Mateusz vs Louise: Ahrefs Use Case Showdown

ladies and gentlemen, welcome back to the ATF's use case showdown. You're watching live from screens around the globe. The third edition of us bringing us two together marketing titans to face off in their most spectacular use cases of our platform. And now today agent A as well. I am your host Constance and tonight we tonight you are not just spectators. Tonight you are the judges and you will crown the champion and you will be taking away amazing ways HRS and Agent A can superpower your marketing. Oops. In the red corner we have in the red corner we with stints at Buzz Sumo PI data metrics and Cisco before landing up at Hrefs. Her time in her days goes into figuring out how AI search engines decide which brands get cited and which ones get ignored. She is our dark horse content marketer at Hrefs. Give it up for Louise Luna. Thank you. Lovely introduction. Thank you. Let me hear you in the audience if you are team Louise. Let's see your reactions and make some noise. Louise, do you have any words for your opponent and the crowd tonight? Uh, I don't. I've never been in a fight before, so I don't really fancy my chances, but I know I'll give it my best shot. I don't have any fighting to talk, Mattesh, but um I know you'll be a worthy opponent. [laughter] Okay. Humble, but do not be um but do not be swayed by her humble words. She will be a contender tonight for our next opponent who'll be fighting out of the blue corner. You may know him from the Atre blog for the last 5 years where he has been breaking down ideas and experiments and what actually works in practice across content marketing, SEO and AI search. Ladies and gentlemen, give it up for Mat Michael Shave. Hey guys, if you are team Maters, please make some noise in the chat. Matiers, do you have any words for your opponent today as well? Uh, I guess it's going to be a it's going to be a fight between two most peaceful people [laughter] on the team. So, yeah, enjoy. Thank you. You guys don't know, but like for all of the best case studies and uh in like impressive blogs that trend really hard on LinkedIn and get quoted by a lot of the other big platforms out there who mark uh who do a lot of work in marketing and uh like in SEO space like these are the people that are doing the real work. They are gentle giants but they really have some of the best like use cases I mean really of HFS and agent A that you'll probably see today. So you're in for a treat you guys. Well, let's then move on ahead to the rules of engagement. This showdown will consist of three rounds. In round one, each fighter will present two use cases. The first fighter will share both use cases and then followed by the second fighter. After both fighters are presented, you audience will vote on your favorite use cases. Whoever's use cases receive the most votes will win the round. In round two, we do it again. two more use cases from each fighter and this time they will get into their AI agent machines and so you see like a different kind of use cases to be presented in round two and we'll vote again to see who wins that round and then in round three the final round each presenter will uh each fighter will present just one use case their final knockout punch their most powerful use case that will blow their competition out of the water right and at the end of the Three rounds, we will tally up the scores and the fighter with the most winning use cases will be crowned the third HF showdown champion. And the loser of tonight's showdown will pledge to donate $500 to the charity of the winner's choice. So now that we have discussed the rules of engagement, um let's time let's go ahead and kick things off. So Louise and Madouch, are you ready? Yeah. Thank you. Yeah. Audience, are you ready? Yeah. Then let's All right. All right. Now, let's get it on. Round one. Louise will step up to the bat. Show us what you got. You have 10 minutes to present your first two use cases. the floor is yours and you can go ahead and take over to share. Thank you. Uh, okay, no pressure now. Right, I'm just sharing my screen and hopefully you can see my slides on my top use cases. Um, so first off, I'm going to start with finding untapped keyword opportunities in Keywords Explorer. Um, this is my kind of go-to way to find untapped keywords that our competitors might not have come across. Uh, and it gives me kind of new ideas for topics, especially around AI. And it's a way you can tap into our entire keyword database and it shows you trending keywords in a matter of seconds. So, I've pre-recorded some of these demos, so I'm just going to walk you through them now. So basically from the dashboard uh you obviously want to head to keywords explorer and when you're there you want to do an open search. So just click search or explore entire database. You don't need to put any keywords in and for the sake of this video I've just excluded sensitive categories. Uh but once you get here you'll basically see every single keyword in our database and but you want to find the top trending keywords that are related to your industry. So head to the include filter and this is where you can add seed keywords. So for example, I might include AI SEO or AI search or SEO or AI marketing because those are the kinds of topics that I want to write about. And when you do this, make sure that you click any word, not all. Just hit any word and then hit apply. And from there you will see all your seed keywords applied and they'll um basically show you all the other relevant keywords. So here you can see there's content marketing, marketing fundamentals, email marketing, but