How Your Brand Can Show Up On AI Search in 2026 with Tifenn Dano Kwan, CMO of Amplitude | Ep. 377

0:00Proving your marketing ROI shouldn't be a guessing game, but when results are hard to track, it's difficult to know if your marketing strategy is working. Meet CallRail. In a few clicks, you can connect every conversation to the exact campaign that started it, so you can focus on what works and drive better marketing outcomes, all in one platform. Try it free today at callrail.com slash prove it. 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 up welcome back to another episode of the marketing millennials today i have a really cool guest it's tiffin she's the cmo of amplitude um fun fact i used to use amplitude at service titan so i've been a previous customer um and excited to chat uh but tiffin i'll let you get into how you got into marketing and then also describe what amplitude is for people who don't know Thank you, Daniel.

1:20I'm really excited to join your podcast. I am Tiffin Denoquan. I am originally from the west coast of France in a place called Brittany. and I have been living in the United States since 2011 and literally spent all of my career in tech and B2B SaaS software, to be more specific. And I am thrilled to talk about Amplitude and all the things that you bring to our marketers and audiences. Amazing. I want to first, let's just describe what Amplitude does so we can get that out of the way because I think it will help shape this conversation. Of course, of course. So when the founders, there were three founders, created Amplitude, it was based on a simple observation.

2:19They were working with a lot of product teams and the rate of failure of products was outstanding. In fact, 95 % of products typically failed. And they wanted to change that. And it was at the time, and it was about 14 years ago, where the digital engines were starting to pick up quite a bit. And they wanted to help product teams build better products and digital experiences. And it took off. It took off so fast that the company actually went public at the end of 2021, early 2022. too. And since then, it's just been a journey to help companies grow through the power of digital products and experiences.

3:08That's a good explanation because I want to go into a topic that everybody's thinking about right now. And I want to first start with the question of what do you think is the simplest way to explain AI search to a marketer who is so focused on Google first right now? It's a great topic. I mean, if CMOs don't look into this topic right now, I think they have to. This is a very critical one. And historically, to your point, all the marketing teams in the world have been focused on capturing traffic, web traffic. There's multiple ways to capture web traffic. You do it through either direct traffic, which you can do as an outcome of activities you run, multiple activities you run, and the awareness that you're driving.

4:08You can do it also through search engine optimization. That was, I would say, the more default typical way of going about traffic. there's typical ways to drive search engine optimization. There is an organic way, and then there is a paid way. And that is because Google created an algorithm that actually created this process, right? And they built a moat around it. So the more you pay, the more you get ranked, and then people can find you. And that as they find you, they go on your website and then you drive click-throughs and hopefully you're optimizing your conversions. That was the way to go until LLMs came into the picture and disrupted it.

4:59So one of the things I've heard a few times this year in particular is SEO is dead. Well, guess what? SEO is not dead at all. In fact, when I look at the data, and I literally looked at the data just before our conversation, search engine optimization accounts for 80 % of my team's traffic, literally. And LLMs account for a mighty 1%, maybe 1.09%, depending on the day, 1.09%. So, the data doesn't tell me that LLMs or AI search is actually taking over the world yet, but it is growing. And what we're seeing, basically, is a diversification of search. There is the traditional, I would say, search engine activities, and then there is the AI search.

5:57It's called also AEO or AIO, depending on who you talk to. Where I see the market going is just an opportunity to expand on our channels of acquisition. It's a formidable way for us to just test and experiment this new channel of acquisition. and we're seeing interesting stats. So, of course, we're seeing some declines in traffic because a lot of the audiences are just using this channel to do some research, but they don't click necessarily. So where do they go if they don't click? Well, they go and they research. And then they end up going back into your website, but probably at a later stage when they're ready to sign up for something.

6:50So, what I love about the way we're going with this is that it's pushing our teams to do much more thinking, analysis on what's working, which channel of acquisition is going to become more critical of others. And then we are observing. We're observing also what I call the LLM wars and the SEO wars between Google and various LLM companies. And I would say that we need to watch that space. Google isn't dead. Not even close. They're actually fighting back and just making sure that their algorithm is winning over the mode and preserving their mode. So I love what I'm seeing right now. It's just super exciting.

