In Episode 83 of the Digital Velocity Podcast, Erik Martinez is joined by Ken Burke, founder and CEO of Microcasting, to explore the next frontier in direct-to-consumer ecommerce innovation: AI shopping agents. As AI rapidly evolves beyond chatbots, Ken shares a vision for an “agentic” future—where AI doesn’t just assist but actively guides customers through highly personalized and conversational buying experiences.
Ken unpacks how Microcasting’s crawl-walk-run strategy empowers brands to experiment with AI shopping agents, replace outdated customer service tools, and ultimately integrate consumer product discovery, checkout, and post-purchase interactions into a seamless flow. Drawing on early results and real-world implementations, he reveals how AI is reshaping engagement and conversion rates through smarter, data-driven insights gather from live shopping journeys.
Whether you're a brand marketing executive, digital marketer, or tech leader, this episode delivers practical and forward-looking insights on AI shopping integration that can enhance customer experience and grow revenue.
****** Get a Free 30-Day Trial of Our AI Shopping Assistant ******
Listeners of the Digital Velocity podcast can try the AI-powered eCommerce Shopping Assistant free for 30 days—boost conversions, reduce support load, and guide customers to the right products faster. Just email ken@microcommerce.ai and mention “Podcast Friends” to get started.
Contact Ken at:
- Website : MicroCommerce.ai
- email: ken@microcommerce.ai
- (personal) Ken Burke | LinkedIn
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:00):
Welcome to the Digital Velocity Podcast, a podcast covering the intersection between strategy, digital marketing, and emerging trends impacting each of us. In each episode, we interview industry veterans to dive into the best hard hitting analysis of industry news and critical topics facing brand executives.
(01:00):
Very different. Yeah. Just in the year where AI has come and it's just amazing. You know, that even just in the last month, it's like, oh, we can now do this, or this. Even chat GPT has gotten so intelligent about what it's done, and literally it's been within the last two or three weeks that it's gotten this intelligent.
(02:00):
Yeah Microcasting's been around for about three, three and a half years. And as with any good company, we were a kind of a customer success platform initially and had some really good success in that. But about a year and a half ago, we started integrating AI into the customer success area and we're like - wait a minute. This is direct application to what we know and my entire team and myself have been in e-commerce for over 25 years. I'm dating myself, but I started my first company in e-commerce in 1995, which was the beginning of e-commerce version one.
(03:00):
If you think about the interactions with Alexa we're first generation and now can we apply this to the front end e-commerce experience and really take what has been going on for the last 20 or 25 years, literally what you and I created years ago together. L-shaped navigation, facet navigation, search on the top, products in the middle. It's still the exact same interface. It's boring.
(04:00):
I was sitting down with a friend of a few months ago, and we were talking about AI and where it's headed and I'm like, I have this question. "What happens to the website when you have AI agents working together?" So basically the concept of an AI agent is an application that's going out and doing something on your behalf.
(05:00):
Yep. I think you just define kind of the next generation of e-commerce and how people will interact with it. Whether it be a device like Alexa or whether it be a phone or a desktop, right? Can it all come together in that ease about, I know what you want and I can go and kind of shop for you.
(06:00):
And you know, we have a lot of data that says that people hate retail chatbots.
(07:00):
First thing that you're gonna do is obviously you're gonna go in and you're gonna have an experience. And first question that a lot of people ask is, " Do I replace my chat bot area?" You know - the floating footer. "Do I replace that with you guys?"
(08:00):
Then I say the run strategy, which people are just beginning with, right? This is very early, is doing augmented search. We're not replacing their search engine, they type in their search. And what we do is what Google does. We come up and we give you our AI version because they may have asked a question, they may have typed in keywords that didn't make sense to the search engine, but will make sense to us. So we're giving them our results. And then the results for the normal search will appear below that.
(09:00):
But it's fascinating that people are starting with their chatbots and it may be just a mix of the companies that you're working with versus some of the companies I'm working with.
(10:00):
So you guys have integrated this now at least half a dozen times at last count. It might be more now since the last time I talked to Christie.
(11:00):
But what we do want you guys to understand is like there's some really fundamental shifts happening in the way this type of technology is getting deployed and how the buyer, your buyer, interacts with it. So can you share some of the data that you have gleaned so far from the implementations you have about buyer behavior, what they find interesting, improvements in conversion rate or engagement rate?
(12:00):
So one of the reactions we've gotten is, wow, that was an eight interaction. We count things as interactions, and I'll talk more about that in just a moment. That's a conversation. So eight interactions and the customer - we've seen this as well, they say thank you a lot. I always find that as interesting because - do, they think they're really talking to a human? I think they get so into the conversation that, at the end of the day, you've really helped me. Thank you so much. Now maybe they know that's going in a backend database and somebody's reading it, but they're interacting with an AI agent.
