contentful customers. They shouldn't then be saying, why buy Contentful? They already bought Contentful. So they should be saying something new. What new features are we launching? We also market to both marketers and developers. We have very different experiences.

9:36So a marketer, for example, if you're an existing customer, you land on our website. We have a great catfish going on right now where we have lorem ipsum is spelled out as our headline. And then it says dot, dot, dot, whatever. And customers, we just had one today actually come through our BDR teams. It was like, oh my gosh, you totally caught me. You know, they thought that we had clearly made a mistake on our website and it was very intentional to break through this noise. And we saw a 275 % increase into our personalization product page in just seven days because of this catfish, right? And it was meant to be kind of quirky and it's worked and it's still on our page.

10:14I just came across it again today. So that's the other thing. Once you find something that works, you know, stick to it. You don't need to change it. So being able to really foster that communication, that creativity, that experimentation. I mean, I think SEMO's really need to be doing that more. And that comes that mindset shift and the culture you want to drive in your team. That's funny. I once did a newsletter about personalization and then we don't collect first name in our newsletter, but I put like, hello, first name in brackets. And I got probably 150 emails like, you know, your personalization is broken.

10:49And I would reply, you know, I didn't get your first name in the form. Like, this is a joke. Yeah. So it's just funny. Marketers notice those things. So I think it's more than other people notice those things, which is such a smart move on your part to get at least spark interest or get conversation going to the contentful team. Yeah, exactly. You said something very interesting, and I've heard another CMO talk about this as well. But marketing teams are going to now have to prove ROI of AI. It's all good and well that we're testing AI internally and using all these tools. But now we've got to actually prove that it's driving some sort of revenue.

11:36We're slowly out of the experimentation phase. We had a little bit of that. We still a little bit are in it. But if someone, a CMO asked you, how do I prove AI is actually working in marketing? What are metrics you're looking at or business outcomes you're looking at that they should be measuring to prove AI? Well, let's start with where I started and what I did first, but that will imply what I think my advice is moving forward. So think about this. It's like, okay, everyone is experimenting with all these different tools. Everyone's using AI. What are they using it for? I don't know. Let's start a use case log.

12:17So our team created just a standard spreadsheet and we had different columns that was, what's the use case? Who is using it? What's the intent of using the AI? And then what did it save you? Was it time or did it save us money? Those are two things that I care about. And then obviously what was the output? How much should we trust the output and so on? And so this was our way of creating the most grassroots tracker of a first understanding. And then what I started doing was looking at that tracker monthly and then quarterly of, okay, what are some really key trends that I'm seeing? Oh, we're seeing a huge uplift in time being saved here.

12:54Okay, how can I be looking at this more holistically? Or I'm seeing a lot of money is being saved here. What's like the total amount we're actually saving? And so a great example of this is we brought a lot of our video work internally here at our team. We have this amazing event that we host with prospects called Contentful at the Cinema. We buy out AMCs all over the world. We host movie premieres. So when you come to the movie premiere, it's super fun. You get tons of really cool stuff. And of course, when you sit down in your seat, what's the first premiere that you see? Well, it's about Contentful.

13:26And we now build these amazing premiere videos about these two to three minute reels that are 100 % built with AI. Now you can see these on my LinkedIn. I posted one recently that we did with Wicked. They are so good. I mean, I could have spent anywhere upward to a hundred, even$250 ,000 to produce the quality that I believe our team did in just weeks had this been five years ago, right? So I have huge cost savings from content production, right? So that's one example. But now what I think my advice is moving forward, now that I know all these things about how my team is using AI, it now needs to be more intentional and there needs to be direct business goals.

14:08So instead of, you know, doing all this work and then collecting and analyzing the money saved, I now want my design team to go in and say, okay, this next quarter, I'm going to help cut the budget by five or 10%. My goal is to save time so I can be doing these other things. Like, I think there needs to be more intention around what is the goal that you want AI to solve for? Obviously it's all about generating pipeline, but what are these different metrics that you can be looking at that's going to tie back to your ROI and starting out with that goal before even going into the tool is really important to define.

