9 Jev Use Cases for Work (Plus the Mistake to Avoid)

Those of you that want to spam with AI, just keep in mind when you use Jev as a publish gate, only 7% of the drafts meet my quality bar. And I'm going to talk about how you can use it from a marketing standpoint, right? So I'm going to give you a couple of use cases for marketing and what we're actually using in production right now and how it's been helping me. And I mean, it's been helping from an SEO, AEO standpoint, which will be the first example. And I'm going to go over Maybe towards the end, I'll go towards a thing that you have to keep in mind. I was actually using it the wrong way. So let's go ahead and get into the first example here. And what I'm going to show you on my screen here is a couple of examples and then I'll go into a little numbers, some numbers I can share with you as well. So we'll probably start with the third one over here. So you can use Jeff to scan your external content to determine the best content spikes, then draft the best, then have Jeff evaluate content quality, enforce content guidelines, and avoid duplicate content. This is quality SEO AEO content at scale. And so what I'm literally doing right now is I'm using a Grokbot workflow. And what it's doing is it's pulling the external and internal content. But not only that, because I have this now, I can look at what competitors are using or competitors are creating. I can quickly classify them. So if 2 or more competitors are making the similar type of content, I can then look at the SERPs, look at how competitive it is, and then I can attack that content, right? That's another example of work that you can do. So you can look at your competitors' content, you can look at your external and internal content. You can also look at content that's decaying on your website too. So content that's decaying, maybe it's high-intent content, but you can see that the traffic is going down, whether it's from an AEO visibility standpoint or SEO visibility standpoint. But these are all things that you can do now. And you can gate the quality around the content that's being published, even the image quality that's being published as well. And so this is extremely helpful because— and I'll just talk about some numbers over here. So, Jev basically read about 3,200 pages. To be specific, 3,208 pages in 178 seconds. This is on my website and I'm just reading it off Grokbot right now. The first version flagged about 2,200 pages flagged as needing a fix, which was useless, right? But then it got a little better. So the real culprit was our scraper. So then after we fixed that, the fixed list fell to 2,666. So, and they basically fixed the 86 pages that actually got the most Search Console impressions, right? Yeah. So basically, JEV is a good classifier, but you're going to need to work on it a little bit, which is why towards the end, I'm going to talk about why it's really important to have these gates at different levels. It just can't be at the— it can't just be after a piece of content is created or after something's analyzed. You want to do it before and after. I'll give you the reasons on that. Okay? So that's on the AEO SEO side of things. And to go a little deeper on this one. Every draft that we publish gets that yes or no check, right? For like, are we, one, are we inventing stats? 2, is this a duplicate of a live page already? Which is a huge thing, right? If you're thinking about AI, one of the biggest pieces of anxieties that I had was, okay, I can publish all this stuff, but A, is it going to be crap? But B, are we duplicating what we have already? And also, should we be keeping? Should we be consolidating? Should we be deleting? Should we be upgrading pieces or updating pieces of content? Those things should all take into— you should take into account, right? Because updating has always worked. So every draft gets a yes or no. So we check for duplication. of a page, thin evidence, and AI tells, right? We don't want a bunch of em dashes, if this, it's not X, it's Y type of stuff, or invented stats. So over 3 days, we basically had 11 of 149 drafts cleared and 138 were held. So the lesson is that most AI drafts should not shift. So if we do the stats on this one, about 7% of AI drafts actually cleared. Here's the AI tell over here. If something is, do we have anti-slop skill in there? Is it ship ready? Do we have images that pass? What the justification is, any links to content that— this is like from content that I've published. So is it, you know, are there any links to content that I've covered already? And then here's the title as well. And then it shows the ID and then we can kind of preview the content as well. And then we can see the target keywords as well. So you can see like the keyword volume over here, right? 