Shawn Miller, Inboundr AI: Employee-Led Content With AI

Watch on YouTube · Listen on Spotify or Apple Podcasts

Shawn Miller, co-founder and CEO of Inboundr AI, says buyers now trust people more than brands, so the companies winning on LinkedIn are the ones getting their founders and employees to post. He tells host Liam Lawson how Inboundr, a Slack app, keeps AI-written posts authentic by drafting them only from what you have already said in meetings, podcasts and conversations, and how five years of door-to-door sales and a home-cooked meal app that ran into regulators led him there.

Key takeaways

• Door-to-door sales made starting a company less daunting. Shawn spent five years selling home services door to door, the last three running his own franchise, and says getting over the fear of rejection pushed him toward entrepreneurship. (01:18)
• Buyers now research on social media, not with sales reps. Shawn describes a trust recession: buyers can get about 90% of the way to a decision through social media, reviews and social proof without talking to a sales rep. (14:40)
• Build trust first, then mention the product. The people doing well on LinkedIn post about their own experience, Shawn says, and only once they have earned trust do they bring in the product their company sells. (16:52)
• Inboundr writes only from what you have said. The Slack app builds a library of your meetings, podcasts and conversations, learns your style from past LinkedIn posts, and aims to take a post about 90% of the way there. (18:35)
• 250 beta sign-ups, mostly from outbound. At the time of recording, 250 people had signed up for the beta, around 60 to 70% from outbound outreach based on signals of interest in creating content. (21:52)
• Senior marketers and compliance teams push back most. Senior marketers who love the craft resist AI help, and technical teams worry about recording conversations, which is why Inboundr is opt in and pursuing SOC 2. (26:48)
• Solve a problem you have yourself. Shawn's advice for AI founders is to build quickly around a real problem you have, not to start with AI and hunt for a problem. (45:52)

How did door-to-door sales shape Shawn Miller as a founder?

Shawn Miller says five years of door-to-door sales taught him to handle rejection and build perseverance, and made starting a company seem less daunting. He sold window cleaning, painting, gutter cleaning and other home services from late high school through university, first for two years under someone else's franchise, then for three years running his own. Those summers meant hundreds of customer homes, hundreds of thousands of dollars in revenue and thousands of doors knocked, and taught him what running a business involves.

Liam Lawson shares a similar start: five years of mobile phone sales from the age of 16, which he calls pivotal in spurring him toward entrepreneurship.

What happened to Shawn Miller's home-cooked meal delivery app?

Shawn Miller's first startup with his Inboundr co-founder was an app for ordering home-cooked meals, launched about a week after the two met. His co-founder built the app quickly, they signed up home chefs and users, and a 15-minute live demo on a cooking show in Ottawa brought a spike in sign-ups.

Within days of launching, a government health and kitchen inspector threatened them with a cease and desist for not following kitchen laws. Shawn says in hindsight it was a huge mistake, because home cooks do not follow restaurant kitchen regulations. Knowing that first $20,000 threat would only build, they shut the business down rather than fight an uphill battle.

Where did the idea for Inboundr AI come from?

Inboundr AI came from Shawn Miller's own work as a LinkedIn ghostwriter. For 10 months, alongside a sales role at a crypto company, he wrote LinkedIn content for its CEO and its head of law enforcement affairs. Each post meant interview calls, questions to draw out their views, and then shaping their experience and opinions into a well-formatted post with a strong hook. The content became a huge lead generation source, and he felt firsthand how much work it took.

He realized much of the process could be automated without losing originality. The hurdle was the generic tone people associate with AI writing, so the first goal was to remove it. Inboundr's answer is to pull from your meeting transcripts, podcast transcripts, Slack conversations and emails, where you already share your thinking out loud, and format that into posts. Shawn describes it as a ghostwriter, a researcher and an intuitive editor in one.

Why is employee-led content replacing traditional outbound?

Shawn Miller says the predictable revenue playbook of raising VC money, hiring armies of SDRs and making 100 calls a day no longer works the way it did. Inboxes are overflowing, cold outreach success rates have fallen, and SEO is less effective as Google adds more AI to search. He says the companies staying afloat are the ones putting a huge emphasis on founder brand and employee-led marketing.

His examples include Adam Robinson, who he says spends around 75% of his day creating content as CEO of a $2 million ARR SaaS company, and a sales agency that, he says, got eight team members posting and saw four million impressions in 90 days and $80,000 in added MRR. Inboundr's bet is that within five years every company will train its whole workforce, from intern to CEO, to build a personal audience online.

Why do buyers trust people more than brands?

Shawn Miller calls it a trust recession. Sales reps used to be seen as partners and are now seen as sellers, so buyers do their own research. He cites stats that 75% of people turn to social media before a buying decision, and says buyers can get about 90% of the way there through social platforms, reviews and social proof. He says he has bought the tools in his own sales stack after seeing LinkedIn posts, without ever meeting their sales reps.

Blog posts suffer because readers do not know who wrote them. The people who do well on LinkedIn, Shawn says, post about their own experience rather than their employer's product. Once readers trust them as experts, a mention of the product they sell carries far more weight.

How does Inboundr AI work?

Inboundr AI is a Slack app, built that way so users do not have to log into yet another platform, says Shawn Miller. You choose which Slack channels, meetings and podcasts it can record, either by pasting in transcripts or by letting it record automatically. It builds a library of your conversations and learns your writing style from your past LinkedIn posts.