(13:00):
But it's not 200 facets, it's four or three. It might be size, might be color, might be thematic, like Comfy Fit is one of 'em that I recently remember seeing on this interaction. And they click on Comfy Fit. So it can be very touchy feely. These are facets that are not predefined. In fact, we don't want predefined facets. We want the AI to come up with 'em. So sometimes I call 'em facets - the real name for it is refinements.
(14:00):
Now, you asked about conversion and other things. We have to be a little bit careful about that 'cause I don't wanna take credit for something that's not totally ours. Here's the deal. When somebody's going and having an interaction with our AI agent, they're already an engaged customer.
(15:00):
What you really want to be able to see is AB testing of, again, people that use it versus people that didn't use it. The early results are that we are seeing a tick up in it, but I think it's still too early to say. You know, as we kind of work through the technology, but I wouldn't say it's dramatic because remember, the pool of people that are actually using it is still small.
(16:00):
Do you find that being able to put it in search as an example to get that usage up accelerates the learning of the AI engine behind it?
(17:00):
Maybe not there today, but definitely the ability to make that happen tomorrow. So, as we're talking about this - a question keeps coming up in the back of my head. If you are the buyer of this technology. I'm sitting here and I'm putting my CMO or VP of E-commerce hat on, and I'm looking at your technology versus, what's being deployed on Amazon and stuff. What are the critical decision factors?
(18:00):
But, if I'm having that conversation with you, "Ken, I'm like, Hey, look, I'm a very technology forward person. I'm like - I'm in." Versus the person who's like, "Eh, you know, I see all these shiny tech tools and I've spent lots and lots of money on tech over the years that, yeah, doesn't always quite achieve its promise."
(19:00):
Yeah. I go back to the crawl, walk, run strategy again for anyone, and I would say that if you're a little bit skittish about this, but you want to have the customers to start to experience it - just so you can get data. That your sole purpose is to get data. Maybe it's to ultimately increase conversion or abandoned carts or what have you.
(20:00):
So that makes the implementation and then to turn it on, you click a button and then it turns it on. And yeah, there's some configurations on the back end, but your vendor - us - are gonna make it look good and make it look right with a lot of configurations to make it look just exactly as you want it to.
(21:00):
When I go back and I look at a thumbs down, nine times outta 10, it's because we have inaccurate data or lack of data, I should say not inaccurate, but lack of data. Maybe one times outta 10, you have some disgruntled person that would probably be angry at any response that they get. Right.
(22:00):
You only see kind of, back to your earlier point, you only see the actual aerial component of it where they take an action and you see, oh, well, every time they take an action - their value goes up.
(23:00):
Remember, I wanna make sure that everybody understands cause we didn't cover this - how this AI stuff actually works. We have the option to go out and use a large language model. We're not using that and we're not doing that. We can tell the agent to go out and get external information, but you as an e-commerce person do not want external information coming into your website at all.
(24:00):
That means you're installing the tool and you are having to collect data for a period of time. What is that data collection period look like before you start getting effective results?
(25:00):
All we're doing, if you think there's magic to this, there isn't really. We are doing what we call rag. Basically all it is, is a very complicated search on your data to get the closest results possible. Then we send it out to the AI agent. That AI agent then makes sense of it and brings it back. That's where the training and other things come in.
(26:00):
That's technology that's leveraging all of this infrastructure and we just keep coming out with more and more stuff, and others are as well. But I did wanna highlight that. The actual underlying AI technology is kind of commoditized at this point. I know it sounds strange, but it's commoditized.
(27:00):
Yeah. And it's gonna evolve towards that vision that we painted at the very beginning of this, right? Where the website is not the place you're interacting. You're interacting on your phone with a voice assistant talking about whatever product it is that you're interested in buying.
(28:00):
So it's kind of like everybody will have their own Amazon, if you will, with all the sophisticated tools and all the things that Amazon has and all the infrastructure. Cause I can just talk into my Alexa, but instead of buying from Amazon, whether it's Alexa or something else, I can buy from the actual retailer and quote unquote cut out the middleman.
(29:00):
Ah! It drives me crazy!
(30:00):
Now it's majorly influenced. I get it. When it first came out. I don't know - I can't speak for them. But at the very beginning though, and I think that's where Chat GPT is according to their announcement, is I've actually interacted with it. They partnered with Visa to, I guess, enable transactions in some way. I haven't actually bought anything from it yet.