4,400, 720, 9,900. And I'm using the Ahrefs API and the Data for SEO APIs over here. So these are things that we want to go through. And then we even have this measurement thing as well. So measurement here, this measurement bot, what it'll do is it'll actually measure the success of how we're doing. So is this being indexed? Is this gaining traffic as well? And eventually, is it driving conversions as well? And so I'm constantly thinking about how to add pre-gates for Jev and post-gates for Jev. By the way, if you're enjoying this right now and your agents don't know how to talk to all the tools that you have, you need an API connector. And that's why we have the Single Brain Gateway. So just go to singlebrain.com/gateway. We have a secure connector there that you can use. And not only that, if we figure out what your goals are, we can put the right workflows in front of you from a marketing and sales standpoint so you can generate more leads and you can generate more pipeline ultimately. So just go to singlebrain.com/gateway and we'll see you on the other side. Now, the next example I want to go into a little more details on is using Jeff to scan all the people who follow you on Instagram, right? So founders, investors, influencers, recruits. I kind of made a tongue-in-cheek joke here. You can even use it for people you want to date. And so the cool thing about Muse is it allows you to just, I mean, it's connected to all this stuff that you have with Meta Platforms, right? And so I think it's extremely powerful. So here's an example of people that follow me. I basically use Jev to classify real quick. And then I realized something that I actually needed to classify deeper, not just based on, are they a collab fit, a buyer fit? Are they agencies? Are they LA local? So I added in all these classifiers over here, which is cool now. I didn't have it before, but I only had like one classifier before where maybe it was based on like follower count and it classified it really quickly on 116,000 people that follow me on Instagram. And now it's like, okay, it's checking if they're verified. There's a follower count here as well. And so this is cool, right? So I can figure out who does it make sense to collaborate with. And so I can see over here, okay, the Marbella Hills Estates. Okay, this is maybe somewhere I might want to travel to. Look, there's like a million followers on this one. I'm like luxury real estate, Marbella, private sales, right? Maybe that's a real estate account. But there's like Cooper Simpson over here. Like I like his stuff around AI and you know he he's look it's categorized him as a creator. Collab fit is like a I don't know this is I think zero through four so he's a three. Buyer fit probably not a buyer fit for us. Is he an agency that maybe we want to acquire? Is he in LA? Like can I meet up with him? But this has taken a lot longer to. Get this from, but this is helpful to me because if someone follows me already, let's say they're a recruit or let's say they're someone that might be ICP fit to do business with us or someone who's in LA, maybe I want to invite them to the tennis things that I do and the little tennis things that we do in LA, or maybe they're an entrepreneur. Maybe they can join the entrepreneur poker game. So there's a lot that you can do with this. And sometimes you don't know who follows you and now you can do something with this. You don't have to pay for any additional software. You're just paying this API key. For example, I really like MixExcel. Miss Excel. She creates a ton of content about Excel and she has like, I don't know, millions of followers. So that's really cool, right? And I didn't know until recently that she follows me, but you can just come down the list here and then you can even have an AI just simplify this even more for you once you have this data. And you can just ask questions to Muse about this, right? So I think that's really cool. Anyway, all that to say, I also even have it running for Meta ads. Okay? So what I'm doing right now is I'm having it mess around. By the way, look, there's Jeff Gates in here, right? So, you know, this is a— are these ads a buyer fit? Is there a good hook here? Is there a good voice here, right? So it's scoring it before and after. And then what I'm doing here, I'm just saying, hey, you know, look at the— because it's connected to my Instagram, I want you to look at my organic posts that have performed well and then figure out the ones that look like they're good to maybe be posted as an ad. And then tell me what you need from me, right? Tell me what you need from me. Do I need to give you a call to action? Okay, are you going to make the landing pages for me? Okay, if you're not going to make the landing pages, What is the handoff MD file that I can give to my other harness? Eventually, there's going to be one kind of alpha harness you have that's managing all of your agents over here. But that's what you can do with this. You can have— look, I put like, I want no AI slop. I want JEV-gated, 2 clean variants. So a lot of this stuff now to me is JEV-gating. You're JEV-gating things to make it better for— so you don't have to make as many decisions. And you can make decisions at scale as well because a lot of stuff is just kind of like a yes or no or like a confidence score type of thing. So I think it's great. Look, it's already gone into my Meta account and already made these. So I don't need to wait for my team anymore, right? So these are from my posts and it used computer use. And then we used Jev to basically gate the ads. The ads that we— I said, hey, the ad should have these criteria over here, right? It should be related to attracting people that are interested in either AI-native marketing services or AI implementation. And it picked the right videos that have the right engagement. It didn't pick videos that are maybe more for a wider TAM, like a wider total addressable market. So that's cool, right? And now there's a landing page for this too. Not live yet. Okay, it's being built with 2 thank you pages over here. This is going a little bit into Muse, but I'm just saying you're going to need a classifier here so that it can reliably make decisions for you before something gets published and after something gets published. That's ultimately what it is. Let me tell you why that piece is important too. One more thing. If you need help with marketing from