It works proactively and reactively. If three podcasts in a week share a theme, it suggests a post on that theme. You can also ask it to pull a specific snippet from a conversation, or, after 90 days of conversations about prospecting, to write your top 10 prospecting tips. It drafts the hook, body and call to action, and Shawn recommends editing the result. The aim is to get about 90% of the way there without losing authenticity, because it is not generating ideas from scratch. That also means it works best for people with lots of meetings, podcasts, Slack and email.

How is Inboundr finding its first users?

At the time of recording, Shawn Miller said 250 people had signed up for Inboundr's beta, with a three-week test period planned before collecting feedback. Around 60 to 70% came from outbound and 30 or 40% from inbound. Every feature so far, he says, has been built from customer feedback.

The outbound approach targets signals of interest, not job titles. A post offering top 20 prompts for writing LinkedIn posts can draw 200 comments from people asking for them, and those people are clearly thinking about content, so Inboundr reaches out offering a tool to do it more efficiently. Inbound comes mainly from Shawn and his co-founder posting content almost daily, plus some referrals.

What pushback does an AI content tool face?

Shawn Miller says the strongest reactions come from senior marketers: they are either very on board or very against it. People who love every part of creating content resist AI immediately. Younger marketers are much more open to the help.

The second source of pushback is technical teams worried about compliance, since recording conversations raises privacy and security red flags. Inboundr is working toward certifications such as SOC 2 and makes everything opt in, so users control what it records.

What sets Inboundr apart from other AI writing tools?

Shawn Miller says 99% of AI content tools miss context: they know nothing about the person they are writing for, so the output is generic. By querying a library of your past conversations, Inboundr can learn your point of view on different topics, and by reading your past posts it writes in your tone and style.

The biggest misconception he has to overcome is that AI content means the recognizable ChatGPT style people can spot from a mile away. On LinkedIn, he says, the content that works is raw and authentic, not a buttoned-up blog post. On trust, he compares Inboundr to AI meeting note-takers, which faced the same early adoption hurdle over recording and CRM access, and now there are a million of them.

Why did Inboundr start with LinkedIn?

Shawn Miller says Inboundr focuses on LinkedIn because he understands the problem deeply from ghostwriting, and because he uses the product himself, so he can judge every draft it writes. His previous AI startup helped people buy the right clothing size online and helped clothing brands reduce returns, but he had never run a clothing brand and had to rely entirely on user feedback.

That taught him to look for founder-market fit. Image generation to go with posts, suggested by early users, is being explored, and some users want Instagram and TikTok support, but he wants to nail a problem before building a solution in search of one.

What advice does Shawn Miller give people starting an AI business?

Shawn Miller would tell them to start: build quickly and solve an actual problem you have yourself, rather than starting with AI and hunting for a problem. He says starting an AI company is fairly affordable right now. His own worry is that superintelligence, which some predict by 2027 or 2028, could disrupt many current AI use cases, including his own.

He says he is driven by competitiveness and the thrill of taking something from zero to one. His first startup received $12,000 in seed funding from Western University, which he credits with strong support for student entrepreneurs.

In his words

"It made starting a company seem less daunting" (02:06)

"We're just in a huge trust recession now" (15:23)

"Personal brand and all that stuff is definitely here to stay" (11:33)

"It's not generating ideas from scratch it's pulling all the words and the thoughts that you already have" (21:01)

"Don't use AI and then go hunt for a problem" (46:03)

Chapters

00:00 Introducing Shawn Miller and Inboundr AI
00:52 Five years of door-to-door sales
04:01 The home-cooked meal app and the regulators
06:25 From LinkedIn ghostwriting to Inboundr
09:46 Why employee-led content is winning
12:04 Companies doing founder-led content well
14:12 The trust recession and how buyers buy now
16:52 Building trust before selling on LinkedIn
18:25 How Inboundr works
21:09 Beta growth and customer feedback
23:28 Finding users through outbound signals
26:00 Inboundr's five-year vision
26:34 Pushback from marketers and compliance teams
29:59 The technology behind personal style
31:47 Co-founder dynamics
33:52 How Shawn learned about AI
35:29 The biggest misconception about AI content
37:16 Expanding into photo and video
38:30 Why LinkedIn, and founder-market fit
40:28 What sets Inboundr apart
41:36 Getting people to trust AI with their data
43:05 Superintelligence and disruption
45:35 Advice for starting an AI business
46:51 What drives Shawn, and Western University
48:50 Beating namesakes on Google and wrap-up

About the guest

Shawn Miller is the co-founder and CEO of Inboundr AI, a Slack app that turns meeting, podcast and conversation transcripts into LinkedIn content. He spent five years in door-to-door sales, three of them running his own franchise, worked in sales and as a LinkedIn ghostwriter at a crypto company and previously founded an AI startup that helped people buy the right clothing size online. He studied at Western University. Shawn on LinkedIn

About the host

Liam Lawson is CEO of The AI Report and host of The AI Report Podcast. Meet the team

Full transcript

Edited for readability. Quotes above are word for word from the recording.

Liam Lawson (00:12): All righty, today we are back with another episode of AI Tool Report Live. I'm very excited for this one, because today we're talking to Shawn Miller, who is the co-founder and CEO of Inboundr AI, which builds AI agents that turn transcripts into LinkedIn content. I might have got that right, I might have got that wrong, but we'll find out today. Welcome to the show, Shawn.

Shawn Miller (00:36): Thank you very much for having me. You nailed it, that's exactly what we're doing.

Liam Lawson (00:43): There we are. That might be the first time in the history of the podcast I've nailed the intro, so thank you for that. I really want to get into everything you're building and doing right now, but before that I'd like to get some context on who you are. I was looking across your LinkedIn, and the oldest thing I can see is that you were a door-to-door sales rep for College Pro. Can you tell me a bit about that work experience, and the skill sets or mindsets it formed in who you are and how you conduct yourself today?

