Aaron Conant  0:18  

Happy Tuesday, everybody. My name is Aaron Conant. And I'm the Co-founder and Managing Director here BWG Connect, a giant networking knowledge sharing group with thousands of brands. Going to do close to a hundred in person events this year, ping us if you want to be in Vegas, any of those who are in tier one cities across the US. Couple 100 webinars, have a podcast where we capture these exact topics. So you can just download on your free time, it's just thedigitaldeepdive.com, you want to check that out. We're gonna start here in just a few minutes after the hour, and we're going to wrap up with a few minutes to go as well. One thing we want these to be as educational informational as possible. So drop any questions you have in the chat or the q&a. Or if you got my email, aaron@bwgconnect, just email them to me directly. And we'll try to tackle as many of those real time as we can. And I mean that I spend my time talking to brands, and hey, what's what's the topics we should be talking about. And the same topics come up over and over again, we host an event like this. You know, AI don't think anybody is really surprised. This was a topic that's come up. Jordan, you and I are talking, you know, at the IHA executive event here and another month or so. Lot of feedback in the brands I'm talking to coming out of etail was there's a ton of AI talk. And then there's articles all over the place of, hey, are people waiting on ChatGPT and AI and all this other stuff is just exhaustive. So anyways, you're kind of a thought leader in this space. we've chatted a few different times on it. So I'll kick it over to you kind of a brief intro on yourself and Coalition. And then we'll kind of kick off the conversation sound good?

Jordan Brannon  2:02  

Yeah, sure. Yeah. My name is Jordan Brannon, I think I've had a chance to connect with many of you at some point. President and Co-founder at Coalition, we are a digital agency primarily focused in eCommerce, mostly working with well established small businesses and up in terms of our focus area. I appreciate sort of the thought leader call out on AI. I think we are aspiring to be a functional leader in terms of AI rather than a conceptual one. So we're trying to use AI tools and a lot of different areas within our business, both from a marketing practice, but also just in operations and things like accounting and recruiting and places like that. And so certainly, I think that the conversation to probably talk a little more about eCommerce and digital marketing. But if there's other questions, you kind of want to bounce off me later. Feel free to because again, we're we're testing and using AI in a lot of different ways.

Aaron Conant  3:04  

Yeah, I just typed this in to ChatGPT, where you're talking to give us an intro for a webinar on AI with Jordan Brannon. It starts off, ladies and gentlemen, distinguished guests and enthusiast of the ever evolving landscape of AI, Jordan Brannon is a seasoned expert renowned for his profound understanding. It's, it's pretty impressive, I decide to avoid generally.

Jordan Brannon  3:23  

Now I have to stop disappointing our AI overlord.

Aaron Conant  3:27  

So, I mean, one of the things, though, is, we are hitting the busiest time of the year. And, you know, at this point, you know, there's all of these different things we're required to do and yet AI is still top of mind. You know, are there practical use cases, you know, as people are looking to pull levers into Q4? You know, they're popping up, we'd love to hear some. I mean, obviously, we need to have this conversation right? Now, and anybody wants to tackle one more thing during the season and make it even more complicated than last year, but they got to have this conversation, right.

Jordan Brannon  4:07  

Yeah, I think there is a lot of value in that. And I think you're trying to get I think last year, about this time, I think we were having a conversation about, you know, Quarter 4 planning, holidays, things like that. And in some ways, it's just as there's a lot of pressure, you've got more Prime days you have the month of November is now sort of the holiday selling month instead of just Cyber Week and then you still have Cyber Week. And so I think in some ways, getting some traction with some of these AI tools, from a practical standpoint now is sort of the time and the opportunity to do that. And, you know, again, hopefully we can make this valuable in that regards where people can start to get some some rubber on the road and find some ways that they can use AI now and then hopefully makes quarter 4 a little more effective and and easy, rather than challenging and problematic.

Aaron Conant  5:02  

So if we just kick this off, if we can start at the like the most basic level, just for everybody that's on today and people who will listen to this later on. Can you offer like an over simplified version of like the basic building blocks around AI that people should be thinking about? And I also say this from the standpoint, even the seasoned people, like who have to explain this internally, like, Hey, by the way, executive team, here's how you need to think about AI.

