Introduction
Ryan: Welcome everyone to the AiFounders podcast show. Our podcast is dedicated to celebrating the remarkable accomplishments of AI innovators, entrepreneurs, and visionary founders and the captivating stories behind the movements that they’ve built. I’m your host, Ryan Davies. I have the honor of hosting today’s episode from Algorithms to Checkouts, unveiling the AI Magic in advertising and marketing with our special guest, Depesh Mandalia. Thanks for being here.
Depesh: Yeah, pleasure.
Ryan: So, a little bit about Depesh. He’s an AI entrepreneur and CEO of SM Commerce, a Facebook ad agency with a remarkable track record of helping clients generate $100 million in online revenue. He’s also the CEO of ZASR Digital, where he shares his expertise in marketing through a proprietary framework. As an international speaker and Facebook insider, he leads his own Facebook ads community and events. He’s on a mission to empower entrepreneurs to build highly profitable businesses. He is an expert in this area, talking about AI and being able to really integrate it and revolutionize the way we do things. Marketing and advertising are some of the key ways that we’re seeing it. So, how has AI become an integral part of SM Commerce marketing strategies and success?
AI in Business Growth
Depesh: I’m going to dive straight into the heart of the matter. I’ll share that I’ve transitioned from having a team of media buyers running ads for me, costing anywhere between 3-6K a month, to a setup where it’s just my marketing assistant and me handling the ads. The power lies in the fact that my marketing assistant isn’t a Facebook advertising expert. However, she uses my AI tools, where we’ve uploaded years of knowledge on running ads, frameworks, copywriting, and more. She asks the right questions, and I guide her through the process.
Traditionally, bringing in a media buyer would mean spending 2 to 4 weeks on training. Now, I can give a marketing assistant access to my AI tool and say, ‘Go and ask these questions.’ Quickly, she and others (as I now allow others to use the tool) can become experts. They rapidly launch campaigns, set up structures correctly, understand targeting, and grasp concepts like Eugene Schwartz’s five stages of awareness. They can handle things like setting up targeting, following advice on campaigns and pixels, and swiftly reaching a proficient level of running ads.
Consider how many people worldwide, especially small business owners or those running digital products, struggle with Facebook’s complexity. It’s like Facebook gives you a Ferrari and says, ‘Here’s a cool, fast car. Enjoy!’ But once inside, you’re overwhelmed by all the controls, unsure of how to start it. This is the common experience many have with Facebook advertising.
AI enables a natural conversation akin to chatting with GPT. It serves as a common interface, guiding users through questions on where to get started 24/7. The power is in enabling millions worldwide to gain expertise and access they may have needed help to reach. This is particularly impactful in countries like Brazil, with a growing digital economy, or in India, the Philippines, and other places where a $2000 course might be unaffordable. Here, they can afford $20 a month on AI to start learning and applying these concepts much faster.
Ryan: That’s exactly what AI is designed for, and you’ve taken it to the next level. Is that accessibility, like you said, right? It’s bringing. It’s leveling the playing field a bit, getting people to be able to get through this and understand things that may have been out of reach for them. And to that effect, you talked briefly about the targeting aspect that takes place. And how AI can help you not only with the creation and the walking through the campaign but also with AI-driven targeting. How is AI able to refine and improve those targeting strategies and the role it plays in making sure we can reach the right customer with the relevant content and offers that are out there?
AI-Enhanced Marketing Strategies
Depesh: One crucial aspect I emphasize when explaining AI to people is the existence of two ‘brains’ or databases that AI operates with. Whether it’s OpenAI, Bard, Grok, or others, they all rely on their knowledge banks or databases. The raw format of AI is inherently intelligent, but its true potential shines when you provide it with a contextual brain.
In my case, this contextual brain comes from a training methodology I’ve developed, specifically in the realm of Facebook ads, known as the BPM method (Brand-driven Performance Marketing). This methodology encompasses my approach to advertising, focusing on brand play, understanding avatars, customer research, and employing concepts like Eugene Schwartz’s five stages of awareness and Maslow’s Hierarchy of Needs for targeting.
The next layer involves performance marketing—constructing ads within my ad account, targeting audiences at the top, middle, and bottom of the funnel, and integrating all the pieces seamlessly. I’ve been refining this framework for the past ten years, utilizing it in our agency and training others on it. Recently, I’ve taken this entire context and uploaded it into my local AI, including case studies, client experiences, frameworks, and more.
When I interact with the AI tool, I not only leverage my framework and knowledge but also uncover forgotten insights. For instance, it might recall a strategy I implemented six years ago that I had completely forgotten. The AI’s ability to remember past experiences is fascinating. It provides new angles and opportunities, constructing a view of the campaign and suggesting budget allocation and targeting strategies.