(31:00):
Oh, and Amazon, of course we can't lump out Amazon. Amazon and Google have the most data right. Apple has a great amount of data and they were saying it's kind of terrible how awful Apple is at artificial intelligence at the moment
(32:00):
And what Ken's company's doing is kind of starting to bridge the gap between the now and the future. And if you are an e-commerce retailer or you're any kind of retailer - you were just talking Ken about Home Depot and how there's nobody in Home Depot. Well, shoot.
(33:00):
It was just so frustrating because I'm gonna hit a button and wait, you know, there were probably six or seven aisles of things in this particular area that were all locked up. And then I couldn't find what I was looking for. Even if I could get through the case, I'm like, screw it. This is, I don't even know why I came here and I left.
(34:00):
So what Ken's also saying is that those of you in the e-commerce space are doing fine. You're great. you're well positioned to compete against the marketplaces by deploying these types of technologies. For those people who really want an authentic brand experience, because let's be honest, buying on Amazon isn't about price anymore it's about convenience.
(35:00):
Yeah. So, we've covered a little bit about Ken's technology at Microcasting, which I think, try it, talk to his team, experiment, have him do a demo. The demo is really pretty impressive I gotta tell you. I've done a couple of them now and I am really impressed with the initial results, and that's without any beautification of the page. Just the simple results and going, "Holy smokes, I wish I could have done that."
(36:00):
You did a better job than me of explaining the phenomena that happens every time we put this thing in. Is that customer service VP or success manager or whatever - calls us or gets a hold of it and says, "I want this for my customer service team." Inevitably, every time it happens, and I gotta tell you, the customer service teams forget the front end people.
(37:00):
If you can find a solution that's similar to what Ken's pitching. And again, I've seen the demo it's pretty cool. I've seen some of their live sites. It's pretty cool. Is it perfect? No, nothing's perfect, but they're getting better.
(38:00):
And then you started writing with it and it's kind of the same thing. And that's what we're experiencing now and it's just gonna get better. You're right, we do stumble on occasion. We're not perfect. Although, I will have to say the speed at which we're evolving is ridiculous.
(39:00):
Extremely rare. We are at the beginning of that same cycle when you and I first met with AI, except it's gonna go a lot faster. It's not gonna take 15 years. Cause I lived through that from 1995 to probably 2010, where we were still putting people on for the first time.
(40:00):
So all that tech still has to be compatible. The thing about AI is that it can be compatible with more tech now than crypto can be in any regard.
(41:00):
No, I know. No. Yeah, no. I was saying that when I was saying the first 15 years I was forgetting about the rest of it. I wanted to forget about the rest of it.
(42:00):
So thank you very much for your time. Y'all have a fantastic day. That's this episode of the Digital Velocity Podcast.
Welcome to the Digital Velocity Podcast, a podcast covering the intersection between strategy, digital marketing, and emerging trends impacting each of us. In each episode, we interview industry veterans to dive into the best hard hitting analysis of industry news and critical topics facing brand executives.
(01:00):
Very different. Yeah. Just in the year where AI has come and it's just amazing. You know, that even just in the last month, it's like, oh, we can now do this, or this. Even chat GPT has gotten so intelligent about what it's done, and literally it's been within the last two or three weeks that it's gotten this intelligent.
(02:00):
Yeah Microcasting's been around for about three, three and a half years. And as with any good company, we were a kind of a customer success platform initially and had some really good success in that. But about a year and a half ago, we started integrating AI into the customer success area and we're like - wait a minute. This is direct application to what we know and my entire team and myself have been in e-commerce for over 25 years. I'm dating myself, but I started my first company in e-commerce in 1995, which was the beginning of e-commerce version one.
(03:00):
If you think about the interactions with Alexa we're first generation and now can we apply this to the front end e-commerce experience and really take what has been going on for the last 20 or 25 years, literally what you and I created years ago together. L-shaped navigation, facet navigation, search on the top, products in the middle. It's still the exact same interface. It's boring.
(04:00):
I was sitting down with a friend of a few months ago, and we were talking about AI and where it's headed and I'm like, I have this question. "What happens to the website when you have AI agents working together?" So basically the concept of an AI agent is an application that's going out and doing something on your behalf.
(05:00):
Yep. I think you just define kind of the next generation of e-commerce and how people will interact with it. Whether it be a device like Alexa or whether it be a phone or a desktop, right? Can it all come together in that ease about, I know what you want and I can go and kind of shop for you.
(06:00):
And you know, we have a lot of data that says that people hate retail chatbots.
(07:00):
First thing that you're gonna do is obviously you're gonna go in and you're gonna have an experience. And first question that a lot of people ask is, " Do I replace my chat bot area?" You know - the floating footer. "Do I replace that with you guys?"
