#57 - Meet the Startup Revolutionizing A/B Testing With AI

Are you tired of spending way too much money on A/B tests? Alison AI can help you out.

Okay, so having this, uh, 2 people with the same, uh, strengths, the same, uh, weaknesses, the same mindset, it's not always the best thing to do, right? You need some balance and you need somebody to challenge you and somebody to, uh, complete you. The initial concept, the initial mission was let's replace A/B tests altogether from start to finish. Let's replace All of it by using AI.

What is the statistical significance across all these videos? That's what, what's working, what's not working. And I think that is the hard problem that you have solved.

Hi everyone, welcome back to this episode of The Tech Optimist. We have a Meet Startup episode for you today. We talk with Alison AI. Now, the host for this episode is a newcomer to the show. Meet Brooke Stroud. He's a senior principal here at AV. And, you know, the person of the hour is Asaf Yanai, co-founder and CEO of said company today, Alison AI. And then of course you'll hear my voice scattered throughout the episode today. My name is Sam. I'm the guidance or narrative writer for the show. Now we have a long one for you today, but I urge you to listen to the whole thing because it is very, very fascinating. The more we get into it, the more stuff just comes up here and there. Brooke and Asaf today have a really cool conversation about Alison AI's technology, how it's innovative to the world of marketing, and, you know, how it might apply to your company. So they talk about the issues in how expensive and how fast-paced the sort of landscape is for A/B testing and for video marketing and everything else, that they incorporate AI into, uh, their sort of marketing suggestions with your branding. Um, So this company does this in a really fascinating way. We're going to learn so much more about AI and machine learning today in this episode, but also we're going to get into, you know, some success stories with Alison AI and how they truly have helped companies really revolutionize their marketing. So super excited to sort of share it all with you. But yeah, hang tight, sit back and enjoy this episode. Let's get into it. As a reminder, the Tech Optimist Podcast is for informational purposes only. It is not personalized advice, and it is not an offer to buy or sell securities. For additional important details, please see the text description accompanying this episode.

Welcome to the Tech Optimist Podcast. My guest today is Asaf Yanai. He is the CEO and co-founder of Alison AI, a company that is revolutionizing creative workflow through AI and data so that companies can stop spending and wasting so much money on A/B testing. Asaf, welcome to the show.

Thanks, Brooke. Thanks for having me, and thanks for inviting me. It's a pleasure.

Well, let's dive right in. Well, when we met up, uh, and we were discussing Alison AI and In addition to being blown away by the technology and the product that you showed me, we opened up your laptop in the coffee shop and we just went right into it. One of the things I came away from that conversation with was such a deep admiration for the enthusiasm and level of just this can-do mindset that it was magnetic to some extent and also just so powerful to hear you speak about your vision for the company. And get the sense of like what you described as being a boxer to some way about taking punches and bouncing back. And so, I would love to hear like on a particular day or week where it's, for whatever reason, to use that analogy, there's a lot of punches being taken. Are there certain things that you do or practices you have in place that over time have helped you cultivate and build this mindset of being so resilient and having this can-do energy? Because, you know, I imagine that's not just something that you can just turn on and off. It's probably something you work on and something you also have learned as a child.

So absolutely. So first things first, I mean, I'm I'm not new to startups and it's not my first company, not my first startup. And I think that really builds a lot of resilience because you're, I wouldn't say used to it, but you've done this before. Maybe not the same, maybe not exactly the same, but you've done this. You know what you're supposed to do. You know what the right path, the right trajectory looks like. I've experienced, you know, those exactly the same punches, maybe in a different company, but you've experienced this already. And I think this is a very good thing that builds resilience for me. Like, I never get scared. Seriously, like never. And even my wife, you know, she thinks I'm crazy. Like, how come you're never scared of anything? And I'm never scared because I automatically put myself into action. So not scared, but actions. How do I get out of it? What do I do now? What's the best course of action? Like, what's step 1? Or even what's step 0? Where do I begin? So I'm not thinking about how difficult it is and, oh my God, how challenging. Not at all. I have a situation. I try to first solve it, handle it, tackle it, and then move on. And I always say to my employees and my team, I always say to them, look, going this marathon, but winning the marathon, right? Having a good positive outcome for your company. So I always say that we're not conquering Rome within a day, right? And Rome wasn't built in a day. It takes time. So we need to accumulate consecutive small wins. Don't go and aim for the big win because it's too big. It's too far away. You can't even see the path to it. But build the steps, small consecutive wins. They will get you to this big win, the holy grail that you're looking for. And every founder is looking for a different thing. Well, all of us are looking for a successful company, I think. But some are looking more in the financials, some are looking more towards scale and it excites them. Some are looking to just be international or those kind of things or solve big problems. But in any case, small consecutive wins, I think, will get you there and will build this resilience. But like you mentioned, for me personally, it's also my personal background and it's also my childhood. I grew up in a highly, highly, highly achievable, highly professional family. So both of my parents are entrepreneurs together. They founded a company 35 years ago, bootstrap. I'm, you know, the firstborn. I'm the oldest. And I've kind of seen, you know, I've experienced firsthand how this thing evolves. You know, imagine what it is to to have a bootstrap company 35 years ago and grow it to international standards, international brand standards. It's a roller coaster that has a lot of, you know, again, wins, challenges, punches altogether. And I think I've seen how my parents are overcoming challenge after challenge after challenge after challenge not giving up, always keeping their minds very positive, very mission-oriented. Okay, what do I do now? I'm not afraid. I'm not scared. This is not a challenge. I need to do something. What do I need to do now? And if you put your mind into it and you train your mind into it, so I think this is also something that really built me and the experience and looking at role models as my parents. it was phenomenal for me. And I even joke about the fact that I'm 37 and I have 37 years of managerial experience or almost 37 years because I've soaked everything in from the family. Because when you're a couple, mom and dad, right? Having one child or a small family and you also run a startup, you run a bootstrap company, You know, you bring the shit home. You know what I mean? You bring everything home. You bring the challenges, you bring the wins, you bring those punches we talked about, the misalignments between each other. You bring it home and then you talk about it. You fight about it. You celebrate all together. And I think, you know, looking at it year after year, challenge after challenge, win after win, achievement after achievement, This really, really builds resilience and shows you how, you know, the bigger picture, the marathon is more important than every single mile in it.