7:38Many CMOs like myself have conversations and I'm sure are just staying close to their digital and growth teams for that reason. yeah you know what it's so funny like when people say seo is dead um the question i always go back to is where do you think these ai is looking for getting the content to be fed in their um in their in the lm they're looking at your blogs they're looking at your website and it's exactly the same as what you were doing with seo is optimizing your website optimizing your site optimizing your blog answering questions so you show up it's just how you show up is different and it's just funny how people think that okay now um seo is dead but these these ai agents or ai bots or ai are looking through your website trying to figure out if this person is typing in this prompt it does does it answer your question the question on your website or not that's right and that's what it's really doing and um if you're that's how you should you should we should have been optimizing our site to answer these questions anyway as marketers for the longest time.

8:46That's exactly right. So I think, first of all, I think that citations is going to be the way to just look at the world right now. Where are you cited? How often? How good is your ability to be cited? So it's based on the prompts that you are generating. And more importantly, it's based on the content that you're generated. So you have to be cited. But the web is a vast environment. It's a universe, right? So you have so many opportunities to be cited in so many different third-party sites and places. But the key in my view is to optimize for citations. And it means that the content that you're creating through its various forms, it could be blogs, it could be interviews, podcasts like today, it could be articles, it could be FAQs, it could be all kinds of content.

9:44It needs to be cited and needs to be positioned in the places that matter. And what's going to be very interesting here is that it's going to help us identify the places that are just really growing and maturing. Reddit, for example, is a clear place where conversations happen, where a lot of content gets exchanged, where a lot of opinions get really captured. So that is a very clear area. G2 is also a great place when you have a lot of reviews, especially if you're in the world of software. G2 is a fantastic place where we see a lot of conversations, so we want to be cited there. But there's so many other places.

10:27LinkedIn, of course, X, all those places truly matter. And, of course, your website. I want to know, so you've been talking to a lot of CMOs. It's a small part of your traffic, but not a big part of your traffic. But it's going to start eating a little bit more as this year goes and next year goes. So when you're planning your marketing strategy for 2026 and marketers planning their strategy, what are some things that they should be thinking about to be optimized to show up in LOMs? What are some things that they should be not missing out on doing? I think that one of the things that's going to become critical is the content that you feed to the LLMs, right?

11:16Why? Because LLMs are powered by content that feeds research. People go to LLM to research. They do a lot of research on it. They ask questions. The questions give them answers. The answers are part of a big element of research that they do. And because we see a decline in clicks, we have to make sure that we diversify the way we get cited, the way the content is created, which means that the amount of times your company's brand is mentioned and cited is critical, which means that it's putting now the emphasis on brand awareness. It's one of the biggest shifts that I'm seeing right now, which is fascinating and incredibly exciting for all the CMOs right now, is the power of the way your brand gets really referenced and cited.

12:16It's a complete reinvention, I would say, of the function. If you think about it, especially in the world of SaaS, all of the CMOs were mandated over the last 10, 15 years to focus on performance and pipeline. That was the number one, still is the number one goal for a lot of us out there, which is making sure that our marketing team is an engine of growth, profitable growth for the company. And you could see, I would say, a lot of big differences between the brand CMOs and then the performance CMOs. There's a lot of those conversations happening. Those functions are getting blurred right now.

12:58The focus on brand, brand is becoming a performance engine, finally, which is really incredible. For the longest time, you could hear a lot of founders or board teams questioning investments in brand. Why? Because we could not measure the actual impact of brand on revenue. It's been the same conversation for now a couple of years. It is the first time now that we have the ability to actually track the impact of brand. And we do it through critical tooling right now called AI visibility. Everybody is talking about AI visibility. And the reason for it is that AI visibility gives an ability for CMOs to track the impact of brand and to actually show it and compare it to other competitors.