Shawn Miller (01:18): You definitely did your digging. I was a door-to-door sales rep for five years, selling window cleaning services, home painting services, gutter cleaning, siding cleaning, lawn cutting, really any sort of home service you can name. It was great experience for learning how to handle rejection and hearing no, and building up that perseverance. It taught me the foundation of the skill set required to be good at sales.

That experience pushed me onto an entrepreneurial track at a younger age. It was a huge step outside of my comfort zone when I first got started in late high school, and I did it all throughout university. It made starting a company seem less daunting, once you can get over that initial fear of rejection and putting yourself out there. It was honestly a really pivotal experience for me.

Liam Lawson (02:19): There are a lot of parallels with myself. My first real job, when I was 16, was five years of mobile phone sales. We had a shop called Carphone Warehouse in the UK, and my job was sales as well, upselling people to take insurance and all these other things. For me that was extremely pivotal in becoming who I am today and spurring me on to explore entrepreneurship. You said it made you interested in entrepreneurship and gave you the confidence. What was your first entrepreneurial venture after College Pro?

Shawn Miller (02:56): College Pro is actually a franchise. I did two years of door-to-door sales for someone else who owned a franchise, and after that I went and purchased my own franchise, so for the final three years I was running that business myself. I had all of the experience of hiring people, leading cleaning crews and continuing to do door-to-door sales. We did hundreds of customer homes every summer, hundreds of thousands of dollars in revenue, and knocked on thousands and thousands of doors. It was really great experience right out of the gate in terms of everything about running a business. Of course the product is sort of trivial, we're selling window cleaning, but it taught you all of the aspects of what running a business looks like.

Liam Lawson (03:50): I can see from your work history that you've done so many different businesses, and from all of those you've probably learned different skills. I love one of them specifically. It says: Uber Eats for home-cooked meals, full stop, got shot down by government. Tell me about that.

Shawn Miller (04:11): I started this business with one of my co-founders from Inboundr. We actually met a little over a year ago now, and Simply Eats was the first business we wanted to start together. Ironically, we met and about a week later we launched the startup. It was sort of: let's just try and build something quickly, let's see if we can take something to market, let's see if we can get users, let's see if we can build the foundation of some sort of tech startup. [Name unclear] is a brilliant coder. He built that app faster than anyone I've ever seen, and we did pretty well on the business side too, in terms of getting home chefs who were willing to cook meals and list their food on the app, and of course the user side too.

We built up a little bit of hype around it in the city, and we were able to go on a live cooking show in Ottawa to advertise what we were doing. It was all of 15 minutes. We demoed what the app was going to be, live, and as I'm driving home that day we're seeing a spike in signups. Everyone's excited, so we launch it and start facilitating a few orders. Of course we were the delivery drivers early on.

Within a couple of days we get a call from the government, someone who works in health and kitchen inspections, who threatened us with a $20,000 cease and desist for not abiding by kitchen laws. In hindsight it was a huge mistake. People who are cooking food out of their home aren't abiding by any sort of kitchen regulations like you would in the restaurant industry, so there was a huge issue with health and safety and all of that. There are other companies doing this in the States, and we heard they were fighting these legal hurdles too, so we were expecting it to some degree. But given where we all were in our careers, we knew that first $20,000 threat would only build. We figured this was going to be a huge uphill battle, so we might as well shut it down and do something else.

Liam Lawson (06:25): I can imagine the amount of pressure, especially from a government institution, is enough to make you rethink some of those things. And with that you obviously pivoted. Was the first thing you pivoted towards Inboundr? How did you first hear about AI, and why was this where you went?

Shawn Miller (06:46): We didn't really start Inboundr in the sense of: let's use AI for some use case we can find. Before starting Inboundr, and during my time at Simply Eats, which was kind of a side hustle, I was also working for a company in the crypto space. I was on their sales team, but I was also working for their CEO and their head of law enforcement affairs as a ghostwriter for LinkedIn content. For 10 months I was writing their LinkedIn content, and the process took a while. I had to set up all of these interview calls with them and prompt them with a set of questions to understand their perspective on the industry and the way they view things. Then you have to take all of their experience, insight and opinions and sculpt it into a piece of content that's going to work well, is well formatted, has that great hook, all of that.

It became a huge lead gen source. We're starting to see this across the industry, where this founder-led content, employee-led content, or employee-led marketing narrative is becoming more and more effective, and I could go into a lot of reasons why. We felt that too, we saw that too, and I had also felt the problem firsthand in what's required to create great LinkedIn content and thought leadership content. I realized there are a lot of processes here that could probably be automated without losing that originality or authenticity.

Having met my co-founder at Simply Eats, we started to talk about it a little, and went back and forth on an initial concept of how an AI tool could help this process. The big hurdle is that when you hear AI-generated content, your mind immediately goes to "harnessing the power of AI in this digital economy" and those types of classic phrases. So in building an AI tool for content, the first focus was: let's get rid of any of that generic crap and find a way to keep it authentic.

That's why our approach has been: let's take your meeting transcripts, your podcast transcripts, your Slack conversations and your emails, where you're saying all of this interesting stuff day to day, all of this thought leadership, essentially verbally. Let's pull what you say, build an AI that's good at formatting content, and then it can take your insight and produce a piece of content. We like to think of it as a ghostwriter, a researcher and an intuitive editor all in one.

Liam Lawson (09:25): I really want to learn more about the technology, your growth so far, what it's like co-founding a company and how you structure that. But something you mentioned that you could talk about forever is why companies are now shifting towards employee-led content, and why that's succeeding. Can you delve into that for me, please?