Wow. Very different than a marriage and being co-founders with your wife. It does remind me of just thinking through, like, figuring out who to work with as a co-founder, because that's, you know, a different type of marriage and through business. And also wanted to say that it's amazing that both your parents co-founded and built a highly successful company over 3 decades. But I wanted to ask you, when you were thinking about who to build Allison with, were there particular things you either learned from your parents or just from your previous successful startups that helped you, you know, narrow in on traits or what you wanted to see in your co-founders, whether that be complementary skill sets or extreme alignment on work ethic and vision?

So I think one main thing that I took from my parents when it comes to this is see how your partner or co-founder complements you. Okay, so having this, 2 people with the same strength, the same weaknesses, the same mindset, it's not always the best thing to do. You need some balance and you need somebody to challenge you and somebody to complete you in a way. So this is a very big thing I took from them and their experience and how I was brought up. But from my personal experience, I took, I think, other factors into consideration. And you've just mentioned a few of those. So it's work ethics. Absolutely. It's having big dreams and aspirations because if your co-founder doesn't have big dreams or big aspirations, so those big dreams and those aspirations and the vision will always come from you. And there's a lack of this complementary aspect that I'm thinking about and I'm speaking about right now. So work ethics, absolutely, it's one. 2 is, as I said earlier, be experienced, but also with the environment that we're going to run on. So we run on a high-paced environment, startup environments, this organized chaos. You know, you hire, you fire, you grow, you open new offices, you attract new customers, you have challenges, all of this. And I think having somebody with you that has experience and also has, you know, can keep his cool and he's calm and he's thoughtful and responsible. I think it's highly important for a company sustainability and the company's growth. These are the main things I was looking at. And by the way, I was dating, I call it dating. I was dating a lot with potential co-founders before I met Kobe and before we decided to get married in a professional way, obviously. But yeah, I was interviewing, I was meeting a lot of different ex-founders, new founders, entrepreneurs at heart. And I think what triggered me with Coby is that he told me, look, Asaf, I'll be honest. My professional passion is AI, machine learning, tech. That's what I love doing at work. In my personal time, in my hobbies, or just, you know, with myself, my personal passion is video. I want to do— I want to merge those two things. And when he said it, I kind of— oh, it was an aha moment. And I felt like, okay, that's the perfect match. Somebody that brings a lot of experience around tech, but has this fire and has these big aspirations. when it comes to his personal goals and personal achievements around video. This was highly exciting for me. Plus, again, Coby is a phenomenal tech expert. I call him a superstar in many occasions. He's done different paths with— by the way, most of the startup companies that he founded or joined, became unicorns, most of them. So obviously he was also successful. Obviously he had a lot of challenges. And I looked at the bigger picture. I looked at this person as a whole, and I thought that we complement each other. And he also brings a lot of things that are important for me when choosing a founder.

Well, let's dive in and talk a little bit about Alison AI. One of the very overquoted, but I think relevant things I think about when looking at your technology was the quote that I think David Ogilvy probably popularized, but someone, John Wanamaker, famously said, half my advertising spend is wasted. The trouble is I don't know which half.

Yeah.

And when I think about the opportunities that you're providing to customers and the ability to deliver this Very granular level of marketing intelligence, and particularly in the world of digital marketing and video. Talk to our listeners about that aha moment, what you saw as the opportunity then and today, and just give us a little bit of background on what Alison AI is solving in terms of pain point for your many customers and many large multinational customers.