you're looking for trending keywords. So you want to sort by either three month traffic growth or six month traffic growth or 12 month traffic growth. And from there you'll start seeing interesting topics that you might want to write about that are trending. So for example here I might want to write about AI search tracking or brand visibility in AI search because we have brand radar which enables users to track their AI search visibility. So I might want to write about that and maybe I want to answer trending questions like which AI tool is best for marketing. Um and as you can see the keyword difficulty column there's like there's very low keyword difficulty or no keyword difficulty. So these are real untapped opportunities. But now we've seen the trending keywords. I'll show you how to find the trending questions uh because obviously that's so much more important now in a with AI to answer your users questions. So you can basically add a word count filter. So I've gone with seven words to find longtail questions and that shows up questions like for example which AI tool is best for marketing? How do I track my presence in AI search? And those are great questions that I would would want to create content around. But you can also filter by our pre-built categories. So for instance, I've gone with search engine optimization and marketing and monitoring software. Um, and then I've also added a question preset. So that essentially shows you keyword modifiers like who, what, where, why, when. Um and it will show you questions um like for example what is AI visibility or how to optimize for AI search and then you can obviously search by 3 months again 6 months and 12 months to see the trending terms and yeah this is just a great way to find untapped keyword and question ideas. So that's my first use case. Um, and now I'll show you my second I think. Is that is it over to me now constants again or should I hand back over? No, no, no. You can uh uh present your first Y cool. Uh so my second um my second use case is finding press links, brand mentions and PR opportunities. And obviously PR is uh has become a much bigger deal with AI. We've done our own research where we've seen that brand mentions for example are some of the strongest correlating factors with AI visibility. So this is my top way of auditing our press links and finding content ideas from PR um and obviously um new PR opportunities. So I've recorded another video here and I'll walk you through that. Um, so like before start off in the dashboard again and then hit site explorer and then you want to enter your domain and then from there head to the backlinks report and uh then what I do is I add a list of media domains. So uh in this referring page URL filter I'll drop in a list of major UK and US media publications. And how I do that is I head to Chat GPT, my good old friend Chat GPT, and I ask it for a commaepparated list of 100 major US and UK publications. Uh, and I'll give it a couple of examples of what I want. And then it will hand back this full list, which I then copy. Um, and then I go ahead and paste that back in the referring page URL filter. Um, and before you hit apply, you need to make sure you click any um on that filter, not all. And then you hit apply. And basically from there, you'll be able to see links from all your major publications, all the major news publications. And you can sort by page traffic or first scene, so that you're looking at your most recent links. And uh these are some of the links that we've acquired recently. So there's one from Forbes there, there's one from CNET. Uh, and then we can also see the journalists that have written about us that we could try pitching to in future because they're essentially warm leads. Um, and then we can look at the anchor and target URL and see which kinds of content they're referencing. So for Forbes, they're referencing our longtail keywords research piece. Um, and CNET, they're linking to our piece of research on AI overviews. Um, so and so on. So this analysis can kind of inform the types of content you create, the types of content you pitch, and the type of people you pitch. Uh, and then you can repeat this whole workflow for your competitors. So in this next tab, I've actually repeated the same report for Semrush. Um, and I've included the same domain list of publishers. Um, and I can see that actually getting some coverage from the BBC. Uh, and I can see what they've been cited for. And then I can open that. Um, and I I actually see that it's a piece of research on AI overviews. Um, and so maybe we could think about doing our own 2026 version of this study if we wanted to try and pitch to BBC. Uh, there's another BBC link they've acquired. And when you open it, it shows that um it's about chat GPT traffic data. Uh, and we actually have our own data on chat GPT traffic. So maybe this is something we could pitch to those BBC journalists. Uh and that's basically how you can get content ideas from competitor PR, but obviously it's not just about links. Um this is the content explorer tool and I've just added HFS in there and I've filtered for incontent mentions. And then um I've dropped in that media domain list. And basically this just shows me all of the mentions of HFS in the press. And I've added a highlight unlin filter. Uh, and if I come down here, I can see, for example, scrolling down there that Axios has mentioned uh, Hrefs, but not linked to us about a piece of Reddit research that we did in relation to Google AI overviews. So, that might be a study that we repeat in future or a person we could even pitch to in future. Uh, and then if I scroll