13:55That is an incredibly powerful tool. And I can tell you, just looking at our own recent AI visibility launch, that there is massive, massive demand for those kind of tooling. So I would say the content is critical, the infrastructure of your web is, and equipping yourself with an AI visibility tool will be critical because it gives you an ability to track how your brand is getting cited and mentioned out there. Yeah. And I love this shift of the pendulum because it always shifts the pendulum. Like 15 years ago, the pendulum was brand because the trackability of performance marketing was just coming up.

14:41And now the pendulum's shifting back to brand. But I think we need to just go back to the middle where all marketing is performance marketing and brand is performance marketing and demand gen is performance marketing. And having a tool like AI Visibility where you're tracking how you're mentioned. And like you said, the only way to track mentions is people talking about your brand name and your brand name showing up in conversations. Like Reddit, your brand name showing up on social, your brand name showing up on blogs. And you could do it for multiple ways. You create great content that's shared on multiple blogs.

15:16You create a great product that is getting shared by a lot of people like GT, you mentioned. All this stuff matters even more now, having a great product, showing up in conversations, which I think is super cool. So AI visibility, could you explain a little bit of what it's tracking? And what are some new metrics you're looking at when you look at... What are some new metrics we're looking at when we're looking at AI visibility? Of course, of course. So when you have an AI visibility tooling like Amplitude, it gives you an ability to quantify how often your brand appears in AI responses. And that is done across all the most important LNMs out there and literally hundreds or thousands of prompts.

16:07So it actually creates a score that compares you to others in your own space, in your own category. So that's incredibly powerful because it's literally based on prompts that are out there in the market. So it's literally aggregating that and it's creating a score. So, if you make this a no-star metric for yourself and you incorporate this AI visibility score into your ability to track the performance of your brand, then you become credible. You have a seat at the table and you're able to say on a regular basis, this is how we're scoring against our competition and why we are scoring. So, you have an ability to track traffic and ROI more easily.

16:54So it means that you have an ability to track your AI-driven visits. Those are driven by LLMs to conversions and revenue outcomes a little bit more easily. So it's just another tool in your arsenal to just build, I would say, that sort of funnel visibility all the way down to revenue. And I think what you said too is now it's giving you the ability to put investment into doing more brand stuff, doing more content stuff, which was hard to really quantify. I mean, you could do it a little bit with SEO, but now you can quantify it more by saying we show up here. And also another thing that's going to happen, you mentioned that Google search is not gone and all these searches are gone.

17:44But a lot of LLMs is zero click. So people are going to search Amplitude in an LLM and then they're going to go to Google and go to the website and search Amplitude. So brand search, you should start tracking, like, are brand search mentions going up over time if I show up? That's right. I would say because the conversations are happening outside, you see? Yeah, exactly. So you have to find a way to correlate it back. So the question is, how do you do this? If your traffic's going down, should everybody panic? Well, the answer is no, because everybody's traffic is going down. What I think matters is how you convert in general.

18:24So one of the things you should start to create is a sign-up mechanism. Just create a mechanism beyond your web traffic to ensure that you capture your top of funnel. It could be a free trial. It could be a starter offer. Find a way, a mechanism to drive a sign-up mechanism. And check if you're seeing an increase in sign-ups. That increase in sign-ups could be an indicator that actually people out there are talking about your brand and your company. It doesn't mean that they're going to go necessarily and visit every single web page you have. But if they sign up for your offerings, that should be an indication that it's working.

19:14So the AI visibility is just an indicator that you have enough content, you're part of clear conversations, you are being cited. And those citations create an ability for your company to be of interest. you know, cited for research for customers who are looking for your company's offering. And you have to find a mechanism to capture that interest. So what matters most is that sign-up mechanism that you need to create or that caption mechanism that you need to create. So it's pushing, I would say, the impact further down in your funnel. You don't just look at web traffic, You look at signups and capture mechanisms.

20:00As marketers, we're all trying to answer the same question. Which investments are actually driving revenue? And too often, the honest answer is an educated guess. CallRail replaces that guesswork with proof. With the help of AI, you can connect every lead to the exact campaign that drove it, analyze your conversations, and capture leads around the clock. That's why over 220 ,000 businesses use CallRail to market smarter. Stop estimating impact and start proving it. Try CallRail for free at callrail.com slash prove it. Yeah, so what I'm like hearing too, because this has been happening for years.