Okay, really quick, before we dive into Alison AI's technology, I want to provide a little bit more context on the company's values and morals and, you know, the, like, the just main lessons and I guess vibes you could say that they run off of every day. So on their website at alison.ai/about-us, they have a really good page that really summarizes that, and I'm going to sort of read that and share that before we get into their technology, because I think it'll be really powerful to understand the people behind the technology first before we start to unpack the really cool technology, um, that they have developed. So, we are redefining creativity. Alison AI empowers every step of your creative process. Our platform's transformative AI technology revolutionizes advertising strategies, guaranteeing your brand stands out in today's challenging landscape. Now, so here's their mission and their sort of principle statements here. At Alison AI, we believe that exceptional creatives come from data-driven decisions, not gut feelings. Our mission is to empower professionals with AI-driven insights and data science, enhancing brand performance globally by replacing intuition with precise data analysis. We provide tools for teams to create impactful video content confidently. We are revolutionizing the creative workflow with evidence-based decisions and cutting edge technology. Uh, first time I read this, I really liked that about how, you know, they're, um, replacing the sort of gut feeling and replacing intuition with precise analysis. I think that's something that, you know, um, a lot of companies don't do. So I think this is awesome. As far as principles go, we are guided by values that inspire our team, shape our strategy, and ensure we deliver exceptional results. These principles form the foundation of our business and foster a serve innovation and excellence. So those principles being customer obsessed, pioneering change, collaboration and teamwork, and excellence and quality. We place our customers at the heart of everything we do, prioritizing their success and earning their trust through our actions. We strive to break boundaries, exploring new ideas and innovations to make our solutions more efficient and powerful. We foster a culture of collaboration, valuing diverse perspectives and working together to achieve common goals. We are committed to delivering the highest quality user experience, striving for excellence and continuous improvement in our products and services. Here's some just awards that they have on their website as well. So from here, we're going to do our first ad to sort of get that out of the way. And then we're going to hop, you know, right back into the interview where Asaf really starts to break down their technology. And how this is innovative and sort of how he and Kobe got to this technology, right? We heard a little bit about both of their passions on you know entrepreneurship and sort of video and AI and machine learning. So now we're gonna figure out how all of these are sort of tied together. So you know the drill. Hang tight.

Hey everyone, just taking a quick break so I can tell you about the AI fund from Alumni Ventures. Alumni Ventures is one of the only VC firms focused on making venture capital accessible. Accessible to individual investors like you. In fact, Alumni Ventures is one of the most active and highly rated VCs in the US, and we co-invest alongside renowned lead investors. With our AI fund, you'll have the opportunity to invest in a portfolio built entirely around advancements in AI. This fund consists of 15 to 20 investments in multiple fields where AI is making a huge impact. Including areas such as machine learning, healthcare, education, transportation, and more. To get started, visit us at av.vc/funds/ai-fund. Now back to the show.

Absolutely. So as I said, as I said earlier, I was, you know, my background is specifically within online marketing. In different companies, big, small, you know, different sub-verticals or sub-markets within the online marketing. And I've worked with the biggest names in the industry in my career. So from, you know, the big CPG companies to the huge banks that we have around the world to, you know, online-first companies that have grown so much, you know, from Tinder to Waze and OkCupid and Uber. and HelloFresh and those kinds of companies. And what I've learned is exactly the phrase that you've mentioned earlier, is that 50% of our marketing budgets are going to waste. That's a fact. Now, let's try to see what we're lacking on visibility towards those 50%, where we're spending the 50% on hunches. on guesswork, on tests, on things that are not proven yet and we're trying to figure out. And what we've learned is that A/B tests, A/B tests and the operation around A/B tests when it comes to online marketing specifically takes huge amounts of budgets and overhead like headcounts. So in order to fuel this A/B test monster machine, You need a lot of people, you need to create a lot of different creatives, you need to dedicate testing budgets, and it's heavy, right? It's heavy and it takes a lot of time to get a concrete insight, if any. Sometimes you're just shooting in the dark. And then understanding that A/B test and the creatives around online marketing These are the only black box in today's marketing, online marketing. And everywhere else, media placements, audiences, we have enough data, we have enough visibility, we have enough that helps us optimize. But everything around creatives, there's absolutely zero visibility. And what's happening is that all the social platforms, all the media platforms, they take advantage of it. They know it and they take advantage of it to increase their own revenues by masking or not showing you what's working behind the hood. So you can't really optimize on your own and you would 100% need an external tool to help you do it. And this is how we came up with Alison. And the initial concept, the initial mission was let's replace A/B tests altogether from start to finish. Let's replace all of it by using AI. And we've done it. I mean, we've done it even in our, you know, before we completed our first year, we've successfully replaced A/B tests. And, you know, our customers are reporting back to us. And as you said, they're multinational, you know, big, huge brands or companies. They report back to us that around 50% of their marketing overall marketing budget was saved. 50%! It's crazy. It's madness. It's huge. You know, nowadays, the marketing cost line within your PNL—it's the second largest cost line. Now imagine you cut it by half, by 50%, just by integrating and you know optimizing and augmenting your. current workflows and replacing some of them, you know, you, you can create such a bigger efficiency boost to the company, and that's just about cutting costs. The other side of the equation is uplifting performance, increasing actual revenues. Now, that's the reason why I call it the dual effect— how you reduce costs and increase revenues at the same time. So you basically increase this delta, you increase this gap between what you're spending and what you're generating. And that's, at least in my world and in my mind, that's extreme efficiency. And understanding that AI could yield and generate and onboard this extreme efficiency for a lot of different companies, this was my aha moment. It's not just an A/B test tool replacement. It's it's it's it's re augmenting and and and and putting revolution revolutionizing your company, putting your company in a completely different state, mental state, headcount state, cost state, production. Everything around it is gonna change. And I think this is the big premise around Allison, and I'm very happy that you know we have a lot of customers that are already. seeing those huge uplifts and this extreme efficiency I'm talking about. And honestly, it just gives me more fuel, gives me more passion to see that it works and to see that it really— we have with a simple, I call it, with a simple SaaS tool, you could revolutionize your marketing organization altogether. It's crazy. And this is a part of the things I'm looking at, or I looked at when I started building Allison. And that's just one part of the equation because introducing efficiency, cutting costs, increasing performance revenues, that's all nice and dandy. But when you speak to a lot of customers, that are new to AI or yet have adopted the AI tools within their company, it looks like witchcraft. It looks like witchcraft. They say, wait, should I use my studio or shouldn't I use my studio when I use Alison? Should I use my analyst or shouldn't I use my analyst? Should I fire 50% of my marketing organization because you can replace it with AI? So typically we say no. I mean, that's not what we're aiming for. We're not aiming for replacing everybody now and, you know, just having a one-person team that has a very sophisticated tool on his behalf. What I'm— what I think we're aiming for, I know we're aiming for, is you already have your teams, you already have your processes, operations, and workflows. Let me, let us at Allison show you how you can do with the same headcount, with the same people, and with the same capabilities of those people, experiences, expertise, skillset, you can do 300 times more. And I'm not joking, it's 300 times. Now, at Allison, and I won't, I won't say much about it, but at Allison, it's all about extreme efficiency internally within our company and externally with our partners, with our customers, because I believe that this efficiency is the new age, right? In 5 years' time, 10 years' time, I don't think we'll see companies with thousands of people in different departments. I really don't think so. I think it's going to change. And I think the only way it's going to change is through sophisticated intelligent tools like AI or GenAI and adoption. So I think also, you know, education and adoption is very important for us at Allison, like showing the market, showing, you know, our customers, showing everybody that it can be done. This big dream that we all have, this holy grail of online marketing is achievable and it's not just achievable in 10 years or 5 years' time, it's already achievable now and the only thing you need to do is change your mindset and just try, adopt it instead of trying new creatives on A/B tests or new metrics and methods, try a different tool. that might help you gain different mindset and different insights and understanding of how you should be running your marketing organization.

