Does Research Make the Best Marketing?

Matt Dornfeld (00:02) I have AI in my title, but I don't rush to include AI in anything, right? I think it's an accelerant, it's a catalyst to an outcome. And sometimes if you don't have the foundation set up in the best possible way, like having your data in the right place or structured the right way, you can accelerate to a negative outcome, which obviously we all want to avoid. Ben Ard (00:46) Welcome back to another episode of Content Amplified. Today I'm joined by Matt. Matt, welcome to the show. Matt Dornfeld (00:51) Thanks for having me, I appreciate it. Ben Ard (00:53) Yeah, Matt, I'm excited. You are someone who's at the cutting edge of what we're going to talk about today. I think our listeners are going to absolutely love what's going on. But before we dive into that, tell us about yourself. Tell us about your work history background. Let us get to know you a little bit. Matt Dornfeld (01:08) For sure. Yeah. So I work in the e-commerce software space. been in it for a number of years now, but I've spent the last five at Commerce, which was most recently called Big Commerce. Now Commerce is our parent co. And I currently lead our global AI go-to-market org. So really anything you can think of related to leveraging AI in systems and data to make our revenue engine run faster. That's my world. So I'm spending a lot of time with internal teams and the sales org to figure out where do they spend a lot of their time, how do we make that go faster, and how do we make it more personalized. Ben Ard (01:42) I love it. So you are someone who is getting this firsthand experience of here's a problem. Can I use AI to solve that problem? Can I get it to make our lives more effective, efficient and better? And honestly for quite a massive large organization. also scale plays a role in all that kind of stuff. So we're, we're going to talk about the AI and the LLMs, but we're going to talk about, and here's the title and it's the longest title. And I apologize, but this is kind of what we're wrapping this episode around. So everyone has context. Matt Dornfeld (02:01) for sure. Ben Ard (02:11) how to personalize the webinar experience from registration to recap using workflow automation and LLMs. Okay, so when we say that title, we're talking about personalization in the entire webinar workflow basically. So where do people typically miss the bucket when it comes to this? Are they just not personalizing? Why is this important? What are your thoughts here in general? Matt Dornfeld (02:33) Yeah, I mean, I think even going beyond just webinars and speaking for myself, a lot of my time is spent on figuring out what customer or prospect data do we have access to and what can we use as signals that help encourage us to take a certain action at a certain time. So at its most general level, right, it's just what do these people want and how do they show us what they want through certain behaviors and then what do we do with that information? And so with a webinar, You know, I've been on who knows how many in my career and attended them, but they're all generally the same. Here's the landing page. Enter your info here. Some auto-fill, some don't. Sometimes you get a registration confirmation, sometimes you don't. I put my own placeholder on my calendar sometimes when that doesn't come through and then you attend it and then you multitask your way through attendance and you're half listening and then you get an email later with the recording and that's it. And then some people wonder why didn't conversion happen? And I get it. It's like you have to do these fast and at scale for lot of especially small organizations. And so there's just a lot of work and areas for opportunity where things could go wrong, where things could be optimized. So for me, what it means is customers or prospects or whoever's showing up to your front door to register are coming to you for a reason. And first thing you want to do is figure out hopefully why they're coming to you, and then based off of what you can find out about them and what they care about, then start to build a story before, during, and after that experience that engages them early, gives them more reasons to attend and participate, and then off the back of that session, finding ways to enable your teams to connect in a way that leads to whatever outcome everyone's looking for. Ben Ard (04:12) I love that. Okay. So let's stick with a webinar example. And obviously I love that this is applicable to other channels and things like that. Let's start with the before. So we'll break down the before the during and the after before they ever show up to the webinar. How can we make it more personal? How can we use LLMs? How do we improve the workflow? How do we really just separate ourselves from the, this is a typical landing page. fill out your info and pray that you get that confirmation email or else you will not show up. Like, how do we do that? Matt Dornfeld (04:40) Yeah, for sure. I mean, I think in my world, I have AI in my title, but I don't rush to include AI in anything, right? I think it's an accelerant, it's a catalyst to an outcome. And sometimes if you don't have the foundation set up in the best possible way, like having your data in the right place or structured the right way, you can accelerate to a negative