Shawn Miller (09:46): Some of this experience is firsthand. Of course, I'm earlier in my career. But if you follow a lot of these go-to-market leaders and do a lot of research on the industry, which is really what we've been doing every day since starting this, you can see a huge trend from the early 2000s up until where we are now. That's when the predictable revenue model was coined. It was: go raise a ton of VC money, hire an army of SDRs, have them build a list, blast out messages, make 100 calls a day. If you want to scale that up, add 10 more SDRs, increase their quota and keep going.

For a while that worked. Even when I started as a sales rep in tech five years ago, I could get away with limited personalization on the emails I was sending out. Today that's not going to get you anywhere. All of our inboxes are overflowing. By the time this podcast is over, we're all going to have 20 more cold calls in our inboxes. There's a ton of research now that the success rate of traditional cold calling isn't what it used to be. Even flipping over to the marketing side, things like SEO aren't as effective as they once were, because Google is implementing more AI by the second and it's crushing search results.

The reality is that the way you would drive customers to your website, product or service 10 years ago, or even five years ago, isn't what it is today. It seems that today the companies putting a huge emphasis on founder brand and employee-led marketing are the ones staying afloat in the economy right now. So our hypothesis, I guess you could say, is that personal brand and all that stuff is definitely here to stay, but five years from now we believe every company will train their entire employee base, from the intern up to the CEO, on building a personal audience online. We think it'll become a crucial strategy of the emerging go-to-market playbook. That's the vision we're building in line with.

Liam Lawson (12:04): With that being said, what are some examples of brands, companies or individuals doing this to the nth degree? Who are the best performers in founder-led brand?

Shawn Miller (12:16): Of course the first one that comes to mind is Adam Robinson at RB2B. He's got a great story and a great case study. His first 12 months of creating content for the e-commerce ICP didn't deliver the results. Eventually he pivoted over to speaking more to the B2B SaaS founder, and of course has skyrocketed on LinkedIn ever since. He says he spends around 75% of his day just creating content, and if the CEO of a $2 million ARR SaaS company is doing that, there's something to be said about it. They're doing a great job. RB2B grew by something like 5,000 users in three months, pretty much entirely through Adam's content, and they do have an open playbook too.

Another example is [name unclear]: 33% of their demo requests come from the CEO Todd's LinkedIn content. ColdIQ, which is a sales agency, did four million impressions over 90 days by getting eight members of their team to post content, and that led to $80,000 in added MRR. Clari is another one. In the month of May they had somewhere between 100 and 200 people at the company create content, and they got over 3 million impressions in that 30-day period. I don't know the numbers on the inbound lead and MRR side, but I'm sure it was quite substantial.

So there are a lot of companies who have started pushing this playbook, and they're getting a lot of great results from it. The problem we're really trying to solve is that creating content is hard, and not everyone's great at it. If you really want to do this at scale, there's going to have to be some sort of tool that can help people who aren't so great at creating content to be better. And for those who are great at creating content, this tool is a way to speed up that process for them.

Liam Lawson (14:12): You're obviously so deep in the weeds with everything here, the psychology behind it, understanding why this is. What has been the shift in the consumer market, the psychology behind people connecting so much more with humans? And as a second part to that question, what types of content do you see being more successful in the future? I'm seeing a big trend for video on LinkedIn, for example.

Shawn Miller (14:40): The trend is that the way buyers used to buy years ago is totally different to how they buy today. Even if I look at my own tech stack: we're using Clay, lemlist, RB2B, Apollo, PhantomBuster, Browse AI, and a bit of [name unclear] sometimes. I've never met with any of the sales reps at these companies. I've seen LinkedIn content, I've seen someone post about how they're building out these workflows, and then I've made a buying decision based on that information.

Of course it's not just unique to me. There are stats now pointing to 75% of people turning to social media before they make a buying decision. The reality is that we're just in a huge trust recession now. Sales reps used to be seen as partners, and now they're seen as sellers. Everyone knows we're just being peddled some new solution, and we're not sure if that sales rep believes they can help us or not. It's pushed buyers away from sales reps and more towards doing their own research, and it's never been easier in history for a buyer to do their own research. They can go on all these social media platforms, look at all the reviews and all of the social proof, and get basically 90% of the way to a buying decision without ever having to talk to a sales rep.

So the companies that have pivoted over to this thought leadership side, this employee brand side, are the ones meeting their buyers where they are, which is on social media, and that's why they're the ones having success. That's the first half of the question. Sorry, I missed the second half, about where content's going.

Liam Lawson (16:21): You mentioned that it's a trust recession. People have a lack of trust in institutions and are looking more to individuals with personality. So based on that thesis, the type of content people engage with, connect with and trust should change to, I imagine, more video-based content, because they can see the personality behind it. Do you have any more perspective on what's performing well now, and what might perform even better in the future?

Shawn Miller (16:49): First of all, the problem with a lot of blog content is that you don't know who wrote that blog. Was it the social media manager, the intern, the CEO? We're really not sure, so it's hard to take that advice and trust it. A lot of the people doing well with content on LinkedIn aren't pushing out the brand they work for or the product or service they sell. They're pushing out: here's all my experience from years of doing whatever, maybe sales. Now I'm following this person who's been great at sales for the last 10 years, and oh, they also work for a company that sells some sort of sales tech.