14:41Yeah. It's funny because hearing you talk, I could see first with the cinema experience thing, I can see the event marketer, but like event marketers have to, are so tight with budgets and having to prove every part of the budget. So I could see the way you think is exactly how event marketer. And I think more event marketers should, should become up the ladder like you. It's pretty cool to hear. I just wanted to mention that is something just seeing, seeing how you talk. It's very event marketer background. I can't let it go. I mean, it's really, it's the one muscle that I'm like. The experience of cinemas and it's pretty cool.

15:23But also something I wanted to mention is some marketing teams now are dealing with, they have 30 plus AI tools or 30 tools in their marketing tech stack. But a lot of the most effective marketing teams often operate with a smaller stack. So how should marketing leaders decide what tools actually belong in their stack versus what's noise? So our research found that the sweet spot is six to 10 in your stack, which was interesting. I was at a dinner with a bunch of marketing leaders during ad week last fall. And the response to that stack couldn't have been more different ends of the spectrum. I had someone sitting next to me saying, no way, that is way too many.

16:09And someone's like, to your point, I'm running on 30 more tools. Like, how is that even possible, right? So it kind of starts with what are you trying to solve for? I think the best thing to think about it with your own tech stack is where are you finding ways to get rid of the mundane tasks of your team and your marketers so that way they can be working faster and working more efficiently, but also working on the work that they care about? Where are you giving them space to be more creative, to be creators, and really reaching your audiences in new ways? And that starts by trying to deeply embed into your workflows.

16:43So if you can take an assessment of the tools that you're using, I think it's fine to still be in some form of an experimentation phase. That'll never, I mean, look at how fast this technology is changing. You use one tool today, it's wildly different three months from now, right? So I think it's good to have some experimentation maybe happening in one bucket, but companies really need to start deeply embedding AI into their workflows and having that commitment. And that starts with your existing stack and what's going to be seamless from data integration to ease of use. You know, what are the tools that you have already that now have AI features that don't have before?

17:21So if anything, I think it's a hard look at what your current stack is today before you start adding to it. I mean, being in marketing ops, the tools that you want to use have open APIs, easy integrations, are doing things like you are doing at Contentful, which is having customer events where you're launching new products, showing new products off, showing that you're investing in the product. A lot of companies aren't investing in their product and they're closed box products that don't connect well with others. And once you build on that and get stuck in that, it's so hard to migrate off of that.

17:59So you want to start from this or rethink it, especially in the age of AI, where if it works out very well, it's going to efficiency gains, personalization gains, all these things that you could do better as a marketer. It's funny that a lot of, despite all the AI tools available, a lot of brands you still see out there send the same content to everybody and still do the same playbook as everybody else. Why do you think personalization at scale is still so underutilized? And what's the simplest way a marketing team can start doing that today? So if you think about personalization, this is nothing new, right?

18:42Like you mentioned the email, right? if like in every email I get, it should say, hi, Elizabeth, or you hope that you have their first name to, you know, to say hi, Daniel, that's been around forever. That is an expectation. Now, when you think about buying behavior today, a lot of research right now is not done on your website. It's done on LLMs. So by the time that they come to your website, they actually have higher intent. So if you think about it as a journey, you know, it's like I did all my research here and now I'm going to make a purchase or contact sales or whatever is my next action to engage with a brand, I want to make sure that is meaningful and I want to already know things about you to make sure that that stands out very differentiated.

19:24So that is a very easy place to start in personalization. I think this is an area where a lot of CMOs that I chat with, they tend to overcomplicate it. It's very easy to let your brain go to the matrix of every product you have, every persona you need to market to. How do I make that happen across the board? And they make it too complicated. It seems too hard and they don't know where to get started. And the best thing to do is to think big, start small and move fast. So a great example for that is thinking in the highest level of segmentation. Folks that have been to your website before and folks that have never been to your website before, right?

20:02Like these are two just very specific different segments and buckets, but you can have so much impact by understanding where they came from. So another example for you is we launched last year, a brand campaign called feeling so contentful. Well, if you came to my website from an ad that said feeling so contentful and it's all sent out, I want that same vibe to then carry over to my website. I don't want you to then see something completely different. And so it's understanding the data of, of where you came from? What are those multiple touches where you get to the final resting place to making sure that you have that connective tissue?