Jordan Brannon  5:33  

Yep, I do, I participate in a number of different AI forums and in different formats. And there is a lot of terminology, and a lot of disagreement about the proper application of that terminology. But I think, you know, probably from a basic components standpoint, you know, most AI's are sort of built on, you know, maybe three or four things I think you have, sort of the data, it's the raw information that an AI can use to essentially learn or operate off of, and data can really exist in a lot of different forms or formats. So when we think about data, one of our disconnects, is we sort of innately go to, you know, spreadsheets, sort of, you know, I think in my head when I first started thinking about data as a resource, I thought about, okay, Excel files, Google Sheets, CSVs, that sort of thing. But from an AI standpoint, you know, you can feed AIs data in a lot of different formats, text, images, audio, video, a lot of different things. And obviously, that kind of gets parsed and parceled into a workable database or data set. Again, most AIs are functioning with some form of a model. And so it's sort of the central piece of AI, as we've become a little more familiar with the term models over the last several months. And they help the AI perform a particular task or function. It's, you know, LLM 's or large language models is sort of the one that's been probably most in the news, that it can be built in a bunch of different ways. So that's one of the things I think about as we kind of arrive at these AI tools and solutions, because of different sort of underpinnings for how their model is created. And there's a lot of innovation actually happening here as well, because they can be quite resource intensive to create. And so machine learning, deep learning, as well as phrases that you're familiar with natural language processing, maybe some examples there to kind of have as a reference for models and how they're created. And then beyond that, you have the infrastructure. And so if you've ever tried to deploy an open source, AI of some kind, there's actually quite a few of them that are quite good. One of my favorite places to play is with some of these open source solutions. You need to have some hardware, a computer that's able to do that, and some software solutions that will help you operate and utilize that either locally or in the cloud. And then finally, a big one. And probably the thing that really sort of set off the wildfire of AI this year, is you need an interface that's allows you to interact with with the AI in some way, shape, or form. You can think about like a GUI, a graphical user interface, API's, even application programming interface. I've been interface there. And then of course, I think ChatGPT, of course, is that sort of simple text interface that we're all really familiar with. And I think, again, that sort of simple text based interaction is really sort of maybe what made the man behind the curtain suddenly seemed very accessible to us.

Aaron Conant  8:42  

So I mean, even from a novice standpoint, is it? Like how important is it to understand you know, these concepts? Is it worth doing? And then again, binders, people, if you want to drop questions, you can drop them the chat or the Q&A, and we'll tackle them as we go.

Jordan Brannon  8:56  

Yep. Awesome. I think so I think, fundamentally, trying to parcel out what AI solutions to us and which ones to pursue and thinking about the news stories we see, I think you have to sort of have some basic concept for, for how these API's are built. Because it ultimately will influence sort of which ones are more relevant to you as a DTC brands, especially if you're not operating with an enormous budget to sort of run your own AI discovery sort of effort. A lot of us are sort of, we've got busy day jobs, we have metrics to hit. We sort of hear all the buzz in the news about AI and how useful it is. And so we're trying to find a way to sort of learn and figure that out alongside of still delivering great row as and so I think when you start to get a good understanding of some of the basic components, it does help you to start to filter certain tools in or certain tools out. And I'd really highlight I think the data side of it, because of the data one I think is probably the area where Most good AI solutioning is living and dying for DTC brands or people working with them.

Unknown Speaker  10:07  

Yeah, I mean, if we

Aaron Conant  10:10  

talk about then, like the 2023 chat GPT, right, their origination of AI for most people is what they're looking at and using those components to help substantiate, you know, their roles more clearly. Like, I think people are still trying to wrap their head around. Yeah, what, what are the tools? How do I use them? Should I be creating my own tools? Should I use wait for service providers to build it into their services? And then I just acquire them? But anyways, like? Yeah, yeah, I

Jordan Brannon  10:47  

think, yeah, maybe think about like chat, GBT is sort of an example. And that also sort of maybe trying to answer that question of, should I be building my own? Should I really be looking at other solution providers and service providers first? You know, I think, again, if you sort of think about, like, what makes chat GPT successful using those sort of four components, one, they have this massive data set of really content that's been published online, primarily. And then they all sort of sort of accumulating additional data as people interact with it. You know, of course, that also feeds into the privacy and security concerns that you get with chat GPT, they just launched a enterprise or business product, I think in the last couple of days. You know, chat GPT uses sort of that natural language text. And I think one of the things that makes it deceitfully powerful is that almost everything we do relies on natural language, right. And so there's sort of this perception that chat GPT, and other tools, like it can do anything, because we communicate everything. And the reality is, of course, that we kind of know that there can be kind of ghosts and Gremlins in the system there that create false positives, right? Their own mythologies, and they can show up if we are overly reliant on them. And again, you know, chat GPT is a Text user interface or tweet is sort of been very approachable. And so I think, again, if you're sort of a novice, or if you think about how to use AI, today, thinking about sort of what the interface is for you, and how you can tie that into what you're doing as a workflow can be really important. And then, of course, a lot of the AI tools, and this is probably where a lot of businesses, especially ones that don't have huge budgets, probably start to look at services instead of building their own chat. GPT has its sort of own infrastructure, right? They have their own servers, connections to the internet API endpoints to rely on. All of those things are sort of provided for you for for pennies, in theory. And so I think that's where a lot of businesses again, are opting to go for services instead of looking at you know, kind of a build their own sort of thing.

Aaron Conant  12:54  

Yeah. Awesome. John writes in here is kind of like top of my mind to from marketing perspective, what are the top two or three practical applications of AI? And so I'm thinking, you know, exactly, John, where I kind of wanted to go is if we turn our conversation to the direct to consumer side, eCommerce, there's two sides of that, right? There's like, how much of it is a bunch of baloney? And how much of it is like actual substance? And that falls right into what are some stuff out there we can use it for and look for?