Once I build up trust with this person, and think this is one of the most knowledgeable sellers currently in my network, then any tech stack he suggests I'm going to take with a lot bigger grain of salt. So it's first building up that trust that the person is experienced and knowledgeable, or has at least built their own trust in whatever career they have. That's when you can start to throw in: by the way, this is the company I work for, and this is the sales tech product we're selling. It starts with posting about your experience and proving you know your stuff. Once people recognize that about you, it becomes much easier to throw in the product you sell on the side.

Liam Lawson (18:18): I totally understand. With this entire process, talk to me about how Inboundr works and how it helps. Imagine I'm a new client coming in. Tell me about the product, how it works and what it could do for me.

Shawn Miller (18:35): We've actually built Inboundr as a Slack app. We did this first of all so you're not logging into yet another platform every day. We're already doing enough of that, so it goes right into your Slack channel. As soon as you log into it, you first decide what it can and cannot record. Is it going to record all of your channels, none of them, or one of them? You also pick which meetings it's going to record. You can copy and paste in transcripts, or you can turn it on to auto-record your meetings. Same thing with podcasts: you can copy and paste in podcast transcripts, or you can let it record them too.

Basically, it's building a library of all of your conversations, and the AI queries that when it comes to writing content for you. It's both proactive and reactive. Let's say this week you do three podcasts, and there's a common theme between all three of them, a lot of interesting information all under one topic. Inboundr will proactively suggest: based on your three podcasts, here's a post that would convey your thoughts and what you explained. So it will suggest topics, it will suggest posts, and it will write them fully for you. Or you can prompt it: hey Inboundr, I just had a podcast with Liam, I want you to pull the snippet where I went off about topic X, Y and Z. Then it will draft that post for you.

It's really speeding up the time to create content by querying this library of all of your meetings, conversations and perspectives, all in one spot. After 90 days, let's say I've had a ton of conversations on prospecting tips. I can go to Inboundr and say: write a post for me about my top 10 tips for prospecting. It's going to go through the library of all your conversations, pull out all the times you mentioned prospecting, take the most interesting and insightful parts, and put it into a post for you, with a great hook, a great body section, the CTA, all of that. Of course we always recommend that you edit it or add anything, but its job is to take you about 90% of the way there without losing any authenticity, because it's not generating ideas from scratch. It's pulling all the words and the thoughts that you already have.

Liam Lawson (21:09): I love that it's a Slack app. That's massively helpful for me, because if you're like myself or my co-founder Martin, you've got a million different tabs open at any one time. What that gives me the sense of is that you're detail-oriented, and you have pillars or values or north stars you're following. So what has growth been like for you so far, and what are the pillars you're sticking to as you continue to grow? What are the values or user experience things you always want to stick to as you build this?

Shawn Miller (21:52): On the growth side, as of this morning we just hit 250 people signed up for our beta, which goes live on Monday next week. We'll be onboarding hundreds of people between the 16th of September and the 11th of October, so it's a four-week period, and everyone will have three weeks to test out the beta. Then on November 1st we'll pull it off everyone's workspace, collect all of this great feedback and see where we stand. Do we need to go back to the drawing board? Have we missed something? Or is the product almost there, and we just need a couple more tweaks before we can really take it to market? It's going to be dependent on that feedback.

Going back to what you said about pillars, everything we do is entirely based on customer feedback. Every feature we've built so far, the Slack approach, all of it, has been built for the customer, to make it as easy as possible to use. On the AI side, the one thing we're continually trying to make clear is that unlike ChatGPT, where it produces its own ideas, this AI is not doing that. It's only pulling from a library of the stuff you have said. So it's going to work better for people who have a ton of meetings, do a ton of podcasts, and are really active in Slack and over email. Otherwise it's not going to have that context library to write for you. But the more you have these conversations, the more it's able to pull from, and the more powerful the output gets.

Liam Lawson (23:28): For all of these users running through the beta period, are you reaching out to people on an individual level, or are they just signing up on a landing page? And from the one-on-one conversations you've had, the individual feedback from people who have tried it so far, has anything surprised you or really changed how you approached this?

Shawn Miller (23:53): On the user side, we've probably gotten around 60 to 70% from outbound, reaching out to individuals, and 30 or 40% coming inbound. On the outbound side, the approach we've taken is reaching out to people based on some sort of signal that they're interested in creating content. I'm pretty sure I reached out to you under some sort of signal.

Liam Lawson (24:20): Sure, we'll have to have a look, actually. Let me check.

Shawn Miller (24:26): Yeah, I'm curious.

Liam Lawson (24:30): Let's see. You mentioned a comment that I'd made on someone else's post, and then you mentioned your platform as well, and we got chatting from there.

Shawn Miller (24:47): That approach works really well. We're not picking people based on their title and reaching out to them. We're reaching out to people based on the level of interest they've shown in some sort of content solution. Maybe not a solution, but at least they're interested in content, it's top of mind. So we're looking for posts. A great one that always works really well for us is when someone posts: here are my top 20 prompts for creating a LinkedIn post, if you want these prompts, comment below. Then you'll see 200 people commenting to request the prompts. Clearly creating content is top of mind for them. So we reach out to those people and say: it's obvious you're looking to create content with these prompts, we built a tool that's going to help you do this more efficiently, would you like to try it? They're probably going to try it. That's been a huge success for us.

Then on inbound, I'm the only one, well, actually my co-founder is too as of recently, so there are two of us creating content on an almost daily basis, and that's been the primary driver of our inbound leads. We get a little bit of referral business too.

Liam Lawson (26:00): What is the true north star of Inboundr as a business? What's the end goal here?

Shawn Miller (26:07): The vision we're following is that five years from now, everyone's personal brand is really top of mind, every company is training their employees to build personal brand, and Inboundr is the tool that unlocks that transformation and powers all of it. So that five years from now, when a company thinks, we need to get our employees creating content on LinkedIn, Inboundr is the tool that comes to mind.