down a bit more, there's also this uh, link uh, mentioned from the Daily Mail. So, if we wanted to get in mainstream media, we could learn from this. And then you basically want to rinse and repeat that process for your competitors. So I've done the same process for Semrush. I've added in the incontent filter. I've added in the domain list and I've added a highlight unlin mentions filter. And the article that caught my eye here was from Business Insider and it was on Vibe Coding. Obviously, we've just built agent A, our Vibe coding tool. Uh so I want to see who the journalist is here that I could potentially pitch to for future coverage. So, it's Kelly Burch. So, uh she covers a relevant beat clearly, so we could potentially pitch to her in future. And that's a really quick whistle stop tour of how you can use uh back links and content explorer to find PR and content ideas. I'll hand over to my worthy opponent, Mattesh. Now, uh that is use case one and two from Louise. Can Mat answer back with an equally powerful counter strike? Mat, the ring is yours. Uh 10 minutes on the clock for you as well. Show us what you got. Thank you, Constance. Um great stuff, Luis. Um let me just try to share the screen. There we go. I hope everyone can see what I have here. So, uh the uh first use case would be to find relevant Reddit threads with highest search visibility. And um you might be wondering why Reddit. So I don't mean to scare you, but Gil Google made it so that you can have like a um fouryear-old thread from Reddit posted by someone who was um uh banned from the entire um Reddit platform uh ranking on page one for you know branded uh um keywords about you. So that's kind of a that's kind of a anecdote about why Reddit is so um important nowadays. And it's not only like you know it happens to uh particular brands just everywhere. I mean if you um we're going to look at actually two ways that Reddit is important. First of all, it's um how it shows up in the organic SERs and the then the other one will be the threads that mention you and get cited in uh AI answers. So if you like zoom out and look like um at the uh like the size of Reddit, it's just amazing. It's like the sec the sixth I think the largest website on the on on the web. Um tons of users. um tons of searches, you know, this uh um people like search with the Reddit modifier and then Reddit is like on uh 77% uh of the time is always there when somebody looks for um products and even Reddit has have they their own uh internal research about how many um threats are of commercial nature and it's like 40%. So um as for AI visibility and and and that is something that by the way this is something that was like one shot made by agent agent A. I will show you later how uh I did it in like less than a minute but Louise discovered that even if Reddit isn't cited um in chat GPT it's it's actually there but in the training data. So, Reddit is there when people search for you, when they click on on on Reddit, and even even even when they don't click because that's that's already in AI answers. So, um there are two ways to uh find the most relevant, the most impactful u threats um via hrefs. And in site explorer, it's actually like three clicks. You just enter the theuh domain uh as the target. Um then go to organic keywords and then you will have all of the keywords that Reddit ranks for. Then if you um enter uh like a topic or or a keyword in the the keyword filter, you will have all those relevant threads to you. So that is that is for SERs and as for AI visibility it's also like that's even also like three clicks you just go to brand raider um click on the uh on the left hand side you will have the uh uh Reddit um SER visibility then uh you can also uh set the country uh right here it's United States then you can put in the filter your brand name and actually let me show you um how it looks live. Um there we go. Yeah. So, uh you will see all of the threats that are cited in AI and you will see them right here. Right. So, for example, this is probably um this one right here probably not that important, but this one with all the citations, that is something where you could go and you know, maybe have a say, you know, in in the in the conversation or maybe like fight back, um or just, you know, learn why, you know, um people complain or, you know, um have anything important to say about you. The second one would be track specific page performance in AI search. Um this is a very simple um yet powerful um feature. I have like a special relationship to this one because this is in uh brand radar because of an of an experiment that I did. Um let me show you an anecdote. So we did this uh experiment where where we um planted some self-promotional you know uh blog posts about hrefs evolve just to see what happens right you know that kind of uh post and the funny thing is is that um and I'm actually very glad that this is happening because this right here is like the final nail to the coffin of you know uh self-promotional listicles. you can get like six citations from your content and zero mentions. So, thanks to this feature, I know that self-promotional listicles is bas basically like, you know, free marketing to your to your competitors, right? Um, let me show you how to pull that up. And you may have missed that because I think it's like maybe on the platform for like one or maybe two months. So you go to sited pages uh then you click on your tracked prompts and then you click on that watch list tab. I should uh have it somewhere here live as well. Um let me pull it up real quick because I want to show you one cool thing about it. So you will have