Jordan Brannon  13:27  

Yeah, I maybe just answer your question. First, I think like the, I think anytime they see AI sort of presented as a superhuman in a can, you're dealing with BS. So if someone is inviting you to the webinar, and saying, Hey, this is going to solve all of your marketing problems, just one click install, and voila, that's probably going to just be some dishonesty in marketing. Someone's probably trying to do some fundraising with a VC fund someplace. And so that's sort of an instant skeptical thing for me. I think if we were to sort of look at sort of the most accessible AI tools for brands to deploy right now, I would probably say, chat bots, really low hanging fruits. Again, LLM is being one of the major places we're seeing a lot of just utility for, for AI with D2C eCommerce. I think AI personalization tools are again, really a strong area, something that was pretty difficult and time consuming for brands. Enhanced product search, again, fairly heavy lifts, and fairly time intensive resource intensive to really get right. Marketing content generation is one that actually have a little bit more of a coin flip attitude towards. Same with marketing automation. Really like a lot of the code creation tools that are coming out with some of these AI solutions. And then also data analysis. I think, you know, kind of those are sort of the core areas we're focusing on. We're actually seeing some tomorrow is some good or some bad. I think those are all areas that have a pretty big impact on D2C eCommerce right away. And so I think getting some footprint in there helps you as tools improve, you begin to sort of understand what your benchmarks are for some of these AI solutions. And so I would probably, strongly recommend beginning look at AI additions or supplements, for most brands sort of, in a way that begins to help supplement their normal day to day workflow. And I think for a lot of people, that's probably gonna be the best point for point of entry for AI for them.

Aaron Conant  15:32  

I'm thinking of a future like webinar, and then I want to dig into that a little bit more. And then we have a question that comes in about recommending AI tools. And it's actually that one made me think about it was, should we do two hour webinar with 15 minute segments where we're talking about the AI tools that you recommend, right? Because that's where everybody is. There's a trillion out there, if you just go out there, like, but you deal with so many brands, and you're inundated. You can try all these ones out. And you can see which ones work and don't it'd be cool. And I don't know if it's an hour, or if it's three hours, but these 15 minute segments where, hey, here's an AI tool, that's a chatbot. Here, that's something that we should probably do, I think people and then bring whoever they are on as well, to kind of walk through some examples. A bunch of you are like, Absolutely, because that's where people want to get to is, I know, AI is out there. I know, I should be testing and learning. But when I do the Google search, a trillion things show up.

Jordan Brannon  16:36  

A lot of AI generated spam shows up yeah.

Aaron Conant  16:44  

Yes, and so we should put that together soon. So everybody on this call, look for that follow up? Because we'll try to do that. And I don't know if it's three weeks out, or if it's two weeks out. But Jordan, we should put that together. Because they're saying yes, we would love just to have a recommendation point us in the right direction, and we can at least prioritize work. But if we pull the thread a little bit, and you know, these AI additions or supplements and the things that we're talking about, you know, can you walk us through that just a little bit more that would be, that'd be great.

Jordan Brannon  17:17  

I'll try to call out some of the tools that we are using it but maybe it's sort of like the precursor on that, like, you know, one of the challenges, I think you touched on this era. And I think a lot of people here are probably seeing the same thing is that the more you sort of dig into the AI news, or just as sort of like, it's more bewildering, there's, there's, again, probably because AI content generators are so prevalent, there's more and more content coming at such a fast pace, it's really hard to sort of get a handle on what's actually happening, what matters. When it first started, I was sort of I came to talk to my wife was like, I think I'm suddenly the old guy, I think I've become too old to understand what's happening in tech. You know, I'm committed to my pager, or my old candy bar phone or something like that.

Aaron Conant  17:17  

And it's not like the remote control you just handy or, you know, 13 year old and they figure it out, right?

Jordan Brannon  18:08  

Yeah, I have to go to my eight year old for smartphone problem solving. Yeah, I mean, I've spent a lot of time as sort of going through a lot of different demos, you know, for Shopify apps or stores for, you know, Big Commerce for Adobe, you know, you know, standalone tools that sort of work in conjunction with what you're doing. And honestly, a lot of it just doesn't seem to amount to much. And so, one really useful, you know, kind of filtering tool for me has been finding sort of things that work with the software that I already am using in my day to day, in sort of my normal task list and beginning to sort of look for AI there versus looking for that big AI solution that's going to do it all, solve all the marketing problems dramatically, you know, save massive amounts of dollars on on my ad spend, you know, wipe out the whole creative department just so I can create ads faster, or just more budget efficiently. And so I would say like, again, just maybe as a starting point for everybody here, look for those AI tools that just kind of cover your day to day. For me like that the places I'm seeing some of the most immediate value are like meeting transcription, note taking. I love some of the email clarifications, we have almost 300 people here. And we're big on sort of collaborative feedback. And so I can be involved in email chains that have you know, 30 different people with input and so that the summary of some of those email threads is really helpful. We do a lot of content marketing, a lot of different industries. And so some really helpful like Chrome extensions that help you extract information from webpages, even from videos. I have a couple of AI tools. I'm testing right now with Google Analytics to help with sort of flagging behavioral trends, things like that. And so again, also just data visualization. So if any of you are sort of the smart person in the room at your company relating to eCommerce, and you're running into issues where you're trying to communicate that to an executive team or somebody else through data and they sort of glaze over, as soon as you start to show them a spreadsheet with, you know, ad spend, and you know, performance changes, in particular ad groups, being able to use some of the AI tools to actually help with data visualization is really helpful. So those are some of the places that start. And again, if you sort of look at what your tool set is, that your software's are, and then sort of look for the AI supplements to those specifically, that's usually gonna be a really good place to begin.