Liam Lawson (26:34): With that being said, have you ever received any backlash or negative feedback, anyone from a big company saying, no, I don't [expletive] agree with you? What have those conversations been like?

Shawn Miller (26:48): The best conversations always come out the more senior the marketer you talk to. Either they're very on board with what we're doing, or they're very, very against it. In the content world, some people absolutely love every aspect of creating content. It doesn't matter how long it takes them, it's a process they thoroughly enjoy. So when you bring AI into the equation with a person like that, there's immediate resistance and total pushback on what we're doing.

As I said, it's the more established, more senior marketers who give us this huge pushback. With the more junior marketers it's much rarer. Not junior in the sense that they're brand new to marketing, just that they're younger. They're more open to this type of technology and this type of help. So there's been a ton of pushback on the content marketing side from that perspective: people don't think AI should be helping with this.

Then there's pushback from the technical team on the compliance side. The idea of recording all of your conversations can raise some red flags for privacy and security reasons. That's top of mind for us too, in terms of getting all of our certifications like SOC 2. We're also building the tool so it's really opt in, opt out. You don't download Inboundr and we automatically scrape everything in your library. You can choose, plug and play, what it will and won't record, so the control is still entirely on your side. So those are the two things we're facing: senior marketers sometimes disagree with us on whether AI should be involved in content, and the technical team will be very worried about compliance from time to time.

Liam Lawson (28:34): You've probably learned your lesson with compliance standards at this point.

Shawn Miller (28:40): Yeah, fully on board.

Liam Lawson (28:40): It's interesting you mentioned data. I had a conversation with someone else in the industry yesterday. We were talking about data, emails and things like this, and this is a bit of a caveat, by the way. We were talking about data warehouses and where you can get information, and he was telling me you can basically get someone's real-time cell phone location from these data warehouses. A lot of people are now becoming aware of how much data is actually out there, how much data is being captured, and how careful you need to be with it. So I can understand why getting that SOC 2 compliance is top of mind, especially as people become more aware of it.

For ourselves, we are running enterprise-level certifications, and we find time and time again that the reason people are most interested in taking our course is that they need to get their entire enterprise, institution or organization up to speed on the intersection of AI with data compliance. It's top of mind for every organization out there, so I can understand why you're doing it.

You mentioned how it's different from ChatGPT, in that it takes your data specifically, your data logs. Could you tell me a little more about the background technology, what's happening on the back end? Are you using a proprietary model, are you running from APIs? What is it about your underlying technology or process that separates you from the other competition out there?

Shawn Miller (30:28): I'm not going to go too into detail on how the back end works.

Liam Lawson (30:32): The secret recipe.

Shawn Miller (30:32): Yeah, yeah. There's a level of complexity to creating content and actually being able to mimic someone's personal style. To give you more information: if you were to add Inboundr to your Slack workspace, it's going to go through all of your past LinkedIn posts so that it understands the way you write, your style, your personal flavor, that kind of stuff. Then it's going to take all of your conversations. So when it creates a post for you, it's not only going to be written from your point of view, it's also going to be written in your tone and style. There are a lot of steps in building an AI to do that, each essentially handling a different part of it, and it's a variety of steps that pull that final output together.

Liam Lawson (31:21): We've used similar tools as well, and it's harder to get right than people give it credit for. They think it's going to be a magic bullet, but it's not easy to get right. We've had conversations with people who raised millions of dollars of funding, and they still don't get it right. So I'll be excited to give it a try and see what sort of content we can push out from it, especially as a Slack connection. That's perfect.

Something I'm really curious about: I'm a co-founder myself, with [name unclear] and Martin here at AI Tool Report. What has your experience been like co-founding Inboundr? What kind of issues have you run into, what have you learned in terms of delegating responsibilities and communication, and what's been the biggest hurdle so far?

Shawn Miller (32:11): My co-founder and I previously did Simply Eats together. We had a third co-founder on that venture who's not with us for Inboundr, and with Inboundr we've teamed up with two more co-founders, so there are four of us in total now. Obviously my co-founder and I had past experience together, so we work quite well together. He's extremely technical and I'm extremely sales-sided, so we really didn't have any overlap in skill sets. It really was the polar opposite of strengths and weaknesses, so there were no real issues to overcome from that perspective.

Moving over to Inboundr, of the four of us, two are business-sided and two are technical-sided. Really no issues on that side either. We spent a number of months talking together, meeting together and learning about each other's past experiences before we dived into co-founding something. And even before we jumped in, we figured out what everyone's role was going to be within the organization: what are all of our skill sets, what are we weak at, what are we strong at. That's how we put all four people into four really different segments. So there have been no struggles or hurdles yet. Everything's been pretty smooth sailing on our side.

Liam Lawson (33:38): Wow, very impressive. Sounds like there's some insanely good communication going on over that side of the pond. You mentioned you'd worked in the crypto industry before, is that correct?

Shawn Miller (33:54): Yeah.

Liam Lawson (33:56): All right. I'd love to learn a little more about how you upskilled yourself in AI. What sort of learning resources do you use, how do you learn about the technology, and how deep in the weeds do you go? You said you're more on the sales side, but you obviously understand how the technical product works. Tell me about the first moment you heard about AI, and what your AI education has looked like until now.

Shawn Miller (34:22): I started a company called Sizewise four years ago. It was AI for helping people buy the right clothing size online, so my familiarity with AI has stayed up to date over the past four years. That's where I definitely learned the most about machine learning and AI and how that stuff works together. It was a fairly simple use case for AI, and the complexity wasn't as significant as it is now.