And what really jumped out to me when you describe the opportunity is when we think about startups and companies that can become massive is clearly there's a huge market and an interesting technology wave that can be ridden, but also just kind of getting back to like the really simple ideas, like how does this software save people time, money? How does it unlock new opportunities that weren't there before? And then kind of the fourth piece being like, is it delightful to use? And when I say those elements, I'm not saying 10% time-saving and 5% more, you know, 5% saved money or expanded opportunity. But, you know, really to your point, like which ones can be 10x time-saving? And I would love for you to provide maybe a specific example because I think the element that this is right now video-focused is really important and that is the future of advertising. And, you know, we already know that's not just the future, it's the current. It's the current preferred method. What about video and maybe an individual description of a way in which a company is using your technology to assess their portfolio of many marketing videos they have and how they want to change what they do in the future or learnings that they can pull from themselves or competitors? So I did just give you probably 10 questions at once.

Yeah, absolutely. So let me explain how it works and then we'll jump into the example and it'll be pretty straightforward. So nowadays, and you said it, 90% of the assets of the creatives that we see online are videos, 90%. There's focus on videos from platform side, Google, Facebook, TikTok, et cetera. They prioritize video over other types of content. Us users, we prefer to engage with video content rather than just static content. And also advertisers all across the globe, all across the spectrum, they focus on video ads and video content because you can deliver multiple messages, you can hyper-personalize the types of messages, you can test and you can try new things constantly. And these things are resonating better with your audiences. So that's one. So now when we understand that most of the creatives, most of the assets that we see online are videos, comes the big question, how do we know what's the right video content or what's the right video composition that we should be producing, right? Should be engaging with. So this is where Alison comes in. Now, the first thing that we do is we scan every single video that our customers have ever ran, ever tested, or ever launched. So it could be a few tens and it could be tens of thousands. Okay, then we use a variety, a stack of different AI models to run a process called feature extraction. We're identifying 25,000 features from every 30-second video that we scan. Then we basically correlate those features with the actual performance of the videos. So features could be anywhere from, you know, sound and voiceover and backgrounds and colors and buttons and logos and facial expressions and product attributes, etc. But once it's attributed or correlated or merged with the marketing metrics, This is where the magic really happens and you understand what components, what elements within your creatives are actually triggering your ROI or revenues rather than the ones that you might thought, or maybe you thought these are generating, but they're not really. And if you're not using AI and a high-scale, highly sophisticated AI, you won't be able to reach this level of granularity. Okay, so that's one. But then we've also understood that it's not enough just to look at our customers' creatives because none of our customers are operating in a void, right? All of us, we operate in a highly or in a competitive, highly competitive or competitive landscape, competitive markets. So we all have some sort of competitors. Now, if we only look at our own data constantly and we optimize only according to what we see internally, we're missing the market. We're missing the trends. We're missing what we don't know that we're missing. So we constantly have to run competitive intelligence, competitive analysis. Now, before Alison or without Alison, if you want to run competitive analysis on creatives, it's almost impossible to do it at scale. And a very valid use case that we see in most of our customers is that they have at least 5 different competitors with 200 videos each, or at least 200 videos. So you end up with 1,000 different videos from 5 different competitors, some on Facebook, some on TikTok, some with influencers, some with, you know, real people, some are animation. How on earth are you analyzing it without, you know, high-scale sophisticated AI tools? Again, it's impossible. So We're analyzing both creative sets, the customer set of creative and their competitors. And then we basically correlate the features with performance, as I mentioned, and we come up with insights or recommendations. Look at it as some sort of a recipe for video success. So this recipe will tell you, okay, Brooke, your next video for this product on this platform would be 15-second video with 4 scenes. That's the script for the first scene. And these are all the features and elements that it should incorporate. So that's the first part of this, the analysis, that aha moment when our customers realize what's actually moving the needle as opposed to what they thought was moving the needle. And then when we ask our customers, like, are you using those recipes, those recommendations, those insights, they said, look, this is phenomenal. We've never seen this level of data. And there's a lot of data that could be used and utilized, but it's data or it's insights. And we still have a way— a long workflow to go. You know, after we just gather the insights, we still