Aaron Conant  20:41  

So just in lieu of this next webinar that we should do, any particular tools that are worth calling out, or using today, like top of mind

Jordan Brannon  20:52  

Yeah, personally, I think like, we're using Fireflies a lot fireflies.ai. That's a meeting transcription meeting summary, as timestamps helps to sort of identify key action items that are coming out of meetings, that's really been a useful one. Again, just sort of in a day to day flow. I'm using SaneBox in my personal email, it's been really helpful. I get a ton of emails, I sign up for all of the webinars, so I can sort of be in the know, but then I inevitably get spammed, and I can't unsubscribe fast enough. So SaneBox has been a pretty fun one to play with. There's actually a number of tools that sort of help with better spam filtering, better identification of what's really critical without you having to set up a lot of filters and folders and things like that. An oldie but a goodie is Grammarly. Just if you're looking for kind of writing and writing enhancement, grammar has really been helpful as one that you can, again, just clarify your writing, or have some some writing done for you. I'm a Chrome user or Chrome power user. And so I'm using a couple of Chrome extensions to help with keyword discovery inside of specific web pages, content summary again, or being able to pull out key points of content, pull links from content automatically with AI assistance. And then also, there's a couple that I use for summarizing YouTube videos. So again, if it's an hour long video, I have to get through the the intro outro stuff, I can get a summary from from some of those. And then I'm doing a lot of testing just with the biggies ChatGPT, Bard, Claude, if you haven't played with Claude 2 from Anthropic that's kind of an interesting one, it's supposed to be safer AI. So there's some some ones to play with there. And then maybe also, I would suggest, again, like just going back to some of the software, you're using already, a lot of really great tools for Google Sheets and Excel. Both of those are beginning to roll out some native AI features. If you haven't, opt in to Google's Workspace experiments. If your organization is a Google Workspace organization, inside of Workspace experiments, you can get access to some early release tools. Google has their what they're calling Duet, which is sort of their AI enhancements to Google Sheets and Google Docs, that's a really good one to get involved with, and can help with porting data back and forth between some of these different solutions, taking data out of an Excel sheet, turning it into a presentation, things like that. And then again, for eCommerce, since that's the big focus, a lot of our eCommerce platforms and the marketing tools you're using already are adding in some incremental AI additions. And again, those are going to be your best places to start just because they rather than sort of requiring you to learn something entirely new. There's usually some way of supplementing expediting improving sort of that day to day you're already already taking on.

Aaron Conant  23:44  

Wow, awesome. You know, at the end there, you mentioned like eCommerce platforms in a couple of things going on there. One is it's easier to bounce around between platforms, and it's ever been on. And therefore there's this constant reevaluation, am I on the right platform, and everybody hopes they are so they don't have to re platform. But that kind of begs the question. Is there one of these it's more of a leader in the space of tackling AI as a whole?

Jordan Brannon  24:15  

Yeah, unfortunately, unfortunately. So can we pause recording for a minute? Just kidding. I probably the one that I would say is probably more out in front of some of the others is Salesforce.

Aaron Conant  24:28  

Everybody's thinking like Big Commerce or Shopify.

Jordan Brannon  24:32  

I know. I have been a stick in the mud. So again, I started eCommerce came from running small eCommerce stores that were drop shipping. And so I've been a big believer in open source eCommerce where a knowledgeable person can customize. I was big into the Magento community in the early days, and then saw sort of the magic of SaaS happening and started using Shopify and Big Commerce before most everybody else did. And we're going to develop her marketing partner with But both of them for a long time. But I think honestly, right now, Salesforce is probably the one that has the most robust solutions in place natively. Again, some of the claims are not well matched to reality. But just I think based on client feedback and input, you know, Einstein and some of the other AI tools that are in place for Salesforce, it seems to be more productive. And I think the key advantage for Salesforce outside of sort of the enterprise budget is they overcharge everybody for years, and nobody knows why. And apparently, it was to build Einstein or whatever. But I think it is sort of that, that sort of monolithic mentality, they just have a lot of big, bloated tools that do a bit of everything across an org chart and an organization, including eCommerce, including marketing, including analytics, including CRM. And so you know, suddenly, a, an effective AI tool can be a better way of integrating all of those things together and making that that sort of disparate service offering a little more cohesive. And so I think that's one of the biggest values for D2C brands in eCommerce right now, is sort of pulling together, you know, marketing tools, commerce, tools, sales and customer service type offerings. You know, that actually is a pretty helpful place for AI to be deployed. I think behind Salesforce, we've seen Shopify have some splashy announcement, some are additions, their features aren't particularly unique, or not necessarily super competitive, but they are convenient. And that is, I think, one of the things to look for with some of your AI solutions. Sidekick, which is the Shopify store system seems like maybe one of the better values that's sort of unique to eCommerce offering, you know, having that ability to sort of have a helpful assistant to sort of run some of your things conversationally for you. It seems like a great sort of ad for an eCommerce platform. And again, I think at this point, just if you're on an eCommerce platform, you haven't seen something about an AI content generator, you're not seeing any sort of news about product reviews, summarization, translation, things like that. You know, chances are, you're a little further behind, but but a lot of the ones that we work with are kind of already in place for that.

Aaron Conant  27:18  

And do you do you see, Shopify, Big Commerce like, you think they're gonna turn the corner and just catch up? And it's just the fact the budgets were there for so long. And I mean, I understand the Salesforce side, it is huge bloated. But if you can, don't have to have a person doing all the connections, if you just asked the AI AI to make the connections, and it's made.