20:40And then ultimately, what are ways that you can leverage personalization in your website and also the simplest ways? Another example here is take the contentful on contentful or at the cinema example that I gave you already. One of the things that we do is taking geo tagging and we update our eyebrow header at the top of our homepage by city. So for example, if you're in Austin, it says, hey, Austin, come join us on March 11th, what have you, for Contemplate the Cinema. And this was, again, just something we tried out like, hey, the eyebrow, that's not much real estate. I can give that up and see how this goes.

21:14And we saw a 51 % increase in RSVPs to our events just by geotagging and updating our eyebrow header for our top five biggest cities. So I think that's my main message is it can be so simple, yet so impactful. and so how can marketers let the complexity go and really start to reel it back and and really think big and start small i like the way you think about that because the the end goal is you want every experience to be hyper personalized hyper uh feeling like it's one-to-one but there's so many low touch things that you could start tackling first till you get to the high touch maybe the high touch thing is a longer term thing, but the way you think about it is super smart.

22:01I also think another one that I know you talk about is content repurposing and content reuse as like one of the biggest efficiency gains, I think, and efficiency gaps we can bring down with AI. So most markers out there, most companies are sitting on thousands of pieces of content. So what does it look like a modern content reuse strategy? What should I be thinking about to make sure I get, we call it internally the sawdust approach where you created this chair and there's all that sawdust and what can we do with that sawdust to keep making new and new things? So what's something you, how do you think about it?

22:48You're sitting on a gold mine of customer data and it's doing nothing for you. That's not a data problem, that's a tools problem. Customer.io connects your data to every channel you already use. SMS, email, push, in-app. Then it uses AI to build workflows that used to take weeks. Every message feels like it's written for that customer, not blasted at them. 8 ,000 plus brands already made the switch. Learn more at customer.io slash TMM. So I look at it more of what are the tools available to us today to give us more insights that we've never had before. I was on a CMO forum on Friday and there was a guest speaker talking about similarly like LLMs, where are they pulling information from?

23:37And if you look at top-cited places like Reddit, for example, they gave this example of how to run, let's say, your top blogs against conversations happening in Reddit and have a prompt suggest where are the gaps? Like where's the conversation happening and how can you then be updating your content to do that? And I think that is maybe my biggest piece of advice. It's not creating net new content, it's updating the content you already have. Another great thing you could do that's very tactical is if you think about blog content, if LLMs are reading content for bullets and lists, we'll start by having a, maybe a table of content somewhere on your blog or having a TLDR.

24:23Like what are different ways that you can also be not only serving humans, but also bots in those audiences. And it's really about your existing content that you have and enhancing it in ways that's going to help your brand show up in different ways. And a lot of the LOMs want to see that this is like updated, new, refreshed content. Like this is March, 2026 edition of this piece of content, not something that you put out in 2025, which you could still do this so much evergreen content out there that still works that you just could add a little flair to, or add, just add a couple of touches to, to make it, new.

25:05So I think that's super important because I think most people out there get stuck on, I need to create a new piece of the content. I need to create a new piece of content. I need to, and that's not, you already have so much good content out there. How could I just keep that content being distributed for a longer period? Especially an easy example is, is blogs on that, right? I mean, even if you just started with, what are your top 10, 20, 30, whatever, most visited blogs just start there. How can I just start refreshing that content. And to your point, it might just be a quick update. Maybe it's just a date change.

25:38Maybe everything is still relevant because it's more thought leadership versus it's more product specific. So there needs to be some tweaks, but being able to reuse the content you have is really important. And going into, we talked about some good advantages of AI, But I think that there's also a negative of overusing AI and people using AI too much in their marketing. So where are marketers currently using AI too much and where are they not using it enough? So I would say we obviously are seeing it in the content creation. And I think there is a big human element that is being lost there. When I take a step back, I like to think about what does AI not have?

26:29Well, they don't have lived experiences, judgment, this human effect. It could be trained to maybe think like a human or talk like a human, but in terms of our own lived experiences, that's what we carry forward. So I think that is what is missing is really provoking this human emotion to really tie back to the content is probably problem number one. Now, the second thing is more has to do with data and analytics and being more insightful into how your content is performing or using different tools where AI can unlock access to feedback that you've never had before. A great example of this for us is Momentum.