(08:00):
Then I say the run strategy, which people are just beginning with, right? This is very early, is doing augmented search. We're not replacing their search engine, they type in their search. And what we do is what Google does. We come up and we give you our AI version because they may have asked a question, they may have typed in keywords that didn't make sense to the search engine, but will make sense to us. So we're giving them our results. And then the results for the normal search will appear below that.
(09:00):
But it's fascinating that people are starting with their chatbots and it may be just a mix of the companies that you're working with versus some of the companies I'm working with.
(10:00):
So you guys have integrated this now at least half a dozen times at last count. It might be more now since the last time I talked to Christie.
(11:00):
But what we do want you guys to understand is like there's some really fundamental shifts happening in the way this type of technology is getting deployed and how the buyer, your buyer, interacts with it. So can you share some of the data that you have gleaned so far from the implementations you have about buyer behavior, what they find interesting, improvements in conversion rate or engagement rate?
(12:00):
So one of the reactions we've gotten is, wow, that was an eight interaction. We count things as interactions, and I'll talk more about that in just a moment. That's a conversation. So eight interactions and the customer - we've seen this as well, they say thank you a lot. I always find that as interesting because - do, they think they're really talking to a human? I think they get so into the conversation that, at the end of the day, you've really helped me. Thank you so much. Now maybe they know that's going in a backend database and somebody's reading it, but they're interacting with an AI agent.
(13:00):
But it's not 200 facets, it's four or three. It might be size, might be color, might be thematic, like Comfy Fit is one of 'em that I recently remember seeing on this interaction. And they click on Comfy Fit. So it can be very touchy feely. These are facets that are not predefined. In fact, we don't want predefined facets. We want the AI to come up with 'em. So sometimes I call 'em facets - the real name for it is refinements.
(14:00):
Now, you asked about conversion and other things. We have to be a little bit careful about that 'cause I don't wanna take credit for something that's not totally ours. Here's the deal. When somebody's going and having an interaction with our AI agent, they're already an engaged customer.
(15:00):
What you really want to be able to see is AB testing of, again, people that use it versus people that didn't use it. The early results are that we are seeing a tick up in it, but I think it's still too early to say. You know, as we kind of work through the technology, but I wouldn't say it's dramatic because remember, the pool of people that are actually using it is still small.
(16:00):
Do you find that being able to put it in search as an example to get that usage up accelerates the learning of the AI engine behind it?
(17:00):
Maybe not there today, but definitely the ability to make that happen tomorrow. So, as we're talking about this - a question keeps coming up in the back of my head. If you are the buyer of this technology. I'm sitting here and I'm putting my CMO or VP of E-commerce hat on, and I'm looking at your technology versus, what's being deployed on Amazon and stuff. What are the critical decision factors?
(18:00):
But, if I'm having that conversation with you, "Ken, I'm like, Hey, look, I'm a very technology forward person. I'm like - I'm in." Versus the person who's like, "Eh, you know, I see all these shiny tech tools and I've spent lots and lots of money on tech over the years that, yeah, doesn't always quite achieve its promise."
(19:00):
Yeah. I go back to the crawl, walk, run strategy again for anyone, and I would say that if you're a little bit skittish about this, but you want to have the customers to start to experience it - just so you can get data. That your sole purpose is to get data. Maybe it's to ultimately increase conversion or abandoned carts or what have you.
(20:00):
So that makes the implementation and then to turn it on, you click a button and then it turns it on. And yeah, there's some configurations on the back end, but your vendor - us - are gonna make it look good and make it look right with a lot of configurations to make it look just exactly as you want it to.
(21:00):
When I go back and I look at a thumbs down, nine times outta 10, it's because we have inaccurate data or lack of data, I should say not inaccurate, but lack of data. Maybe one times outta 10, you have some disgruntled person that would probably be angry at any response that they get. Right.
(22:00):
You only see kind of, back to your earlier point, you only see the actual aerial component of it where they take an action and you see, oh, well, every time they take an action - their value goes up.
(23:00):
Remember, I wanna make sure that everybody understands cause we didn't cover this - how this AI stuff actually works. We have the option to go out and use a large language model. We're not using that and we're not doing that. We can tell the agent to go out and get external information, but you as an e-commerce person do not want external information coming into your website at all.
(24:00):
That means you're installing the tool and you are having to collect data for a period of time. What is that data collection period look like before you start getting effective results?
(25:00):
All we're doing, if you think there's magic to this, there isn't really. We are doing what we call rag. Basically all it is, is a very complicated search on your data to get the closest results possible. Then we send it out to the AI agent. That AI agent then makes sense of it and brings it back. That's where the training and other things come in.