need to come up with a marketing brief. We need to iterate on the brief and make different versions. We need to come up with a storyboard that will visualize how this new video is going to look like, and just ultimately we're actually producing the video. There's a long path, a long way to go, and a few steps to go from the insight level when you understand, you realize what you should be doing to the point that you're actually executing. This was another aha moment for us at Alison, specifically for me because I said, okay, so we've successfully created insights or came up with insights and recommendations, those recipes, it's working. Because the customers are using it and we see that it's very impactful. How do we take it to the next level? How do we augment the entire workflow and not just the analysis and insights part? So what we've done is we've built a dedicated AI model that takes raw insights and engineers a prompt. Now, this is a very sophisticated prompt. It's more like a code. It's a machine-to-machine. There's a lot of data that is being missed or not utilized during the prompting process if you just type in a prompt. So we completely changed the way we look at prompts and think of prompts and engineer prompts. But then those prompts we use to generate marketing briefs. Storyboards, and ready-made or commercially ready videos. And that's another challenge because I'm sure you and, you know, all of our listeners and viewers have heard about these GenAI models from Sora by OpenAI to Midjourney and Stable Diffusion and others that are, you know, you just type in a prompt and you get a video. But these are not commercials. These are not ads. These are just raw video, raw footage. Again, there's a long way to go from just taking a raw video, raw footage, and making it an ad. We wanted to go this mile for our customers. We're not just giving them another raw asset that they would still need to work on, they would still need to design, they would still need to use the studios and agencies that etc., we wanted to give them the holy grail of marketing, which is a self-feeding loop internally that you run campaigns, you generate insights, and then you create another video within 30 seconds that you can push back to Facebook, push back to Google, push back to TikTok, and complete the cycle within 30 seconds. And this is a ready-made commercial. And even in a way, we are bringing the power back to the marketing, to their teams. Meaning if you want to create different videos or different versions of the storyboard, you don't have— you don't want to— you don't need to be a designer or a video editor at all. You don't have to have any design experience. You just need to understand a little bit of data. That's it. Because the way we've built it is it's not really a design thing. It's more of the data that is being translated or generated or transformed into a design.

And that's what really— when we were going through the demo in that coffee shop, that jumped off the screen, where I am not a data scientist, and yet here on the screen, I could have a high-level overview and also go as granular as I wanted to, clicking into individual elements within a video. And that was really exciting from thinking back to the A/B testing and all the challenges, expensive, time-consuming. But also, you know, you picked the video B that did better than A. There's probably some elements of that other video that actually would make it do even better. And I think you've hinted at that.

Okay, so Brooke and Assaf have talked a few times about, you know, this one instance where they met for the first time at a coffee shop and Assaf opened up his laptop and showed Brooke, the actual demo of their technology and their software, right? So Allison, this is my MO. You guys have probably have a pretty good guess of what I'm about to do. On their YouTube page, Allison AI produced a really cool video with some awesome, like, animations, and it's a really cool color palette, and the sound design is great. A video of kind of like a demo of their technology, right? So you get a visual understanding of how their technology works and how it can help sort of your business and your media assets, you know, be the best and most efficient assets that you and your company and your group of people can create and put out onto the world, right? So I'm gonna play that right now because I think it's really cool. It's about a minute long, and then from there we'll hop back in. So enjoy.

Would you drive a car with only half the parts? Would you bake a pie with only half the ingredients? So why would you run a performance campaign using only 50% of the data you need for success? Imagine being able to expand your analytical horizons beyond just media metrics. Introducing Alison, a creative analysis technology that fills in the gaps and transforms every creative element into a multidimensional world of measurable data. Alison can recommend which creative component is peaking ingredient of each campaign will work best for which audience, on which platform, in which country. You can detect ad fatigue and maximize your creative's potential so you can achieve your ad goals and KPIs. Alison's unique competitive analysis solution drills down into the performance of each element of your competitors' leading creatives, giving you actionable insights for your next creative masterpiece. Fully customizable to your needs, Alison simplifies complex analysis and allows you to make decisions based on your specific data so you can produce stunning creatives that perform. Eliminate the guesswork and create successful, profitable ads by transforming your art into data. Join the creative revolution with Alison. Contact us for a full demo.

