Using Generative AI to 10x Your Email Open Rate with Larry Kim

0:00Hey guys, Eric Su here. So the following is a clip from the Leveling Up Founders Mastermind that happened in Beverly Hills. and you're going to hear from some amazing founders, some amazing marketers and investors. And this was a private event that happened. So again, you'll get a preview of it. If you want to learn more, you can go to levelingup.com slash founders. Once again, levelingup.com slash founders. And without further ado, enjoy the clip. When it came to starting this new business, customers.ai, we're in the sales outreach automation and data company for, it was, it's a very interesting product.

0:32I think that part, I think we've figured out. It was definitely in need of a go-to-market channel. And as I was reviewing, like, how the heck are we going to get this in front of people? One of the key discoveries was that what worked for you in the past doesn't necessarily work for you today. And not to say that we don't do content marketing, but it's just harder. Like, I have a website that does about 1 or 200 ,000 visitors a month, 6 ,000 domains, and 36 ,000 backlinks. But it's just really hard. I'm competing against... Is Neil here? No. So, Neil, I'm just calling you out here. It's just very hard.

1:07I'm competing against websites like Neil Patel, who's got 12.8 million backlinks and 128 ,000 referring domains. And it's just the idea that I could somehow magically displace these companies like HubSpot or WordStream or Neil Patel. It's just fantastically unrealistic. And even the ads, which were a big part of my go-to-market for my last business. I was buying clicks for$0.05. They're now$50 a click. And Facebook ads were like$0.50 per thousand impressions. They're now$800,$900. It's just I needed a new marketing channel to exploit in 2023 to get this thing moving like how I did before. And don't laugh, guys.

1:44My master plan, at least the original version 1.0 of it, was outbound emailing. The plan was pretty simple. Get a list of 400 ,000 marketers and salespeople from something like ZoomInfo or Apollo. Just download that list and then bombs away. Email these people. That, of course, did not work. It does not work, people. I was shut down by Google. I was shut down by SendGrid. It destroyed the sending reputation. It was quite a disaster, actually. However, I'm very stubborn, and over the last two years, I believe I've perfected what I could describe as a way to actually make this work. I call it, this has to do with leveraging new ideas and new technologies, specifically three areas of concern.

2:25One is user intent data and turning that into emails. A second idea is these advanced email delivery tools. And I'll talk about that in just a second. And the third kind of category is generative AI. And if you can stitch together these different ideas, this is a very high-leverage go-to-market that is very easy to copy in your own organizations. So my big claim today, guys, is that AI-powered, intent-driven, outbound email marketing will save the day. And what I mean by save the day is make you tons of money, leads, profits, sales. If you've never heard of intent-driven outbound email marketing, don't worry.

3:02It's just a word that I made up. It doesn't actually exist. But conceptually, the way I'd like you to think about it is that it combines elements, the best elements of inbound email marketing and outbound email marketing. Suppose, if you will, if you could figure out a way to combine the higher engagement rates of inbound emailing with the unlimited supply of emails and low costs of outbounding. And this is where I've landed here. So remember I said there were three components to this machine? The first component is leveraging user intent data. This has to do with harvesting intent signals from users and then turning them into email lists.

3:43That's actually one of the things that, among other things, that my company does. We have developed a way to identify the anonymous website visitors that go to your website. It's crazy. So those email lists that get generated, therefore, they have a high intense signal because they're not just random people. They're people who are in market for your product or service, and they've heard of you. So this is like almost a layup here. This is just one example. There's tons of different data sources that any company could leverage to generate these outbounding lists. Suppose you're an e-commerce consulting agency.

4:19One of your user signals, the signals that your users gives off prior to becoming a customer, might be the installation of something like Shopify or Klaviyo. So using products like Built With, you can subscribe to kind of technology additions or removals, depending on what the signal is for your target buyer. Suppose you're like an office furniture retailer, okay? Who are you going to sell your office furniture to? You can just subscribe to the SEC business financing records. When someone raises$5 million or whatever, you have to fill out this form, and it shows up on this SEC website. And those are the kinds of people that are going to buy your office furniture.

5:00LinkedIn Sales Navigator, of course, very good at providing data around new hires, promotions, work anniversaries, company in the news, that kind of stuff. And there's tons of crazy data products out there that you might not be aware of where you can subscribe to life events like Baby, New House, Death in the Family. You can subscribe to physical locations, people who enter a Goodyear store on a certain street. You can go after people's search history and get their identities even. And there's a huge difference here between static interest-based email lists, which are used to power conventional cold email outreach, versus dynamic intent-driven email lists, which is used to power these intent-driven outbound marketing strategies.

5:42And the punchline here is that the intent-driven lists have 10 times higher engagement. More clicks, more opens, more replies, lower complaint rate. It sends all of the good signal to the Google's algorithm, which improves the deliverability. The second big idea here I have to share with you has to do with email deliverability tools. So these are tools that are specifically designed to undermine the Google Promotions tab AI. Okay. The one that I use is called instantly and it's quite a racket. Basically you sign into this network of instantly.ai and then you're part of this network of a cabal of emails that are like sending emails to each other, like a big, a big network of sending emails.

6:26And occasionally something will get found in the, in the spam folder and it'll just confuse the heck out of Google by marking that as not spam and pulling it out of the inbox. So this is like a sporting analogy. You can't just have a strong offense and no defense. So the intent-driven list is the offense. It's a better list to begin with. The deliverability tools is like the defense. It's when something got classified as spam, we're going to have to pull it out there. My third big idea today is to customize the email content using generative AI in combination with the data intent signals that you've previously collected.

7:07As an example of this, companies like my own and other companies do provide a lot of consumer data. So for any one of you people, I have hundreds of data points of like where you live and how many kids you have and your marital status and your income. So you can augment the data that you have on these people and then leverage generative AI use cases such as this to generate spectacular outreaches. This one is actually me. So this is write an email for a guy who lives at 15 House Street in Dennis, Massachusetts. in the email, working to the copy, a reference that would be recognizable to someone who lives at that address.

7:45Also mentioned that he has got interested startups with two kids at home. This is what it ends up coming up with. It's crazy. It says, as a resident of Dennis, Massachusetts, you're no stranger to the notorious shark sightings at nearby West Dennis Beach. What the heck? This email, not only does this increase the delivery because it's impossible to fingerprint, because every email is completely different, it also is a delight to receive. An email like this gets 60%, 70%, 80 % open rates. I'm just scratching the surface here. There are many other generative AI use cases for intent-driven outbound emailing.

8:22For example, we could use generative AI to generate the text for text messages. But why stop there? You could use AI algorithms, the new AI text-to-speech algorithms, to have a friendly female voice read that text message and then turn it into a voicemail. Like, you could do video sales letters, record them once, have the AI customize your speech patterns and the sound to be customized at scale for every single recipient of that email. So what we're talking about here is, for the first time in the human existence, we can enable one-on-one hyper-personalization at scale, the benefits of which are super ridiculously high engagement and deliverability.