outcome, which obviously we all want to avoid. So when you think about setting up the form, One thing to consider is making that form dynamic based on simple logic-based rules. for example, if your ⁓ visitor is a B2B or B2C retailer, if they can delineate that they're B2B or B2C or both, the next question about data they give you might change. And so what that does is immediately helps personalize for you in the back end who is engaging, what might they care about. And based on the following questions, those are more signals or data inputs that you can then use further down the line as part of this process. And so if I complete a form and I say I'm a B2B manufacturer, I care about these things, here's my name and my email, you may go so far as to say, just tell us your business type and email address. And off the back of that, behind the scenes, you're collecting all of that information. And we talked offline earlier, there are number of ways to enrich data. There's so many tools in the world that help you find out more contact information. So with any database you have, what you could do is collect that information and then enrich it aggressively with all relevant things across social channels and personal, professional contact information, whatever is available. And then using that information, you begin to build out an archetype of what this person is and maybe what they want to learn about. So, know, stage one is figuring out how do I capture, now that I have this person's attention, how do I capture the right information based on exactly who they are versus forcing all visitors down one swim lane, which likely overindexes or underindexes on certain kinds of information that would otherwise be helpful. Ben Ard (06:42) Okay, I love that. So the first step, collecting dynamically the right information, enhancing and augmenting that information. So I really understand who this person is and what they're looking to solve. What do I do with that? After that? Do I still just send them to the normal webinar and hope that they get the information they need? Or what does that look like? Matt Dornfeld (06:54) Exactly right. Great question. now, the now what, right? Now they've signaled interest. They want to learn something here, whether they attend or get the follow up. They care about what you have to say on that day at that time. Between the moment of registration and delivery of the presentation. There's lag. Sometimes it's two weeks even. That's a long time. I forget things that I signed up for all the time in a two-week window. And so what would be helpful to me as someone who wants to get this content is pre-read materials. Things that are relevant to my business, to me personally even, that you now know about me based on my inputs at registration. So for example, if I'm that B2B retailer, maybe you're within a week of the webinar sending me a couple of case studies relevant to something I might learn about. and relevant to my category. Because I'm in the manufacturing space, give me something related to manufacturing, maybe not something about home goods and how they might appear in my house. And so by collecting the right logic-driven information early, I now know what to say between moment of registration and webinar to help reinforce attendance, remind them about the brand and the value that I can offer without actually directly selling. I'm just providing value. Here's relevant things to you about the business you told me about already. Ben Ard (08:16) I love that. Okay, that's super cool. now I know who they are. I've provided unique personal value and then the day of the webinar shows up. What can we use that information for during the actual webinar and then after that in the in the follow up process? Matt Dornfeld (08:31) For sure. Here's where I think things can get a little advanced. You may be able to do this in-house. You may want support. depends on how you work. Ben Ard (08:38) I'm all about this, I'm super curious. Anytime someone says that I'm like, all right, I'm all in, perking up my ears, I need to hear this. Matt Dornfeld (08:44) So I think what I haven't seen a lot of webinars do is break things out into pods or separate rooms. And so you might have an intro section, like a lecture. You have the intro, then you have several subsections, and then a conclusion. Especially if you want to drive more active participation, you could split the rooms out into very specific pods related to the overarching topic. So based on the input information we receive, we can now say, well, we have a B2B client and a B2C client or prospect who are both attending. They may want to talk to only people who care about their relevant area of business. When we host summits ⁓ at commerce, we have advisory boards that are all agency and all tech, right? Because they want to learn from people in their ecosystem and share so they can excel. And so, Doing this in a webinar format also allows you to have a conversation actively about the topic with people who already know care about this one thing. And so that's a huge and easy thing to do, which is just use the information you have, split the rooms ahead of time. In addition, now we're getting to a place where LLMs are powering agentic conversations. And it's not very challenging to create chatbots that are based in the facts about your business and