Jordan Brannon  27:41  

It's interesting. So so Big Commerce, has really been pursuing sort of, again, they they're big into that sort of complexity story. They're advertising today is more focused on enterprise and b2b. But price competitively, you know, there's still a great solution for small businesses, if you have some complexity built into sort of your catalog, your customer base, your transactions, I would make a lot of cases, I would recommend them over Shopify, if you sort of complexity is your thing. And so, but but again, like part of their thing is they do this open SaaS mindset, which is they really look for partners to provide the best solutions. And so with them, I be excited to see them introduce like an AI tool to sort of help you manage the backend of your Big Commerce store, kind of like Sidekick. But my guess is they'll probably take the approach of looking to third parties to sort of help elevate sort of AI solutions for them. Shopify, trying to be everything to everybody, I expect them to see a lot of like, sort of that low hanging AI, you know, background generation for images, sidekick content generation, I expect them to sort of be a little more aggressive on that front. The one interesting play, and especially with Klaviyo's, you know, IPO and all that sort of thing. Shopify is consumed a lot of other sort of software solutions. And so I think, again, there may be some signs from them where I see a conversate, a question came in about an LMS in a Shopify can very easily begin to sort of think about what Salesforce is doing, and start to try to accumulate additional software we haven't natively thought about as being part of an eCommerce platform, but are really important to running a successful eCommerce Store. And so that that might be one where I'd say like, expect Shopify to be a little bit more aggressive on AI tools and new feature announcements to sort of support them.

Aaron Conant  29:42  

Awesome. So you've mentioned things like Klaviyo and I'm thinking about it categorically. You know, are the best tools that are out there for you if we kind of like slice it up?

Jordan Brannon  29:55  

Yeah, I think um, I think but maybe by category just going to get through sort of me like Um, kind of what the tool specifically does.

Aaron Conant  30:02  

Yeah, yeah, exactly. So yeah. Not, not beauty if I should have taken the different word, but not beauty or fashion or anything?

Jordan Brannon  30:09  

Yeah. So I think again, just by volume of experience, you know, we're mainly working with small and mid market eCommerce brands. So, you know, we have, you know, I think our largest one is largest client is into the nine figures. So, not small. But we predominantly by volume are working with small and mid market, eCommerce companies. And so I think I'd say again, probably that categorically that the highest ROI I'm seeing comes from product search, product recommendation, and customer support, kind of those are the top tier. And I think going down from there, I'd say, you know, content creation is, again, middling, which surprises a lot of people when I'm talking to them. You know, and probably some of the content marketing solutions as well, a code creation is really helpful. But if you're an established eCommerce brand, the solutions may be a bit more redundant. So it may not be super helpful, unless you're very tech, heavy, and tech forward in terms of what you're doing. If we have some merchants, we're in a sort of very headless and very composable in mindset. And you know, how the tech team to support it. And so some of the code creation is really kind of fun for them, and it fits in well to what they do. And then again, data analysis and reporting are big for me, I think those are some areas where you can get a lot of value, but just on their own part, necessarily producing that that ROI that seems to be advertised, they need a little bit more of that hand holding that sort of assist. And then probably the same thing is true for personalization, sort of customizing that the shopping experience merchandising content to fit a particular person or persona. I think a lot of those so far for us, we've seen sort of a middling outcome, you know, in terms of cost of investment, and what we're able to get back from the AI itself.

Aaron Conant  32:02  

If we pull the thread on the content marketing side, it's just interesting, because, you know, that's top of mind for a lot of people. And it seems to be the one everybody's diving into first. And it's the most prolific just even conversations I'm have everybody's trying to tackle the need for content and how to use it. And you said, hey, it's kind of like in the middle, would love to hear? Is it worth doing after for people as well? You know, it's kind of comes to mind?

Jordan Brannon  32:29  

I think it is. But I think the biggest reason we're seeing sort of that middling outcome is rapid adoption is happening without really consideration or education around, you know, how they work, why they work, why they don't work. And so I was going to pin sort of the downsides on a few things, I just say, it's, it's what I'm seeing a lot of happening with AI marketing, AI content marketing is all it's doing is allowing bad marketers to produce more bad marketing faster. And that it seems effective at first glance. And so that's I think it's probably where I'm seeing sort of a lot of the downside of for, for that sort of content marketing experience.

Aaron Conant  33:13  

Do you have people or are there ways that you have like AI assisted content generation marketing, or people are using it well?

Jordan Brannon  33:23  

Yeah, I mean, I think it this goes back to sort of evaluating and sort of figuring out which AI solutions to use is, you really want to understand that data that's driving what the what is behind sort of the AI and how it operates, you know, LLMs, and other sort of generative AI tools, that there's this big external data set that's being used to produce sort of a common result and outcome that really seems appealing, again, anything about LLMs are basically kind of a coin flip as to what you know, the next best word should be based on on the dataset they've been trained on. And, and the data is really only half of the value that they can bring in. And so from a marketer standpoint, you have to understand, like the other half of the potential value is making that brand centric, instead of universally sort of appealing, which is, you know, what, again, a lot of these AIs are just sort of based on, you know, billions or trillions of different parameters. This is what is right. And I think that insert in sort of like a brand value, I don't necessarily want to be generic, right? Like, I don't necessarily want to be what Reddit thinks my brand should be, or a company my space should be. And so I think for AI is that sort of have massive volumes of information about people generically, you have to bring the other half of the value as a marketer, and become a little bit more disciplined and discerning and sort of developing a dataset that really reflects your audience reflects your product or service reflects marketing messages and content there, otherwise, you're not really gonna be able to produce a consistent ROI and you're really not gonna be able to produce something that You can build off of. And I think brands are often with content marketing tools are able to suppress sort of the cost of bad marketing with AI, by doing bad marketing at a lower cost is that they just are not necessarily doing something more effective or as a better outcome. Oh, maybe one other thought there. Probably the most important is that the issue compounds, right, as you start to put that in front of customers, they get turned off by a stream of less relevant and less engaging marketing based on their interests and experiences.