One other question that I'm thinking through is, now that you've been in market and you have some very large customers, are there things that you or your team is now like, this is, we're seeing a lot of pull from the market on this element, or has it informed your roadmap to spend more time on certain features than you might have expected? Any interesting, like, big insights from getting the product in customers' hands and seeing how they're using it?

Absolutely. So first, the advancement of generative AI has definitely catapulted and definitely sped up our video generation capabilities internally, definitely. Because now we have a sustainable, scalable, robust enough tools off the shelf that we can use and just work as a layer or multiple layers on top instead of coming up or generating, engineering the entire model by ourselves, which is time-consuming. It requires a lot of money and capital and people, et cetera. So definitely the advancement of GenAI catapulted our product and our video capabilities within our product. That's one. But 2, I'll tell you the opposite. I think there's a lot of excessive hype. And maybe I'm wrong, but this is my humble opinion. And if I'm hurting anybody, I'm sorry here. But I think there's an excessive hype around agents. AI agents. Every single company I encounter, I come by, they have this AI agent that does X or does Y or multiple AI agents that are collaborating with each other. And I think that this is a silly way of handing the technology to people. Seriously, because agents are inherently limited, right? The use of agent is saying, okay, don't do everything. No, no, no, don't do everything. Do just this. That's what I want you to do. Like a designer, right? Or like a travel agent, or like maybe a simple analyst or other copywriter, for example, right? So I think all those agents are just a means to an end. They're a step in the way. And I've heard from a lot of investors like, are you gonna incorporate agents within your AI or within your tools? Or are you thinking about using agents? And I don't think this is where we're going. I don't think this is— this is not where AI or GenAI is going. I think AI is going to more robust, comprehensive, intelligent, faster models. That can react and interact. And I think that, again, agents are not interacting the same way. So on the other side, so I think that what I'm trying to say is that the advancement and the huge usage of AI agents in a lot of different companies has done zero, done nothing to influence us or influence me of adopting the same mentality or the same usage within Alison.

Yeah, I appreciate you saying that. And another thing I've noticed and heard from both investors as well as some large companies is this idea that AI and large language models are going to make software as a service less— not important. SaaS is dead because we have AI now. And I just struggle with that one a lot because, yes, teams have data scientists and lots of engineers, but this idea, I mean, that a group as a side project, let's say, is going to build a product in a very specialized niche, and within the context of that time and that team, have the same level of insights as a full-time company that's been doing this for multiple years, that's talking to lots of other customers that wakes up thinking about this problem, goes to bed thinking about the problem and what they're going to do about it. It's just been this interesting dynamic where I think there's been this, oh, let's try to build, let's try to vertically integrate and build everything internally. And when you're talking about hard problems, there's enormous value to outsourcing to someone who thinks about this all the time, and then with your engineers and data scientists focusing on what is your competitive advantage, what is the best use of their time and energy and your company's resources. And it's just something I've been hearing in the market and I just kind of had to— I wanted to bring that up because I think it's misguided, particularly for the next several years because we are not there. And yeah, companies can kind of spin their wheels trying to build stuff internally when there's really awesome solutions that are focused.

I absolutely agree with you. But going back to the AI agents, I was just thinking of one of the examples I used to speak about when it comes to the AI agents. For example, you remember, you're almost my age, I think, more or less, without revealing our real age.

Without revealing how much less accomplished.

But you remember the Minidisc? I used to

have one. Yeah.

Me too. There was the MiniDisc, and back then we all thought it's phenomenal, right? It's gonna change the world, and this is gonna be the future. The small discs that you can, you know—

Doesn't scratch.

Doesn't scratch, right?

Doesn't skip.