20:44Besides, I'm like, word of mouth has been happening for years. Conversations that you could not track have been happening for years. it's happening on social and dms it's been happening for the last few years now lms are and just other conversations just you talking to a machine not to your friend but people are still going to talk to their friends outside of it so i i i like that so what i'm hearing too is like marketers should start thinking like how can i make my website as frictionless as possible as scannable as possible to get people to the finish line faster than other people are doing.

21:26It's always been the thing that we should be doing, but now it's even more important because people are at the end of their phase of when they come to your site, not at the beginning of the phase a lot of the time. That is correct. I would say you should scan for your web vitals and just make sure that you have a very strong head of digital and web who is going to just re-architect the WhatsApp capabilities that you have very strong engineering support also for your website. So you make it not only L &M ready, but SEO ready. So I'm one of those who believe that you have to actually invest in both at the same time.

22:06Don't account for Google's power to bring their mode back. They're not dead. They're actually fighting really hard to get it back. And you have to trust that there's a lot of money behind this company to figure out how they're going to fight back. So at this point, I'm just looking just at the data. And the data tells me that SEO still accounts for 60%, 70%, even sometimes 80 % of my overall traffic. Even though the conversion rates are going down and LLMs have much higher conversion rates, it's still a significant part of how I'm able to capture demand. That part is not going to go away. If it still is, what companies need to do, the CMO and the marketing team, is just really monitor how each channel of acquisition is responding to it and have the appropriate strategy in place to capture the interest from those channels.

23:10And I think it's also pushing us marketers to rethink how we thought about metrics like five years ago. We used to care more about impressions and all that stuff. And now, like, what is quality coming to our side? Are we showing up like with quality instead of like quantity? Like quality is going to be matter even more now than quantity. That's right. I'm going to put it out there for your audience. I think there is a business to be made around the way to score brand, period, not just AI visibility. AI visibility is one aspect of it. But one of the things I'm discussing right now with my CMO peers, I literally had a conversation with the CMO of DAXCO yesterday about it, is how do we score the health of our brand?

23:57It's not a simple answer. You have multiple, multiple ways of, I would say, tracking this. The way we do it today is we have a composite score. We have a brand health score that is made of multiple metrics. We have a PR metric. We have an AI visibility metric. We have a traffic metric. We have a social media metric. We have a lot of different ones that get weighted to create a composite score. There's a business to be made out there. for whoever is listening, because this is what's going to happen next, is that when you diversify your channels of brand and awareness at the very top of the funnel, how do you know across those various channels, social, web, LLMs, SEO, and whatnot, what's working, what's not?

24:52I'm just putting it out there, but this is a conversation that I'm having right now with my CMO peers, and I'm just inviting the conversation. I love that conversation is happening because that just proves even more why to invest in brand. But it's also like what levers do we need to invest in? How are we best? And then different stages of company. You said like Amplitude has pretty well known in the market, I would say, in the MarTech space. Like if you're a challenger brand, it's a different animal of how to get your brand in the market versus someone who's been that strong force in the market for a lot of years.

25:36So I think that's super interesting. I also want to go into, what are some signals that you have been seeing that matter most for LLMs when deciding, LLMs are deciding which brands to surface in their prompts? What are some signals that are you seeing? I think that you have to look, first of all, on making sure that you're sighted consistently. I think it's the quality of the content that you have. AI-driven content is phenomenal. This allows your company to drive those citations at scale. And actually, I'm seeing in the data that the conversions are actually fantastic coming out of it. So whoever just stays skeptical about AI-driven content, I would invite them to lean in more and test and see how you can drive this.

26:36So that is a strong signal in my view that AI-driven content is going to very rapidly increase to do it. So what kind of content, what kind of elements of content work best? What we're starting to test is FAQs work a lot because whenever you develop your own point of view with very clear answers to questions being asked by an audience, this is where you have the best chance of differentiating and cutting through the noise. So test the format of your content a lot. FAQs work great. Like blogs continue to work very well. A lot of video content is coming out there. And this one is also really great.