Aaron Conant  35:36  

So a couple things, but I want to want to tackle this, can you give examples of bad marketing content created by AI?

Jordan Brannon  35:44  

Oh, yeah, I mean, yeah, we had a brand come to us. So we probably are best known for our SEO work. Had been an early adopter of widespread AI content. And I don't want to say early, because they this was really like, they jumped in sort of late fall of last year, and ended up hiring us in the spring of this year, after some of the helpful content updates that rolled out, they basically had, you know, asked, you know, their AI solution to sort of generate all of the possible keywords that they could, you know, need to cover and then had these long form articles that were just very vaporized, and empty and Harlow that were created to sort of try and spin up these landing pages. So they had all this content, and they been able to rank some of this content, get traffic to some of the content, extremely low conversions, you know, low interaction, they were actually transacting less in February than they had been in October before sort of launching all of this content. So there seem to have been some of backfiring that I kind of had mentioned, where they produce all this content, they put this content in front of not only net new customers, but also in front of existing customers. And the consequences, nobody was really interested anymore. And then they got hit by a helpful content update from Google where Google looked at it and said, Well, this is all AI generated, not necessarily bad, but it's all lacking and uniqueness. It's all lacking and expertise and authority and all these other things. And so it really kind of wiped out some of the ranking on some pages that previously were doing pretty well with less content. And so that's probably a probably the best example that comes to mind. Because there's had to be a fair bit of work to sort of clean that up and kind of help them move away from sort of this thought that AI is going to help me generate this great marketing outcome faster. And by all accounts outside of impressions and clicks, it seems to have really backfired more.

Aaron Conant  37:41  

Yeah, I mean, it's just interesting, because the reality with AI is, you know, garbage in, garbage out. And so there's still a ton of hard work that has to be done. Right, kind of, to your point, the feed into the engine to have a kick out great stuff. And if you trade it over and over again, to be great, to be great to be great, it kicks out great stuff. But it takes a lot of time and effort to train it. You can't just hit go.

Jordan Brannon  38:11  

Yeah. Maybe two thoughts on that. That point. Actually, I think one of them is the interesting thing we're seeing with a lot of AI is especially when you over train them, is you start to get worse outcomes, partially because you get them so narrowly focused on one specific thing, that you sort of lose the value of the broader thing. So you know, sort of finding that balance in that tension between, I've got ChatGPT, which is trained on, you know, you know, more of the human language than any other tool in human history. And then I'm going to try to take that and specialize that so, so much for my audience or brand with prompts or with, you know, a paid solution. All of a sudden, what I'm getting is less appealing to an audience than if I had sort of tried to find that hybrid. So there's a little bit of a tension there. And maybe kind of a related point item, kind of thinking about this conversation, but not directly related to this morning, I made a post that said, you know, if your job is is that your day to day work is sort of really easy to do now, because of AI solution. Your job is the one that's most likely to be replaced because of AI solutions. And so you need to find a way to elevate what you're doing in a way that you offer a distinct value, otherwise, you're going to be out. So those are two sort of random thoughts. Sorry to interrupt your your question there.

Aaron Conant  39:32  

No, that's fine. Because I was just thinking from the standpoint if you're going down that path, and you're creating a stream of less relevant, less engaging marketing materials based on the interests, like that's a problem, like, how do people tackle that problem? Then my guess is it's going to be different from a small to mid size to a large company. Now based on knowledge that's out there size of budgets, how much is already been done, right? It's anyway, so.