like um first you have to add pages to your watch list, right? And this is very handy when you do experiments like I did or like create content specifically for AI search and then watch you know whether it gets cited where whether you have mentions thank to it. So um you will have all of the all of your pages here. uh you will have um you will see how many um answers it's it's been used in and a chart that shows you like a historical um citation hit. And one uh one other thing that you can do with that, you can um take that to agent A uh and do like a like a custom dashboard. Um I actually call it the AI coverage um of how this these pages you know perform, right? So for example you can even see like uh the uh citation rate of uh those pages was like 40 pages were like only four 4% right. Um yeah that's it. Thank you. Okay I'll go ahead and take over uh the screen here. All right. Uh, I need to open up something really quickly. All right, so what a feisty first round. Four incredible use cases from both Louise and Medouch. But now, audience, it is time for you to vote. On your screen, you'll see a popup um poll pop up with voting options very soon. Vote for the use cases you believe were more powerful and more useful, more likely to move the needle for your business. You have 30 seconds to cast your vote. Okay, the clock starts now. Go, go, go. We stopped the the voting in 30 seconds for real. So, the votes are coming in. One of them is really heavily contested. Okay. And wait, last few votes. Five, four, three, two, one. Okay. And the votes are in. How do I count this? Uh let's see. Let's give. So technically since the voting is now closed and results are in, the winner of round one is Louise. But interestingly, the second use case is a tie. Like a full 50% tie. So like it's really really close you guys. Like Louise, you take the early lead, but don't let your guard down as Mat will be hot on your heels. So let's move on. I'm not sure how to count this now. We'll see what happens later. All right. Now the pilots shall enter their Agent A platform for the next round of use cases. Let's engage pilots. And we're here on round two. Both pilots have now deployed into their battle suits of agent A. The stakes are rising and the tension is building. This time, Mat gets to strike first. Mat, the floor is yours. Go ahead and take over the share and show us what else you've got in store. Yeah, thank you. Okay. Um, I guess someone else is sharing the It's me. Uh, I'll stop it. Great. Go for it. Okay. How about now? Yeah, go. Maybe I'll try to maybe I'll try to zoom in. Okay. Okay. The uh first one uh find outdated pages cited by AI. So the thing is that u we know that AI in general uh prefers fresh content but when it when it comes to branded searches searches about your product service your brand it can surface up really old content you know like even seven years right because u it it may not have any other choice right um we know that even now chat GPT tries to improve improve their results by you know asking um querying the uh site itself uh in in in the query right um to to to show like the uh answer straight from the source so I guess that will be having like up-to-date documentation help docs you know blocks with your use cases will be even more important um and uh this is actually a pre-installed installed skill in agent a this is something that I also helped to um develop. So since that is pre-installed what you can just do you can just go in the chat and and uh ask do an audit freshness of pages cited by AI from uh for my site or you can go to the um uh skills tab and then you will have the citation freshness audit. So um then it generates ask you a couple of questions like for example you know which site which which directories and what is crucial here what will be the source of the information about the publication and update date of your pages. That is why uh pretty much that is the the whole reason why we've develop developed that because with sites like ours you know big sites with with different um content directories you will have like um inconsistent ways of telling you know the users and and and search engines what is the date. So, for example, it could be the last mod uh um or it could be like a special schema that you've added on the block. So, um if you go like with your default settings of, you know, find the find the uh um the date of the page, you will probably miss that. But this skill tries to um look in your HTML content for information about um for any um hits about uh your publish date. Right? So it uses it uses um the search the site explorer um snapshot HTML snapshot of the page and uh if not found it defaults to um the data that we have in Brent Rider and uh that is actually live uh lickable thing that I'm showing you right now. Let me show you real quick. So it will generate this app for you, you know, with sliders. Uh you can click around and for example um right here um content that is um uh we have like four 429 pages that have been published or updated um 30 months uh from from this date. Um okay. And the uh second one is something that I call uh source of truth. Um this is like a a wiki a knowledge base for your agent. And this is something that also comes out from this experiment that we did with self-promotional articles. You know, I had to generate mass generate uh content for this for this um experiment. And I realized how you know bad everything is without the right context. Right? So source of truth is that con is that context and it's really helpful when you generate content with uh inside the agent or apps. Um how this looks is it's like this. So