27:26Virality is also a strong signal of interest. The question is more like, how do you convert virality into those LLMs? But virality sends people, right? Sends people to do more research. So those elements are also very, very critical. These are the signals, in my view, that speak to a healthy brand. Is it a brand that's going viral, that gets cited? Is it creating authored content through formats that work and that are repeated and consistently out there? One thing you were saying about the FAQs, which I think is great, is if you go look at a great D2C or e-commerce site, they have to be great at FAQs.

28:07They have to be great with reviews. They have to be great. And if you go look at, if you do product searches, a lot of things that come up are reviews and questions answered of, like, for example, I have a baby. I was, like, searching what is, like, the best high chair for longevity for a baby in my LLM. And I came up with a brand. And then I went on the internet to just, like, verify if the LLM is correct. But I still like if you could show up and it showed me like five reviews of people talking about this, I could use this from zero newborn to nine years. I could use this. So just like you saying, the FAQs, the reviews, sites, the way people are talking about you matters so much when it comes to AI search.

28:53So test out the content. Evals work really well. Benchmark tooling works a lot as well. People want to try. People want to benchmark themselves. They want to see how they compare to others. They want to have very quick answers to their questions, which is obviously the most important. And it sounds obvious what I'm going to say, but you get what you feed into the LLM, right? So the LLM is going to be only as good as the content they're going to scan for, right? So the more you experiment with content at scale, the more chance you have to actually teach the LLM your own language. So you've got to have a very clear opinionated voice so that it becomes your language.

29:43Question for you. Where in the current buying journey of customers you've been seeing, do you think AI will have the biggest impact? When you think of in a current buying journey, where would AI have the biggest impact for a buyer? That's interesting. It also depends probably on the type of business. Are you an enterprise business or are you a small-sized business? I actually think that the research happens typically at the top to middle of funnel. People research your credibility as a company. They research the tools that you have available. That's why evals, trials, all those things really matter.

30:29And they formulate that decision. And I do think that this is happening very much at that stage, in my view. If you go into a more complex sales cycle, which is an enterprise sales cycle, and you have to speak to humans and you have to have proof of value, proof of concepts, all those kind of things, that becomes a little bit more complicated and that you probably need the human in the loop at this point. But if you do some, I would say, first stage research on who the company is, what the product is about, what kind of typical features and capabilities a product could offer, you can find all of that information through LMMs right now.

31:12and i one thing so i wanted to ask too so you you using the ai visibility tool you see like we're showing up in these prompts um and um what are what's like the next step like if you see all these prompts like what is your next step to as a marketer to be like i need to tackle this this and what should i do next do i go after like the highest like queried prompt do i go after the next one? How do you think about, as a marketer, what should I do when I see I'm showing up here but not here? I'm showing up there. This branch, my competitor is showing up way more than me over here. How do you tackle that?

32:00That's an actual interesting question. I don't know that we have the answer for it. We're all testing right now to see which are the channels of conversations that matter most. So if, for example, we say we want to make sure that we feed Reddit as much as possible because that's going to be a feeder for a top citation, it's going to work. I think the strategy my team is having is that we want to be cited in LLMs. We want to be cited in AI overviews from Google. Well, we want to be cited everywhere, right? Like we are really working on the algorithms of all of those places. So that's why I'm saying don't discount Google.

32:47Don't. That'd be a mistake. Like you want them to cite you. Their AI overview is incredibly powerful. Not necessarily everybody is going to the chat GPTs or the cloud. You just need to keep being diversified and learn the type of content that works best for you to be on the top ranking in citations. I think that what we're going to see with our audiences is that they typically have multiple tools at their disposal. If they really want to compare, they're going to go to Cloud, and then they're going to go to ChatGPT, and then they may end up on their mobile phone directly into a Gemini. So you got to have to make sure that you cross-pollinate and triangulate.