Jordan Brannon  40:02  

Yep, got Yeah, yeah, man, it's a really good way of viewing it because I think there is some very distinctive things that you can do that are different, you know, if you're a small brand versus, you know, an enterprise brand, obviously, enterprise brands, there's a lot more legal restrictions, there's a lot more existing sort of teams to sort of work through and work around, there's usually more software solutions. And so that actually creates a problem in and of itself, for a lot of people looking to roll out AI solutions, you sort of have to wait for the organization to sort of adopt a thing, before you're really able to see them be able to sort of adopt some these tools that are a small brand, the challenge is that you don't necessarily have a lot of breadth to your data, you may or may not have a very big audience, you may not have a lot of audience history, you may not have a lot of data gathering tools that are there to help you sort of accumulate. And so I think for anybody sort of wanting to do AI, this is probably the place that it really starts, sort of the low hanging fruit, you know, in the title, start really making sure that you're understanding how you're gathering data across the different touchpoints in your role, and your function your team. And really making sure you can begin to sort of surface that data in a way that you can then utilize as part of your AI aspirations and your AI tooling. You know, analyze who is shopping with you who is buying from you start considering was driving their purchases. Again, some of those AI spreadsheet assistants that I talked about earlier, can be really helpful. If you just pull a data export from your eCommerce store, look at locations names, time of purchases, size of purchase, you can find some correlations there, if you if you're gathering that data, you know, make sure you have some of the basic free tools in place. You know, Google Analytics for as much of an annoyance as it is, you know, is a free tool. If you're on Shopify or Big Commerce, we can help you with implementing GA for problem solving that if you're struggling with some of the reports and things, if you're looking to sort of move a little beyond that, and you're starting to try and develop that first party data practice, which I think is important, and it's pretty approachable now for brands of all size. Look at other tools, when we're using frequently as a relatively recent startup called fueled.io. As first party data platform works really well with Shopify, Big Commerce that has a lot of pre built connectors for other platforms. Look at Microsoft clarity, some good heat mapping session recordings, can look at that, again, as a free thing that sort of you can add in to start getting some data about how people are navigating where things seem to get stuck. And again, if you're really small, it's just you, you know, sort of get eyeballs on a screen to start served with some of those patterns meaningful, you may not have a meaningful, large dataset to work off of, and an AI is going to struggle with a small sample size. If this is again, where you see a lot of those fictions being created as AI is like, Oh, I'm just gonna invent like Aaron has shopped at your store 65 times. But that never happened. Aaron is, you know, not even in my order, my order list here. And so starting to sort of, you know, actually use some human interaction to start to identify some of the patterns you're looking for. That can be helpful. And then again, I'd start to look at some ways of creating sort of person to business or person to person peer to peer type engagements. They can be really great sources for initial data gathering for you.

Aaron Conant  43:31  

Yeah, and that engagement side, like are there recommendations you have for building those engagements?

Jordan Brannon  43:37  

Yeah, so I think a manned chat tool can be really helpful, or an AI chat tool can be really helpful. And again, this is a pretty low hanging fruit for people. I'm still surprised that I'm not seeing, you know, more emphasis on sort of that interactive chat experience for customer support or for sales support, or even in marketing. Actually, some really cool experiential marketing I'm seeing is sort of rooted in conversations with a sort of AI tools. And so, again, having that chat tool, either run by a person when you're on during the hours or buy a some sort of AI solution, chat bot, that's asking some specific questions, trying to gather some specific information as part of sort of the customer service or sales experience were really helpful. There's, again, really good AI assisted chat bots now available that you can provide us to provide support at all hours there will record those interactions. Sometimes there's some plan segmenting based on you know, how many you want to record. And as you're actually getting that data come in, start to use it right. So I think the bygone era chat bots, we sort of let live and die on their own, but start pulling out some of that data, you know, and looking at using AI to help do that. You know, how many of these these inquiries are coming from a particular part of the country? How many are coming from a specific languages, how many seem to be coming from a specific demographic? You know, what are sort of the typical issues or concerns or things, they're asking about the product? Where are their hangups happening, that becomes part of that data set that's really going to be helpful for you in terms of, again, getting an AI solution to really begin to sort of advance what you're doing from a marketing standpoint.

Aaron Conant  45:22  

So awesome. A couple questions that come in any particular chat bots or support tools you'd recommend with AI features? And the other is comes in over email, What tool does he recommend for data visualization? Sorry, I had to shorten it up a little bit. Yeah.

Jordan Brannon  45:41  

Yeah. So I think I think on the the chat side of things I probably look at. Let's see, we are working with a tidy oh there's a company called Ometrics, which now is, has something called Ochatbot. Again, these are eCommerce focused, I'll kind of make that caveat. And then another one that's less eCommerce focused a little bit more universal, which we've been working with more recently is Infobip has been a really kind of a neat one that that's a little bit more conversational interaction focused, maybe kind of related in unrelated the same time, I also make a quick call out for, again, kind of that data gathering idea. KnoCommerce, it's a post purchase, sort of survey type of tool is a free layer they're really recommending for for any mark, sort of transactional eCom merchants to have some post purchase surveys in place just because you can gather some great information about attribution and what motivated people to purchase. And again, can be really helpful in terms of informing some of your marketing. If you are using like a helpdesk type tool, or CRM, already, a lot of them may already have their own chat solution. So again, going back to that earlier point, like pick the ones that you're already working with, because they are things that again, they it's a low hanging fruit, you don't have to get approval and budget and all these extra things, you're probably already paying for it. And so that can be really, really helpful for you. What was it the other question?

Aaron Conant  47:16  

The other one was around data visualization.

Jordan Brannon  47:19  

Okay, and tools for that? Yeah, I mean, we're using things like a Tableau is I think it's probably the one of the most common ones we're looking at, you know, Microsoft's Power BI is another one that's out there. You know, Google has some some great solutions. So there's a few different things that you can use it again, there's some very low hanging fruit that you can get just through something like Google Sheets, if you're not super complicated, or Google Data Studio. If you want to get more complicated, we can certainly talk about some implementations along those lines.

Aaron Conant  47:47  

Awesome. So something that I want to just bring up, because I know organizationally, when you're middle management on a midsize company and not even small companies, depending on how many layers there are. There are multiple limitations that are not necessarily built around, hey, I don't have the team or expert expertise, but around Hey, I can't get these implemented. I can't get these integrated. You know, any thoughts around that mean?