But really, it's not so much that I relied on some outsourced news channel to keep up with it. It's more the technical guys I'm working with. Every meeting there's another technical snippet I learn about. They've got all the code pulled up in front of me, and half the time it's a lecture, with me learning how the hell they even build what they're building. So it's much more learning from my co-founders than staying up to date with some media source.

Liam Lawson (35:23): Got you. So you've got front-row seats to the best education on the market, essentially. When you're educating potential prospects, people who are going to buy or try the product, what are the biggest misconceptions people have about AI, machine learning and large language models? What do people get wrong?

Shawn Miller (35:51): Really, I'm pitching to people who are much more marketing and sales-sided, so we don't go too deep into the LLM side of things. The biggest misconception we're having to climb is that when people hear AI for content, it goes back to that sentence earlier: you think of the ChatGPT jargon it produces today, the stuff you can smell from a mile away, how obviously it was written by ChatGPT. We're all very capable of recognizing that style of writing now, so people are very conscious about using an AI writing tool and losing authenticity and quality, because the content that works on LinkedIn is raw and authentic. It's not the buttoned-up blog post you'd see years ago.

So the hurdle is changing their mindset from what they think about AI content to what we're trying to do. Once people get on board with the approach we're taking, that's when we really start to see the excitement about the tool. It starts with breaking down the initial idea of what they think the tool is going to be like when they hear AI writing, and showing what it actually is.

Liam Lawson (37:16): In terms of the future of the product, you're obviously in the initial stages. Is there any possibility you would delve into photo, video or other forms of creative as well?

Shawn Miller (37:31): We've thought about photo quite a bit. What we've been thinking is: once Inboundr generates a post for you, how cool would it be if it could take that context and also develop an image to go along with it? We've been looking into that a bit recently. It was a suggestion from a couple of our initial users, to see if we could build something like that. It's a great idea. It's a complicated idea, but we'll see if we can incorporate it into what we're doing.

On the video side, we've thought about it. Some people have pushed us to expand our tool into helping with Instagram content and TikTok content, so it's definitely on our radar. We've got a lot more interviewing to do to figure out where the fit is, where the problem is, and how we can easily take the tech we've built and help on that side of things. But it's definitely on our radar.

Liam Lawson (38:30): Why LinkedIn specifically? What is it about the platform that made you tune in on it initially? I imagine it's some combination of your ICP fitting perfectly with what you're producing, but I imagine you also see some growth potential in LinkedIn. I've heard that from a few people, so I'd love to hear your insights.

Shawn Miller (38:54): It's definitely a combination of that, but it's probably primarily due to how much we understand the problem. Having been a ghostwriter with the aim of creating content for employees so that it drives pipeline, because of the challenges with other sales and marketing strategies, we got to learn that very deeply over a number of months. Because we understand the problem so well, we're able to build a solution that can actually help.

The benefit is that I use our solution. Aside from user feedback, I'm able to go in, use Inboundr and actually verify whether I like what it just wrote for me or not. That makes it really easy to validate whether we're on the right track. An example I can give you, as opposed to that, is my past startup Sizewise, where we were helping clothing brands reduce their returns. I've never had a clothing brand. I don't know the problems they face, so I could never verify myself whether we'd built something right. I had to rely entirely on user feedback, and that became a problem at the time.

So with the next business, I was really looking for that founder-market fit perspective. Do I understand the problem thoroughly? Can I solve my own problem and build a solution I would be a user of? That's why the focus on LinkedIn content, just because we understood it so well. We are very open to moving into other platforms, but we need to make sure we really nail whatever that problem is before we try to build a solution in search of a problem.

Liam Lawson (40:28): Got you. I imagine as part of that research and development process you're doing a lot of digging into competitor products and services. Where is your product differentiated? What is your USP, your competitive advantage, compared to the solutions already out there?

Shawn Miller (40:47): There are a ton of these AI content tools now, whether for blogs or for LinkedIn content, and I would say 99% of these competitors all miss context. Their technology doesn't know anything about the user it's writing on behalf of. Without that, you have no authenticity. It's entirely generic, and you're going to sound the same as every other writing tool out there. With the ability to pull in and query that library of all your past conversations, we can actually learn your point of view and your perspective on various topics, which is key to writing content that's authentic to you. That's where we stand out: our ability to grab as much context on you as we possibly can, to make the writing as personalized as it can be for you.

Liam Lawson (41:36): I think the hardest part about that is being able to convince people that AI is not scary, it's not terrifying, and that you want to work with it. This is probably something you're thinking about all day, every day, but what would you say is the biggest way you're stepping over that hurdle of convincing people to trust you, trust AI, let it into their life and let it produce content for them?

Shawn Miller (42:00): What's interesting is that we're actually so similar to all of these AI meeting note-taking tools. They're all recording your meetings and then giving you insights or updating your CRM. Having access to a CRM is probably, in my opinion, the scariest integration for an AI. That's where your company's data lives. If you study all of these companies, Gong and Chorus from ZoomInfo were kind of the two earliest ones, [name unclear] kind of next, and now there are a million of them. They all had a huge hurdle early on with adoption because of what they were looking to record. So we don't really see it as a huge roadblock. We know we will face it, but if everyone allowed these AIs to record their meetings, update their CRM and send an email to their customer afterwards with the update, all of that kind of integration, we really don't see it being too much of a roadblock for us.

Liam Lawson (43:05): Talking of fears, talking of excitement, let's take things meta for a second and talk about the AI industry as a whole. What are you most excited for in the next few years? And on the flip side of that coin, what's something you're scared of, if you're scared of anything in the industry?