Yeah, and you can rewrite, and you can do a lot of things on. And look, just took a few years and they've completely vanished, right? Completely vanished because it was never the end goal. It was never the the end product or the end technology, you know, that companies or, you know, users were looking for that could do like we're trying to do at Allison is take 100% of the workload or 100% of the workflow. And going back to one of the last questions, and it's exactly like the MiniDisc. I mean, coming up with or building internally new capabilities or having on your company trying to build everything internally, everything on your own, just a waste of time and money. Some people do succeed. You just mentioned there's vertical integration. It's not necessarily related to acquisitions and acquiring company in the same supply chain. Sometimes you try to vertically integrate internally and build it your own. Yeah. But I think that, you know, we have a specific vision, roadmap, trajectory, and it's very important to stay focused where other companies, they also have the same kind of thing. They have the same vision or their own vision and trajectory and their own path. And by using each other, you know, sometimes it's a matter of 1+ 1= 3, because if we're not— if we don't need to develop internally, this— imagine this huge GenAI model. We do develop the AI models internally, but when it comes to GenAI, we just use off the shelf. Now imagine how quick, how easy, how seamless it is to our business just to use one and the barrier to replace it, right? Change, try different tools, try different versions of vendors and work with different companies. There's no barrier. It's very slim. It might be a barrier of just a little bit of money or just a little bit of time to get to know the product and the platform and the technology, but that's basically it. So, I think there's huge benefits of outsourcing a lot or just using different companies. And I'll tell you another thing. We could have built ticketing platforms internally, just a ticketing tool to solve tech problems and communicate within different teams. But there are very good solutions for it that cost very little money. Or we could have tried to build our own CRM tool, right? But there's fantastic vendors and fantastic software out there. And I think it applies for a lot of things that we can use and we are using at Alison to stay focused, to stay mission-oriented and outsource everything we can. And I told you before, when we just started this conversation, I spoke about advisors and consultants, and it's also kind of the same. It's outsourcing your thought process in a way, or outsourcing some of the vision or some of the strategy and outsourcing acquiring knowledge. Right? I could read books on my own. I could listen to, you know, a lot of podcasts and stuff. But interacting, speaking to a person that knows you and understands you and solves some of your problems and takes some of the heavy lifting and, and help you understand in a clearer path what your true company's vision and true company's path is, I think it's priceless. Priceless. And last thing, I think a lot of companies pivot Because they've tried so hard to do everything on their own.

Now, I thought this was a great opportunity to sort of hop in and share something. So with Alison and AI, this whole, you know, conversation, I've been thinking to myself, like, what does it look like? Like, what do their AI-generated videos look like compared to what these internal companies are producing themselves? And they have some success stories, um, on their website. And the one that I'm going to show you today is from Uh, the company Laffy Taffy. So Ferrara is actually the American candy manufacturer, um, that manufactures Laffy Taffy. Um, and with this sort of new campaign, um, their challenge was aiming to increase viewership KPIs across all platforms. So their marketing team had already identified many of its top-performing creative elements throughout the years, but was looking for the right combination of tags to complement each other and engage audiences on multiple networks. So with this, what I thought was really cool is they, um, on their website brought up like creative recommendations that, you know, uh, Ferrara had at first for Laffy Taffy, but then compared to Alison's recommendations. So for example, brand tenets, um, their original version was family connection, and then Alison's recommendations were dad joke energy. Which I think is great. Um, and then for sound, the original recommendation was music and sound effects, but Allison's recommendation was music, sound effects, and voiceover. And then as far as font type, which I thought was really interesting, um, I'm a bit of a font knob, font snob myself, so I understand this one. The original version was like not included on what type of font they should use, but Allison recommended using some sort of stylized font. So on this, you know, article that they have on their website, they also provided video examples of the original asset and of the new asset after using Alison AI. So I'm going to play those right now on the, um, the visual you could see on, you know, this video version of this podcast on YouTube, LinkedIn, wherever you can find it. Um, so here is the original video asset from Ferrara. And then here is the new asset after Ferreira used Alison AI. Okay, we're going to finish off and take care of our last ad here, and then we're going to hop right back into the rest of the show. So hang tight.

Do you have a venture capital portfolio of cutting-edge startups? Without one, you could be missing out on enormous value creation and a more diversified personal portfolio. Alumni Ventures, ranked a top 20 VC firm by CB Insights, is the leading VC firm for individual investors. Believe in investing in innovation? Visit av.vc/foundation to get started.

I really think that it's one of, it's one of the main reasons. It's rarely the case that the market is failing or the market is locked, the market is too crowded. I don't really think so. I think the main problem is focus, and focus because you try to solve all of your problems internally instead of using smart people that have solved the same problems or almost the same problems for a lot of different companies. And for you, it's a simple SaaS login and password or a simple integration, and that's it. And once we discovered it, once we changed our mindset that we don't need to build everything internally, we only need to build Alison as the core capabilities, and that's it. And that's also the reason why we now have 15 different proprietary AI models internally, because we spent our time on what's important. It's the analysis, it's the insights, it's the feature extraction, it's the prompt engineering. These are the things that matter the most for our business, and we spend No, the vast majority of our resources exactly there.

I think that's a deep insight on how both founders and large companies can think about build versus buy approach and really maximizing their competitive advantage and their team and their time and what they're good at. And the other thing I was thinking about with the MiniDisc example, which I loved, is I guess If I'm a large company, we've always done A/B testing, that's how we do things here. What would be the MiniDisc version to that solution? And it's, we'll save you time and money, we'll do A/B testing faster. And if I'm hearing you correctly, it sounds like that's been kind of the state of play maybe for some adjacent competitors is that we'll make your A/B testing better, we'll make it faster. Which is like, we'll make the CD smaller. And we'll make it skip less.