(26:00):
That's technology that's leveraging all of this infrastructure and we just keep coming out with more and more stuff, and others are as well. But I did wanna highlight that. The actual underlying AI technology is kind of commoditized at this point. I know it sounds strange, but it's commoditized.
(27:00):
Yeah. And it's gonna evolve towards that vision that we painted at the very beginning of this, right? Where the website is not the place you're interacting. You're interacting on your phone with a voice assistant talking about whatever product it is that you're interested in buying.
(28:00):
So it's kind of like everybody will have their own Amazon, if you will, with all the sophisticated tools and all the things that Amazon has and all the infrastructure. Cause I can just talk into my Alexa, but instead of buying from Amazon, whether it's Alexa or something else, I can buy from the actual retailer and quote unquote cut out the middleman.
(29:00):
Ah! It drives me crazy!
(30:00):
Now it's majorly influenced. I get it. When it first came out. I don't know - I can't speak for them. But at the very beginning though, and I think that's where Chat GPT is according to their announcement, is I've actually interacted with it. They partnered with Visa to, I guess, enable transactions in some way. I haven't actually bought anything from it yet.
(31:00):
Oh, and Amazon, of course we can't lump out Amazon. Amazon and Google have the most data right. Apple has a great amount of data and they were saying it's kind of terrible how awful Apple is at artificial intelligence at the moment
(32:00):
And what Ken's company's doing is kind of starting to bridge the gap between the now and the future. And if you are an e-commerce retailer or you're any kind of retailer - you were just talking Ken about Home Depot and how there's nobody in Home Depot. Well, shoot.
(33:00):
It was just so frustrating because I'm gonna hit a button and wait, you know, there were probably six or seven aisles of things in this particular area that were all locked up. And then I couldn't find what I was looking for. Even if I could get through the case, I'm like, screw it. This is, I don't even know why I came here and I left.
(34:00):
So what Ken's also saying is that those of you in the e-commerce space are doing fine. You're great. you're well positioned to compete against the marketplaces by deploying these types of technologies. For those people who really want an authentic brand experience, because let's be honest, buying on Amazon isn't about price anymore it's about convenience.
(35:00):
Yeah. So, we've covered a little bit about Ken's technology at Microcasting, which I think, try it, talk to his team, experiment, have him do a demo. The demo is really pretty impressive I gotta tell you. I've done a couple of them now and I am really impressed with the initial results, and that's without any beautification of the page. Just the simple results and going, "Holy smokes, I wish I could have done that."
(36:00):
You did a better job than me of explaining the phenomena that happens every time we put this thing in. Is that customer service VP or success manager or whatever - calls us or gets a hold of it and says, "I want this for my customer service team." Inevitably, every time it happens, and I gotta tell you, the customer service teams forget the front end people.
(37:00):
If you can find a solution that's similar to what Ken's pitching. And again, I've seen the demo it's pretty cool. I've seen some of their live sites. It's pretty cool. Is it perfect? No, nothing's perfect, but they're getting better.
(38:00):
And then you started writing with it and it's kind of the same thing. And that's what we're experiencing now and it's just gonna get better. You're right, we do stumble on occasion. We're not perfect. Although, I will have to say the speed at which we're evolving is ridiculous.
(39:00):
Extremely rare. We are at the beginning of that same cycle when you and I first met with AI, except it's gonna go a lot faster. It's not gonna take 15 years. Cause I lived through that from 1995 to probably 2010, where we were still putting people on for the first time.
(40:00):
So all that tech still has to be compatible. The thing about AI is that it can be compatible with more tech now than crypto can be in any regard.
(41:00):
No, I know. No. Yeah, no. I was saying that when I was saying the first 15 years I was forgetting about the rest of it. I wanted to forget about the rest of it.
(42:00):
So thank you very much for your time. Y'all have a fantastic day. That's this episode of the Digital Velocity Podcast.
Popular Podcasts
Fudd Around And Find Out
UConn basketball star Azzi Fudd brings her championship swag to iHeart Women’s Sports with Fudd Around and Find Out, a weekly podcast that takes fans along for the ride as Azzi spends her final year of college trying to reclaim the National Championship and prepare to be a first round WNBA draft pick. Ever wonder what it’s like to be a world-class athlete in the public spotlight while still managing schoolwork, friendships and family time? It’s time to Fudd Around and Find Out!
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.
The Breakfast Club
The World's Most Dangerous Morning Show, The Breakfast Club, With DJ Envy, Jess Hilarious, And Charlamagne Tha God!