The AI, based on extensive data, offers insights on broad targeting for Facebook ads. It guides users through setting up targeting and advocating for a few look-alikes and interests. It recognizes the importance of high-targeted messaging and strong creativity for effective campaigns. By analyzing chat logs, I’ve observed users asking common questions about targeting, and the AI provides instant learning, condensing what might have been covered in multiple training videos into a few conversations.
Addressing a common concern about AI doing all the work and potentially making users less knowledgeable, my approach is different. I’ve designed the AI not just to answer but also to explain its reasoning, mimicking the logic teachers seek in schoolwork. This way, users are educated while receiving answers, promoting a deeper understanding of the process.
Ryan: That’s just exact. I couldn’t, couldn’t have it any better. I absolutely love that comment. We hear that all the time. Well, AI is going to make people obsolete because they’re not going to be able to, they won’t have a purpose, or they won’t know what’s going on. They’re just going to hand off the work, being able to get that feedback. An understanding is incredibly important when we’re talking about feedback in this industry. ROI is the most critical metric in marketing from the outside. How are they able to measure the return on investment specifically attributed to AI Power marketing initiatives? And what edge does AI give them to give them that stronger ROI?
Depesh: I’ll give you an example. So, I work with a lot of freelancers and agencies. Based on our agency’s success, we train others as well. And, in the next 12 months, we’re going to see a rise in AI ads agencies. So I had a conversation with someone this morning, and she said she’s got ten clients. She’s freelancing, but she’s struggling with time right now. And I said, well, here, here’s how you can use AI. So, first of all, it can be your media buying assistant. That’s fairly simple. There are complexities behind it, but you can have an AI that can be there to help you. But one thing we’re working on is you can also train AI to be an expert at the client level. So imagine you’ve got ten clients, and you’ve got 10 AI agents that you’re feeding in all this relative and related knowledge based on that client.
Let’s say client one you have customer research, you have 12 months of data, you’ve got all the reports you’ve sent them. You’ve got all the kinds of websites and nuances that are client-specific. Now, let’s imagine, next week, that client comes to you and says, hey, we want to launch a campaign similar to what we did last year for Black Friday, but we want to do something slightly different. You can go straight to the AI and say, hey, what did we run? Last Black Friday? What worked? What did and what can we try this year? And that’s how we’re using it right now. So now it becomes not just a knowledge engine, but it’s an intelligence engine. It’s giving you additional information that would have probably taken you 30 minutes to open up Ads Manager, find your reports from last November, and analyze what worked and what didn’t. You’re fully capable of doing that as a human, but AI is making it instantaneous. So, within 30 seconds, you’ll know what you ran last November. What worked, what didn’t, and ideas to test now. And that’s the power of what we’ve got. So, a freelancer in that stage could go to 20 clients and go to 30 clients, but AI is doing a lot of the heavy lifting for them. The next level for that is not just taking the knowledge and saying here’s what you should do, but then plugging it into ads manager and doing it for you as well, which is going to be pretty close.
Challenges and Opportunities in AI
Ryan: Yeah, again, you got hyper-personalization. You’ve got that ability to capture all that information so you can take advantage of, like you said, what worked? What didn’t work was revolutionized pricing and promotions by dynamically adjusting pricing, pricing, and promotion to maximize results. Do you have any of those notable instances where that kind of AI-driven pricing strategy has led to some significant outcomes or abilities, places where you’ve seen it in action very successfully deployed?
Depesh: I am still waiting to see it. I haven’t used it yet, and part of the fear is how much control you give away. With AI, my concern is always, will it run away from me? And do I know what’s going on? We are definitely using it to analyze and come up with ideas. I just had a conversation this morning about where I’m launching a new book next week, and when you try to price something, pricing things is so hard because it’s so relative, right? It’s the price you should be paying versus the price you think you should be paying. So there’s a price elasticity there, and we want to find the level where I can get as much of that revenue without impacting the conversion rate. Now, I’ve done multiple book launches, and I say, hey, I’m looking to launch this book next week. I’m having a conversation with the AI, and AI got all the data and all the campaigns inside. Then we can have a discussion and start running those hypotheses and say, right, let’s launch at $9, and then we’re going to quickly launch another one at $19 and $29 and test. So, in the past, I’ve done this manually. So, I have an E-commerce store.
