Dušan Jovanović: Hello. And please let us welcome Sani, the host of No Hacks podcast and an AI adoption specialist. He's a very dear friend with more than 15 years in the industry. And thank you so much for sharing the advice that you have for us today.

Slobodan Manić: My pleasure, Dušan. Thank you for having me on.

Dušan Jovanović: All right, so let's kick it off with a few questions that we have for you. So the first one I was wondering about is like we have most of the DTC founders that say that they're just too busy to experiment with AI. So what will be the first workflow that you would automate in a 7 to 8 figure ecommerce brand? And why would you choose that flow?

Slobodan Manić: Great, great question. And if you're a DTC founder and struggling with finding time, you're just like everybody else. So it's not unusual for people who are way too busy to find time for yet another thing they have to learn. The first flow that I would automate, and it may not be the highest influential flow long term, but it's a good first flow for people to start with, is automating your reviews and analysis of your reviews. So every month, for example, depending on how many reviews your best selling product gets, you can look at all the reviews that you got in the last month and then you can find, you can have AI and LLM analyze the themes in those reviews. Maybe they're praising a certain feature of the product or they're negative about a certain feature of the product and then you can decide if you want to address those in a product page. So this is the first one that I would do because every DTC person knows the importance of reviews, the value of reviews, most of them do not have time to look at the reviews every month. And if you automate that and you get just the summaries and just the knowledge from those reviews and you don't have to read every single one and react to every single one. I think this is a clear, clear thing to automate and it's an absolute no brainer for most brands.

Dušan Jovanović: Okay, that's very useful, thank you. And since we do have so many things that people are talking about right now, it's very hard to distinguish between the hype and the reality. So how does a smaller ecommerce team decide which AI tools are actually worth adopting and which are just like shiny distractions?

Slobodan Manić: The longer advice would be more time consuming. Advice actually would be to learn about the basics and the building blocks of AI technology, what agents are, how to prompt, what they can, what the tools can do. That way you can tell hype from actual useful features. The more practical advice is just pick one. If there's a hundred shiny tools, just pick one that seems popular enough and good enough. Because people will tell you Claude is better than ChatGPT, Gemini is better than. They're all within 5% of each other in every single way. So it doesn't matter if you're just getting started. Pick one. It could be a base model like those that I mentioned. It could be an agent, it could be any kind of implementation, a wrapper based tool, whatever. Pick one. I would say pick a base model first and get familiar with how to prompt it, how to implement it into your workflows. They all have APIs. You can use a tool like n8n or Make to build those no code workflows. Pick one, get good at it and you will not even care about the other ones.

Dušan Jovanović: That's pretty much the essence of it. Thank you. I believe you have something you could share with us which will be very useful for the audience and it's coming quite soon. Would you share what it is?

Slobodan Manić: Yes. So this is first time ever I'm doing a webinar. I did one in 2015 or 14, so a long time ago. I'm doing a webinar with a friend of mine, Iqbal Ali. He's an experimentation coach and the webinar is AI Beyond the Chatbot. It's aimed at people who are stuck in their ChatGPT and Claude chats and see that as the AI. And we aim to teach you how to build an agent, how to build a workflow, how to implement it into whatever you're doing. So teaching you the basic building blocks of using AI in your day to day in your work life.

Dušan Jovanović: August 19th. Okay, so we'll share the link with the audience and highly advise for everybody who's interested in the topic to come and ask you more questions and see how you actually do it.

Slobodan Manić: In reality, it's free to attend, so you just need to register.

Dušan Jovanović: Amazing. So we can stick around for like an additional question that I have. What I was wondering is, do you have any success stories, real life success stories that you might want to share with us?

Slobodan Manić: You'll see all those workflows all the time, workflows on LinkedIn and you know, this is what you can. Yeah, show me the end result. Like I want to know what it does and how well it does it. So success story in terms of AI adoption is something that's almost invisible, that you don't even see or interact with. It just connects with whatever you're already doing and expands on that and makes that better. So I did a project recently, it's in a continuous education organization from Norway. They're their biggest of the kind and they had hours and hours of material from the classroom. So lessons that are recorded, teacher presenting something, one hour, two hour, three hour long lessons. In total they had 60 something hours. And they wanted to create e-learning units and quizzes and whatnot from that video material. So the workflow I built is they just drag a file, a video file into a Google Drive folder. There's a workflow that monitors the changes in that folder. Whenever a new file is added, it transcribes it, it talks to an LLM, breaks it down into a manuscript for the lesson. For the e-learning component, I think it's five quiz questions and a bunch of other things that it creates and puts it back into that Google Drive folder. And what used to take them, I want to say five hours per lesson per unit basically is now done in a few minutes, let's say 10 minutes because it still needs to be reviewed manually by the person who's running the program. But still that's a lot of time saved. So that's the biggest value, is not in giving you a possibility to do something that you never even thought of doing before. I think that's hype. If someone tells you you can have 100 LinkedIn posts in two minutes, I never wanted to have 100 LinkedIn posts in two minutes. I never thought about this. But a practical thing where you know, we have all these lessons, we don't have the people or the money to put that into the platform. Let's find a way to use technology to do that. So same principle applies to ecommerce, obviously.

Dušan Jovanović: Yeah, perfect. Like to a point it resembles what you said in actually monitoring large amounts of data, as those could be reviews, and then extracting some value in a very easy, easy way and pretty much automated. Right. It comes to you. You don't even have to think about going somewhere.

Slobodan Manić: Right.

Dušan Jovanović: It just comes to you. The data comes to you.

Slobodan Manić: With ecommerce, with any business, think about what you have that others don't. And that's your reviews, that's your customer data. First party, zero party data. And try to build on top of that. If someone tells you I'll create ads for you, like a thousand ads for TikTok in two minutes using Google Veo or whatever, just generate some video slop. If someone tells you that and you can do that, anyone can do that. If you can create a thousand ads in a minute, anyone could create a thousand ads in a minute. No one has access to your customer data but you. So work with whatever the moat is, whatever your advantage is, and try to build on top of that and not on top of something that literally anyone can do.

Dušan Jovanović: So this reminds me of one thing people on Amazon like to do. Sellers love to look into the products of their competitors and see what they could improve. Right. So I'm just thinking how hard would it be to actually use a similar tool to basically go through all the new comments or new data you can get about competitors and feed that information into your product so you know what you can do differently or what the downsides of their products are.

Slobodan Manić: I would say fairly easy because you can build an agent or just a simple workflow that scrapes comments and puts them in a spreadsheet, in a database, wherever you want to store them. So I would say extremely simple. And you said, how complicated would it be to build a tool? That's the whole thing. You're building a tool that does the thing you needed to do. That's why I love AI adoption and building with AI specifically, it's not a one size fits all. Everyone should have their custom solution. So if you're hiring for a certain role in your company, it's going to be a person that your company specifically needs. Same thing with AI implementation. It's going to address a need that your company has or you have. It's not going to be some platform that sells you AI agents that anybody can do the same thing with or generating videos or whatever. It has to do exactly what you want to do. So start with the problem, work on the solution. It's always that way. If you start with a solution and then find a way to come up with a problem to use the solution on, you're doing it wrong.

Dušan Jovanović: Yeah, seems like in product development these products never succeed. Sani, thank you so much. We'll share the link of the upcoming webinar and it was a pleasure hosting you and thank you for sharing all the information.

Slobodan Manić: My pleasure. Thank you.

Dušan Jovanović: Bye.