Jordan Brannon  48:17  

Yeah. So how to help people get into AI for the battle start.

Aaron Conant  48:22  

A simple therapy session? Yeah. Yeah.

Jordan Brannon  48:26  

Okay. Yeah, I think again, so if you're running into that sort of organ organizational approval thing. Part of the buzz of AI is that a lot of sort of our central software solutions we use as companies are already sort of moving into that AI space for you. And so if you can't get new approval, and if you can't sort of get new legal sign off, IT budgets, whatever it is, again, go back to sort of what's present. And recognize, you know, sort of maybe how some of those are already solving and look really at yours, sort of what your central software solution is, and explore that one with probably the greatest focus, maybe kind of one of the real valuable spaces is that we're seeing a lot of brands, not brands with software sort of jump into with AI is sort of they're rolling out this CDP or customer data platform feature set. If you've been watching anything in the eCommerce news space, it seems like everybody's launching a CDP. Every SaaS has one and varying degrees of success. But it really is actually a nice thing, especially as you're sort of looking at how do I use AI for content marketing. So these other things, having this customer data platform that's already tied in to some of your existing solutions is really helpful. Probably one of the most common ones is Klaviyo is launched sort of their CDP solutions have you know, helpful for their, you know, IPO probably. I mentioned info bet earlier. Twilio, if you're familiar with them also has a CDP See that's that's come out. A fueled I mentioned earlier again. So everybody I've talked about has a CDP that's out there. And I'm a little surprised that eComm platforms like Shopify and Big Commerce haven't sort of jumped into that yet.

Aaron Conant  50:13  

But that's probably going to come. Yeah, I mean, the CDP space, I mean, it doesn't help them that they were launching heavy like three years ago, touting all these solutions they didn't really have. And so people got burned out on the CDP. Because there's a hop, I think, at one time, there's like, 250 of them out there. Yeah. But I think they're trending up, right. I think there's a lot of them that haven't, you know, persevered. But to that extent, how do you recommend people go about, you know, selecting a CDP like the quick criteria, and that's probably another full webinar, we got it yes, on CDP.

Jordan Brannon  50:50  

I would, again, go back to that idea. Like, you know, if you're in a marketing function, if you're in a sales function, or customer service function, operations function, you know, which one is sort of closest aligned to what you're doing, you know, and then sort of, again, is likely something there, you know, clearly I mentioned is, if you're already a Klaviyo customer, you're already using their solutions for email and SMS, it's already tied to your Shopify or Big Commerce store, could be a really easy one. Infobip like their focus is more on conversational and interactive sort of AI deployments. And they cover a huge range of stuff, in terms of what they integrate to and the features they already have. So if you're more in sort of that interactive, conversational side of things, that's more what you're looking for, again, could be a great CDP. Again, fueled, if you're looking at more of that my emphasis is on the data sources, I want to be able to pull together all of this data from all over the place and put it in one spot field can be a great one. There's a you know, a CD CDP solution there. And again, it's the goal of the CDP is to sort of help track and aggregate data on customer interactions with your brands, so that it can be useful. And again, if you can find the CDP that's closely aligned to what you're already using customer data for how you already interact with customers, it's probably the greatest payoff for you the quickest way.

Aaron Conant  52:06  

Awesome. And as we kind of get to time here, yeah, any like circling back to, you know, AI? And where you see it going? What should people be thinking about as we kind of wrap it up? And I know, I probably have another two hours of content, but in the interest of time.

Jordan Brannon  52:21  

Yeah, I would look at again, if I really hammer on that point of like, what is it that's driving some success for you today, and which AI tools are closest aligned to that. You don't need to have an AI solution for everything. Again, we were working with some Shopify brands, especially where there's like 25 different apps, that all have an AI thing. And half of them aren't useful. You know, just because there's just too much clutter, you just can't get to it in the day. And so I would really stress like, look for the spaces where you see big value, where are you getting most of the revenue right now? Is that SEO? Is that paid? Is that email is that SMS? Is it some branded experience, word of mouth, user generated content? You know, what are the things that are really sort of the engines for you? And then start to think about which AI tools layer in on top of that. And you'll usually find some, some some some great solutions there. And I have, I can provide a lot of other you know, one off examples if you have a specific use case you want to chat about. I'll be available, we'll probably have a follow up email that will go out and can share some of those specific ones for your use case and situation.

Aaron Conant  53:27  

Yeah, for sure. And I just actually dropped the link to the podcast homepage. So in there, I mean, you mentioned Shopify Summer Editions, we did record a podcast on that earlier, just people want to check that out or hear more. But yeah, I'd encourage everybody, you want to further the conversation in this space, Jordan, the team at Coalition Technologies just come highly recommended from across the network for years now, just digital leaders across the board. So I encourage a follow up conversation there. I'd also love to have a conversation with you. I see people on the line today that haven't connected with a while and a bunch of new people. So we'd love to you know, just get together 30 minutes on the phone and just networking knowledge share. So look for a follow up email for me. And with that, Jordan, I think we're gonna wrap up this webinar here. So everybody, we're gonna give you three minutes back in your day, so you're not late for your next meeting. But everybody take care stay safe and look forward to having you at a future event. Thanks again, Jordan.

Jordan Brannon  54:22  

Alrighty.

Aaron Conant  54:23  

See you everybody