We sell personalized jewelry, and the recommendation from our supplier was most people are selling this for $49 for 49. 39, and 29 $39 had the highest conversion rate and the best average order value. So we stuck with that. But that took time. Now, that goes into the knowledge bank. In the future, I can have a conversation with AI and say, do you remember that e-commerce test we ran? Please tell me what the results were, and it will come back and say, well, you ran these tests. This is what you’ve got. This is what you should consider for your next launch. So that’s as comfortable as I’m using it right now. In the future, yeah, absolutely. You could plug in and do dynamic pricing. Why not? Suppose you look at connecting your CRM (customer relationship management) data to AI. In that case, that’s scary and powerful at the same time because now, for example, I know in my CRM there are certain signals based on people being on my list, a certain amount of time clicking on certain things, buying certain things. And I can have a human go in and interpret that, and I can work with an active campaign, which is the tool we use. But now imagine plugging all of that into AI and it coming forward and pushing stuff like pop-ups on the website or emails like you could send automated emails based on stuff that’s happening in the background. It’s exciting, but it’s also quite scary to imagine what it could do as well.
Ryan: The myths are running wild already on that. I remember reading the story about Uber categorically denying the fact that they measure your battery life because when you’re really low, they’ll charge you more for it. After all, you’re more desperate and things like that. Right. So there’s all sorts of craziness out there that, yeah, myth, they’ve denied it, but you could see where this is going. I want to drill more into SM commerce, and just with any business in general, there’s a difference. We have listeners that are all at different stages of the business cycle. Some people are trying to break into a region. Some people have hit their plateau and their flat line now and don’t know how to hit that next level. Turn a million to 10 million or whatever, or people who have hit that now want to experience more growth. Can you tell me a little bit more about the unique challenges and how there’s a common formula between them, a different formula that you look at, and how you can help each be successful along the road?
Depesh: So if we take the AI lens as an example, one of the things we’re looking at right now is across the agency. What are the SOP-driven tasks that AI could do? One of them, for example, is when a client joins the agency, there’s an onboarding process. So a client account manager will communicate to the client and say, hey, we need a list of these things. We’ll set up folders or sign a project list and things like that with AI and with Zapier, for example, you can do all of that. So, for example, this is our process: when a client pays, Zapier can automatically send out the contract when the contract is signed; that’s the trigger for AI to do the rest of the stuff. So, Zapier will allow us to connect to as many applications as we want. But AI could do some personalization and customization, which the customer, the client account manager, would do. Now, for example, the fact that AI can now browse the internet and it can go and look at websites once that client has joined, we’ve already got their website and their names. It could find contextual information from Bing or Google, whichever AI you’re using, and it could add some personalization. Hey, Ryan, I’m really excited to get you on board at SM Commerce. I noticed that XYZ and whatever you can start doing that profile data for you and then start to build that relationship. And one of the things we’re looking at is, for example, what’s the role of a client success manager? They are there to check in on clients, make sure everything’s okay, and things like that. I saw a video of a sales call where an AI was speaking to a human and closed that human without the human knowing that he was actually speaking to an AI voice in the background. Now, knowing that you listen back and say, I heard the delays, and I get it, but that guy basically bought something just based on AI conversation. So you could imagine. Now, the AI bot, let’s say you join the agency, and we have your cell number. Some applications allow you to drop a voicemail where it skips your kind of ringing. So, for example, you can have an AI-driven voicemail saying, hey, Ryan, really excited to have you at the agency and things like that. It could easily mimic my voice.
I have a paid membership for media buyers. We actually ran a test last year where I sent 100 personalized 62nd welcome videos. So that’s me just recording on the mobile.That obviously is time-consuming, and I sent 100 of those personalized messages versus another 100 not getting personalized messages. We had about a 30 to 40% uplift on lifetime value, where I gave them a personal in. Welcome to the business. Now, imagine there’s a tool called Maverick.ai. It could do the same thing as scale. All you do is record one video once, and it can change your mouth movement and also add the name of the person. So I would have one recording that says, Hi David, welcome to the membership… It would change it in real-time to Ryan or Jane or to John and things like that. So now you feel like I just recorded you a personalized video, and now you feel more connected instantly, more connected to the brand and the business, and AI is enabling things like that. So, it’s about more than where we can replace people and things like that. It’s also how we can improve retention, lifetime value, scalability, and things like that and give people a more personal experience as well.
Short-term and Long-term Trends in AI Integration
Ryan: Depesh, I’m talking to you right now. It’s scary where it’s going, though. That’s exactly it. I mean, I’d love to drill down a little bit more into what short-term and long-term trends you see in terms of the integration of AI with marketing and with advertising, marketing strategies, and how it’s going to shape the future of this space.