Shawn Miller (43:27): That's a really good question. I just started reading one of the co-founders from OpenAI, who put out some 150-page PDF on the future of AI. I'm forgetting the name of it now, but it's the idea of superintelligence arriving by 2027 or 2028. I guess the biggest threat is how much this will disrupt a lot of the current applications of AI. Is this going to put a lot of startups out of business, or is it going to fuel a lot of startups to be better at whatever use case they're solving? That's the biggest fear for me, at least. I think it's a really interesting conversation, because a lot of people will have a lot of different opinions on it.

The biggest worry for me is how disruptive it could be. When the first wave of ChatGPT came out, you saw a lot of different use cases and previous business models getting disrupted, because a better and cheaper technology came along. So I guess that's the fear for the future: if some sort of superintelligent AI came along, it could probably disrupt what we're doing, along with 99% of the other AI use cases out there.

Liam Lawson (44:53): Is that also the same answer for what you're most excited for, or is there a different answer for that?

Shawn Miller (45:01): It would be the same answer. I'm not into AI enough, from the actual backbone of the technology, to give a really interesting answer on this side of things. I'm sure if you got all three of my co-founders on and asked them that question, that would be a really interesting answer.

Liam Lawson (45:19): Maybe we'll need to get them on the podcast as well. You mentioned that AGI could potentially take loads of these jobs, or just wipe out a bunch of these startups. If you were to give advice to someone looking to start their own AI business or AI startup, to make money in this AI entrepreneurship space, one, what would you say, and two, would you even tell them to start?

Shawn Miller (45:52): I would tell them to start, otherwise I'd be going against my own advice. I think what you need to do is build quickly and solve an actual problem using AI. Don't use AI and then go hunt for a problem. Actually have this problem yourself, and solve the problem for yourself. You'll be able to get somewhere over the next one to two years with that use case, and then you'll just ride out what happens with AGI. Who knows how that's going to go, and I'm not the person to tell you how it's going to impact the rest of the industry. So I would definitely say go for it. It's fairly affordable to start some sort of AI startup right now, so I would say go for it.

Liam Lawson (46:38): There's been lots of stuff coming out. I don't know if you've seen the stuff with Cursor and Claude, people building AI applications in the space of an hour. It's a massive thing right now.

I want to round off this podcast with a bit more of a deep and personal question for you, and it's kind of a two-parter. One, why do you do what you do? What motivates you on a personal level to continue pursuing entrepreneurship, going through these iterations of businesses and trying? And two, I'm not sure it's directly related, but could you speak on the environment you grew up in, and whether that part of Canada was conducive to entrepreneurship?

Shawn Miller (47:23): On the first part, starting all of these businesses: I'm definitely just a really competitive person. I like to see how far I can push myself and what I can accomplish. I feel a lot more fulfilled going after a startup idea, trying to take something from zero to one, than the idea of working a sales job for somebody else. So it's a combination of thrill and competitiveness that drives me to entrepreneurship.

On the Canadian side, the university I went to was great for supporting entrepreneurs. It really wasn't a huge focus of mine back in high school, but once I got to university they were great. They had all kinds of accelerator programs for funding different startups. My first startup was funded by the university. They gave us $12,000 in seed funding to get it off the ground, and they run cohorts like that a couple of times per year, so they're very supportive on the entrepreneurial side. The people I met at that school were very entrepreneurially minded too, so it worked out quite well with the school I chose.

Liam Lawson (48:39): What school was that, just so we can throw them a plug?

Shawn Miller (48:43): Western University.

Liam Lawson (48:46): All right. Dope. I want to round off with one last creative question, if that sounds good to you. There's no right or wrong answer, it's just a bit of a weird one. I don't know if you watch Formula 1. It's massive over here in Europe. There's actually a pretty famous Formula 1 driver for Red Bull called Liam Lawson. He's just about to enter the Formula 1 rankings, he just came up from F2. So when you search my name on Google, I'm way down the rankings. It's all about this handsome 21-year-old guy from New Zealand. He's beaten me out in every regard.

I was searching for you, trying to find more information as I was doing research for the podcast, and I noticed there are some other Shawn Millers who are invested in their own AI companies. There's one who owns one called [name unclear]. So what's a traditional way you're going to beat them in the Google rankings, and one creative way you're going to beat that guy in the Google rankings eventually?

Shawn Miller (49:51): That's an unexpected question, but a good one. I think to beat them in the Google rankings, it'll be a competition of who builds the biggest company. Google is going to populate more towards whoever that person is, based on the sheer size of their company compared to anybody else. For that reason, it seems the most famous version of any common name, the person who rises to the top, is just the most accomplished. So I don't think there's really a hack. That's just the way it will have to go.

Liam Lawson (50:26): That is the most sensible answer I've heard. I asked a similar question to someone else, and he said, I could build a bigger company, or I could just go around murdering people and maybe that would put me on the Google rankings. I was like, you know what, fair enough, I'll give you that one.

I wanted to thank you for taking the time. I appreciate it. I know you're probably very busy, as every startup is, especially AI startups. If people want to find out more about what you do, who you are and what you're building, how do they get in contact?

Shawn Miller (51:00): I'm active on LinkedIn all day long, and I answer every message, even if it's a cold call, so feel free to hit me up on that platform.

Liam Lawson (51:10): All right, awesome. Thank you so much, dude.

Shawn Miller (51:13): Thanks for having me, Liam. I appreciate it.

Get the five-minute AI briefing

Join 400,000+ business leaders who start their weekday with The AI Report. Subscribe free