an AI-native marketing team, then you've got to check out Single Grain with a G, singlegrain.com. That's where we help people build these AI marketing agents that live inside of Slack or Microsoft Teams, but we also manage it for you. We have a lot of these workflows that can help with AEO, SEO that can help with cold outbound. So all that infrastructure we build, we manage it. And talking about ad creative scaling, we are looking to work in the future and we are looking to completely disrupt how agency work is done. So singlegrain.com with a G and we'll see you. Learned that because I was classifying everything after a piece of content was created. So let's say after the outreach happened, I was wasting valuable tokens, right? So it could be valuable frontier tokens that you're wasting. So if you actually have a classifier happen before the outreach, let's say you're reaching out to people from a recruiting standpoint or a sales standpoint, right? You want to shortlist who you're reaching out to first, and then the reach out happens. And then afterwards, based on whatever response that they give to you, then you can decide whether they make it to the next step or not. So you can actually add multiple gates. And because it's so cheap, you might as well have multiple gates, right? So this is known as the JEV sieve. Claude Code actually caught this for me when I was— when I asked it to figure out outside-the-box things that we could do better. So this one is, I think everyone should be doing this, not just from a marketing standpoint, but from a sales standpoint too, right? So using Jev to scan all your customer calls for case studies, objections, churn risks, et cetera. And what you can do is you can basically, instead of having it only scan like 10 calls, you can have it basically chunk all your calls. So let's say you have a bunch of gone calls, right? Maybe it's only looking at the first, maybe it's looking at like 5-minute segments of all these calls. And then it's looking for trends from there. And then that actually ends up being a lot cheaper versus having your frontier tokens go first and it being more expensive. That's actually the one that Claude Code caught. And then the sieve, I was just talking about that sieve then expensive model. So Jev tagged 517 chunks of calls and posts and passed only 189 to the costly model. So on YouTube transcripts, it kept 198 of 225 chunks and only half of a hand-checked sample was useful. The lesson is to always measure the hit rate. So this is where humans aren't going anywhere anytime soon, right? We still need to be We still need to be the judgment layer that this thing needs ultimately. And then like I mentioned before, using Jeff to scan all inbound and outbound recruits, running them through your talent evals, shortlist people at lightning fast speed. Okay. So those are just a couple more things that you can be doing. And then also when you're reaching out to people, let's say someone became a CMO at a new company or maybe a company raised new funding, right? So we try to reach out from an outbound standpoint with news openers that are actually news. And if there's no news, then we just don't send anything, right? Yeah. The outreach bot that will send a congratulations line in the first line with data news from the last 30 days. So using the last 30 days skill from that. Yeah. And then in one batch, by the way, just to share some numbers with you, 3 of 14 people qualified and the other 11 got no opener instead of fake personalization. So real numbers for you, right? Routing LinkedIn engagers is another— this is a big one actually. So one comment or DM contains— counts as a hand raise and Jeff routes each person. So of 151 people, 0 were buyers and 23 were agency owners. Guess what? Because we're looking to buy agencies and we're looking to kind of integrate how we go about doing AI, 23 of these actually make sense. Let's say someone has, you know, our range right now is like, let's say, you know, talk about 80 employees, right? That's something we're looking for, right? But that might route to the M&A bot that we have inside of GrokBot, for example. But if they're an ICP fit and they've engaged with 3 posts in the last 30 days or so on LinkedIn or X, we can find those people, reach out to those people, and see if we can get a call with those people, right? Because the fact of the matter is they're constantly engaging with our stuff. And I think this is just going to get more and more sophisticated over time. Right now we can look at LinkedIn, we can look at X to see who the engagers are. And the last one I'll call out over here that my little Grok bot's calling out here is the send door. So no bot sends as you without your yes or locked rule with a daily cap. In the test run, it blocked 10 of 10 bad sends on the First real day, it allowed 77 of 83 items and flagged 6. That's what Jev is ultimately. So there's a ton of use cases that you're going to come up with here. But I gave you some ads use cases. I think I gave you some social use cases as well. So ads, social, I think AEO, SEO, any type of outreach as well, generating more case studies as well. That definitely helps from a marketing standpoint. You know, inbound, outbound recruits, but you can use that for influencer outreach as well, partnerships as well. There's just so much that you can do with this, and I think we're just scraping the surface right now. I'm actually curious from you, if you want to hear more from me on Jev builds or what interesting use cases that you've seen, drop it in the comment below, and we'll see you in the next video.