other useful bits of information. And so what you could also do is attach a agentic teammate to the chat, whether that's offline, like follow up after, or even during the conversation, to real time answer questions as they appear in the chat. So imagine someone says, ⁓ I'd love to learn more about this thing in the chat. you could directly send that information within the chat experience or using workflow automation, actually pull down those comments, push them into a pre-drafted email, and then power them up in the drafts folder of whichever team member should be sending that follow-up email. So all of these things about, I commented in the chat, I've engaged, I've clicked a button, I went to a breakout room. These should sound like. more signals and behaviors, which means we have learned more things, which now signal not just fit, which we learned about early about what kind of business they are, but now we're learning more about intent. So they intended to attend, they did attend. Now they're asking for more information. What team would care most about this information? The sales org. The sales org wants to know, this company has signed up. They clicked on the case study. They attended the webinar. They've the session. They've asked a question in the comments. I mean, how many more flags can we get saying, please help me, right? Engage here. And so, know, in using LLMs offline, you could then analyze the transcript, the comments, the engagements, and build a narrative based off of all the things they told you at registration to now write a really custom and personalized follow-up email that would come after the show, after the webinar, with relevant content. and other tools or inputs that that customer or prospect otherwise wants to hear about. Ben Ard (11:40) Okay, so I love this. We just broke it down. This can completely change the workflow when it comes to all of this. So let's get tactical. So neither of us work for companies that do webinar software or any of that kind of stuff. What tools are you personally using? If you're allowed to share, if not, that's totally fine. What tools are you using that is helping you do this workflow? so the audience can kind of poke their heads into it and see if it might work for them. Matt Dornfeld (12:07) For sure. I think one of the easiest things to pick up on is form software. So type form, even Google Forms, as long as you can input the logic components that we were talking about. Those are just simple if-then statements that attach to the questions. Workflow automation software would be something I would consider investing in. We use Make Automation. We're a big fan of them. But there are many others, right? There's Zapier, there's N8n, there's Relay. ⁓ So I think there's gum ball or gum rollers, something out there, gum road maybe is the other one. ⁓ But there's no shortage of workflow automation tools in the market that can help power really unique and custom workflows. So like the whole concept of if this thing happens over here, trigger these three waterfall steps to go send these emails or draft this and all that's doable with workflow automation. I'd want an enterprise Ben Ard (12:38) Yeah, something like that, yeah. Matt Dornfeld (13:00) license to chat GBT or your preferred LLM. I think there are a few reasons for that. One is enterprise license means that your data won't be shared publicly. And so especially for larger organizations, you simply can't afford to have your data being chewed up and training the chat GBT enterprise models at large. So having that walled garden is very important, especially for bigger businesses. And then, you know, off the back of that. To do some of the chat experiences, you're getting far more technical. And so there are softwares that power agentic conversations. It could be less advanced than that, where you're simply powering up, you're linking a chat or a custom GPT via API to a publicly facing instance. So you could even publish your own GPT and provide access to that for those that attend the webinar and just put an access code on the GPT itself. There's so many ways to do this and to find ways to engage with your prospects or your customers using fairly low cost solutions, which I think is the beauty of a lot of this, is that of all the tools I've shared here, there aren't really many that cost over 50 bucks a month. So, the stack is very affordable and will just scale based on the size of your organization. Ben Ard (14:07) Yeah. I love it. Well, Matt, this has been extremely helpful, extremely useful, tactical and strategic all at the same time. Like I promised the episodes go by quick. have run out of time today, but for anyone listening who wants to reach out and connect with you online, how and where can they find you? Matt Dornfeld (14:29) Yeah, LinkedIn. It's my favorite channel. feel free. I think my username there is Matt GPT so kind of on the nose. But I love to engage with people there, share thoughts, comments and content, and then emails get to. So just matt.dormfield at commerce.com. Ben Ard (14:46) Perfect. Anyone listening to this episode, just scroll down to the show notes. We will link directly to Matt's LinkedIn profile so you can connect with them there. Matt, thank you so much for the insights today. This was awesome. Matt Dornfeld (14:56) Pleasure, thanks for having me.