Depesh: It’s going to scare a lot of people. I even saw a message on Facebook today saying someone’s created a Facebook bot, but when this person asked it to fill out some headlines and copy and things like that, it said, well, I’m not going to do it for you, but here are links to go and learn it for yourself. So, someone’s changed the way that AI works because they don’t want to give people the answers. My problem with that is at no point. Did anyone in their math class go and hit that square root button on their calculator and say, oh, I wish it told me how it worked out? Like no one cares. Go and find the square root of 54, type 54, hit the square root button, and get on with your life. And that’s where AI is going. Like, at some point, no one will care about how you figured it out. Just give me the answer. I’ve got better things to do. And this is the thing: there are going to be naysayers who want to hold AI back, and it’s taking away our jobs and our critical thinking and stuff. But what jobs did computers take away? What jobs did calculators take away? They enabled new opportunities. That’s the thing I’m looking at with AI, and I’m fortunate I fell into AI when Chat GPT first did its public launch in 2023 or whenever it was late 23. And I just went down rabbit holes, and honestly, every week, and you all know this as well, there’s a new idea, a new concept, a new way of doing something. And as much as I’m at the forefront of AI, I also know I’m so far behind because whatever publicly gets released, they’ve been working on it for 12 months. So there’s always something new to learn, and you get better at staying in your lane. So, my lane is advertising and business growth. So, I’m looking at AI in that respect. Other opportunities pop up. So now you can create your GPTs on open AI, which are your apps. Now, they’re launching at the app store. Now, you can make money by selling these kinds of AI bots. It’s a distraction. It’s a fun distraction, and someone out there is going to make a lot of money through it as well.
Revolutionizing Business with AI
Ryan: Absolutely, that’s exactly it. Right. I always hear about them. I was talking about this, or I looked at this, and I’m getting all these personalized ads about it. I look at them and go, would you prefer ads about nothing that you care about, or would you prefer something that you might actually be interested in? Like, there are scary elements to things, and that was years ago. Now, it’s going to be hyper-personalized in terms of what we can do and all of that. I’m really excited to see where this goes. I think the future is going to be wild, that it’s a little bit of a wild west of what’s happening, and it’s going to be interesting to navigate these waters, to close off here though. Depesh, tell us a little bit about sm commerce where people can get in contact with you. You dropped the breadcrumb that you got a new book coming out here. Let our listeners know about all that stuff so that we can learn more about you and hear your thoughts about more information here.
Depesh: Yeah, for sure. SM commerce.com is where we’re residing. And where I’m taking my business right now is we’ve gone from a 25-person Facebook ads agency down to seven people, and we’re leveraging more technology and AI. And so not only are we doing that in our business? Because I believe that’s the future. We’re training other business owners to do the same and think the same as well. There shouldn’t be a case where a single-person business owner is going to survive without any other humans in their business. I’ll give you an example. So our customer service, for example, one thing we’re launching next year is when you contact it, the first point of contact will be AI, and it will try to deal with as much as it can. And if it can’t, it will go to a human. So that’s what we’re working on. So, I’m training media buyers on our ad signals membership. So that has our AI tools, training videos, Q and A support, and things like that on the agency side. We’re doing more consulting and coaching to help business owners understand how to leverage advertising and AI in particular, and on the other side, I’m geeking out on AI. I can’t get enough of it, and I do get too distracted at times by it, but I think it’s just so exciting. To me, it feels like when the internet was coming to fruition 20 or something years ago, where people were doing things like, hey, I can do wild HTML stuff on websites, and how cool is this and Front Page and Dreamweaver? I don’t know if the people listening are old enough to remember that stuff, but that’s where it feels like we are right now.
Ryan: Yeah, it’s incredible. And then the new book comes out.
Depesh: So, the new book is designed to get people started with Facebook advertising. So, I ran a five-day challenge, and this is a cool thing. I ran a five-day challenge showing people how to get started on Facebook. Took the transcriptions and got AI to publish those as chapters for a book. And then I’ve gone in and edited it and proved it. And now we’re going to launch that as a product as well. So, a lot of the opportunity with AI is repurposing, and we’re using it. So that book will be landing on Black Friday with a pay-what-you-want offer as well, which I think will be a cool little test.
Ryan: So cool. Absolutely, love that. So, Depesh, thank you so much for joining us here today. Really appreciate another enlightening journey through AI innovation. And we hope that you have been inspired by the story shared today. Remember, the future is driven by pioneers like our guest, Depesh Mandalia, and the limitless possibilities of AI. So, for our listeners, stay curious and innovative and keep exploring the boundless horizons of technology. Before we sign off, a small request for our dedicated listeners: as always, if you’ve enjoyed our podcast, please take a moment to leave a review and subscribe to the show on your favorite platform. It’s your feedback and support that helps us bring you more amazing content and amazing guests like Depesh; until next time, everybody. Thank you so much. This is Ryan Davies signing off. Take care.