Exactly. It's going to be smaller. It's going to skip less, but it's going to be the same minidisc that you know. It's the same process. You put it in the machine and that's it. And yeah, but I think, again, we're focusing on hype and we're focusing on how can I do more of today. Rather than looking at the future and trying to understand what's going to be there next year. The reason I'm saying it, yes, there are some companies in our space, in our market that are saying, okay guys, you're running heavy A/B test operations and we all know that it's heavy. You need to create 30, 40 different versions of videos Let us do it for you, but not with 40. We'll do it with 4,000. It sounds super exciting. It sounds, wow, it's again witchcraft. I'll run 4,000 versions in my A/B test. So I gotta find the right formula. I gotta find the right insight. Not really. And the trick here is understanding how the backbone, and the infrastructure really works. So even if you take the entire production workload off your shoulders and somebody, a company, a SaaS tool hands out 4,000 videos to you, you still need to run an A/B test with 4,000 versions live. Let's say on Facebook, on Google, on TikTok, etc. That means That each one of those 4,000 videos should have a testing budget. And that means that collectively those 4,000 videos should accumulate enough impressions or enough clicks or enough actions to come up with a meaningful decision or a meaningful winner for this A/B test. Let me ask you this. Just Again, roughly, but how much, how long do you think that the biggest platform on earth, Google, Facebook, TikTok, how long will it take them to run a 4,000 version A/B test and come up with a meaningful decision or meaningful output? That's one. 2, how much of a testing budget or how big are the testing budgets that do you think you should be dedicated to this process with 4,000 video versions?

Yeah, the quick answer is I don't know, but the reality is needing it to be statistically relevant. So like, to your point, there's enough impressions with each one of those 4,000 that you can say, hey, this one's working better than that one. But then on top of that, you run all 4,000. Now you're okay, You could stack rank on how they did if there was truly a lot of impressions. But again, we're not getting into the variables of what about each individual video. What is the statistical significance across all these videos? That's what's working, what's not working. And I think that is the hard problem that you have solved. And the point about 4,000, I mean, I can't even imagine how many— Data scientists and how much time and energy and money would be needed to do that.

Exactly. And, and, and by the way, uh, we used to have customers that before they've tried Alison, they used a lot of A/B tests with hundreds of versions, maybe not thousands, but hundreds of different versions in every single test. And it took them too long to come up with a, a, a, an insight, sometimes months. Now imagine you run a test for months, right? The market already changed. By the time you have the insight, you should run a different test because trends and seasonality and market trends are completely different. So that's one, and you're missing the whole point of testing. 2 is because you have to dedicate budgets to each version, you'll need At least $50,000 to $500,000 just for this test, and you run the $500,000 test knowing that $400,000 of it is going to go to waste. Now, I can't think of a single company, even a single CFO or a CMO that would hear those numbers and say, okay, yeah, let's do it. Let's invest more in A/B tests. I mean, rather than, and as you said, understanding billions of variables at the same time because you've already tried all those versions, even if not in a single ad or in a single video, you've tried those collectively. And even if you haven't tried some of those, your competitors or direct competitors have tried. So by looking at every single creative you ever ran and looking at your competitors, we basically cover 95% of the variables that exist. And if you cover all of those and you analyze it using AI and it takes a second, then you have more power and a more powerful tool at your fingertips with Alison than running 4,000 videos in your A/B test.

Yeah, it's— having gotten to see firsthand, we're talking about it, but having seen it too, it is really, really impressive. And I can't wait for what the product, you know, the maturation of this product over the next few years ahead and seeing your company continue to win and grow its customer base. I think this is a good point to wrap, but before we do, I just wanted to ask you if there's Um, and ask to our audience, to our community of listeners, uh, our network, anything that they could do and that they could reach out to Alumni Ventures, uh, which would be helpful for you?

Um, so I think first it's just a mere, you know, a general statement, which is don't be AI shy. And don't— the same as you're spending thousands of dollars on your A/B test, try to spend some time, money, efforts, and resources on using it, adopting it, and seeing the true value and seeing how it could potentially transform your own business. That's one. But 2, I think what really excites me is I actually speak to fellow either fellow founders or marketers who are actually hands-on experiencing those challenges that we just talked about and hear their perspective, hear how they're overcoming those challenges, how they're solving their own problems when it comes to lack of visibility, lack of analysis, as we said. I would love just to hear back how a lot of different people, a lot of different companies, sectors, environments, markets, how they look at their problems and what their tricks, tips, or just tools that they use to solve those problems and overcome those challenges. That's interesting to me because this is how I got into Alison, questioning a lot of different marketers and seeing firsthand their experiencing their challenges and pain points. But also I think as a general understanding, as I said, as experienced as I might be, there's a lot of things I still don't know, a lot of things. And just generally as a person, I think that knowledge makes you stronger and also knowledge makes you a stronger founder and CEO. The more you know, the more you can react and adopt, et cetera. So I would love to hear back from, again, founders or marketers about their specific marketing challenges and how they're overcoming them today and how they think or expect that they would be overcoming them in the next few years.

Well, on that note, Assaf, thank you so much for joining The Tech Optimist today. It's been an absolute pleasure having you on the show and looking forward to doing this again sometime in the future.

Likewise, Brooke. Thanks for having me. It's been a pleasure and looking forward to the next one.

Thanks again for tuning into The Tech Optimist. If you enjoyed this episode, we'd really appreciate it if you'd give us a rating on whichever podcast app you're using, and remember to subscribe to keep up with each episode. The Tech Optimist will Tech Optimist welcomes any questions, comments, or segment suggestions, so please email us at info@techoptimist.vc with any of those, and be sure to visit our website at av.vc. As always, keep building. Thank you.