you it's an app where you and actually uh let me go uh back real quick. You can actually click that link that link that takes you it can actually show that link to your agent and it will read the GitHub um instructions to recreate that in your agent A and you will have an app like this and um so what you do with this is just you give it an URL with uh um important information about your product or your brand. It can come from like a blog post or a help article. You can paste in text. You can even give it ped PDFs. And what it what it does then is um it analyzes your content based on the pre-installed schema. So that's that's what uh makes it different for from like a you know like a Google Drive with your um content, right? Because there's that pre-installed schema that puts things into four categories. Insights and stats in my example concept and mechanism, product info and HFS how-tos and how I use it. Um I can say something like what do SOT source of truth say about AI overviews impact on traffic and there I have it right it's it's my data it's not like you know alternated or or taken from the web it's just my data data that I trust and I can use that uh to um well for example to create an presentation for you guys. So this prompt is I need to show my audience what why Reddit is so so important for AI visibility. What does the SOT say? So that graphic about Reddit that that you seen before that is oneshot agent creating creating this infographic from the SOT files. Um and I guess yeah that's it. Thank you. Yeah. So if you guys don't know like uh many me uh members of our marketing team helped to create all these skills they uh that they use themselves to improve their marketing work. So, if you guys are using agent A not tried it, there'll be a bunch of pre um built skills that are I mean actively used by our team to do the marketing that we do and I think we're pretty uh successful here at HF. So, I think you guys have pretty uh amazing skills to to like take from and and modify it for your own needs. Uh what is agent A? We'll talk about it more at the very end of this. But yeah, I mean that was a really powerful uh first round by M uh I mean round two use cases three and four from Matias with coming in hot with his AI powered marketing tactics. But can Louise fire back with equal force to rival Mateos? Louise, the floor is now yours. Show us what you've got prepared. Okay, here we go. I'm back in the ring. [laughter] Okay, hopefully you can see my screen. Um so my next use case uh also built by agent A is something that can help you find sites with traffic growth and decline. I actually shared this on LinkedIn like an earlier version of this app. Um but I essentially built this app so I could look at thousands of sites in one go uh and compare their traffic and assess the impact of Google algorithm updates or find examples to share on in blogs and on social media. uh assess the impact of programmatic and AI content and also compare traffic share because although like organic traffic is dropping across the board, it is relative and it's still very useful to compare and benchmark yourself against others and also kind of take tips from their strategies. So, uh I've recorded another video um and this is growth scanner and I'll walk you through it. So, Grow Scanner essentially tracks different categories, each one containing up to a thousand SERs, and it shows you whether sites within those SERs have been in flux since the previous month. So, for example, when it comes to AI SEO, 29% of sites have moved over 20% in the last month. And that kind of flux is usually indicative that the sites or the c the sites within that category may have been hit by an algorithm update. So, this kind of dashboard can show you which categories have potentially been hit by recent updates, like for instance the May core update. Uh, and to set a category up, you uh just enter a seed keyword. So, for example, I've gone with flights here and then you can enter related keywords if you want to expand the topic further. Uh, and then you select a country and the number of domains and then hit scan. Uh, and you should be able to see the flights category. Yep, it's just added there. But I'm going to dive into AI SEO. Um, so this category is based on just 170 SERs because it's quite a new cluster of keywords. So it's a bit smaller as it stands. Uh, and by default the data within the table there is sorted by a combination of traffic share and keyword coverage. But I I've also added dual sorting. So you can find high growth sites if you hold the shift button um, and then click on the 12-month column. And now you can see Reddit's no longer top. It's rankability.com. So, you're essentially sorting by traffic share, keyword coverage, and 12-month growth all in one go there. Um, but you can obviously sort by different growth periods, too. So, there's a one month growth there. Uh, there's also 3 month and 5year growth. Uh, but I'm just going to switch those off for now. Um, and yeah, I didn't build Growth Scanner just so we could see the impact of algorithm updates. As I said, I also built it so I could find examples of growing sites or declining sites that we could use in blog posts and on social media. So, for instance, if I click into easel AI on the table here, um I should be able to see the traffic trend over time in more detail. Um and then I can also see algorithm updates. I can switch those on and off. Um, and then I can also download this as a PNG and that will show me a fully designed chart that I can drop straight into a blog post. Um, and then if I scroll down here, I can also see pages in relation to traffic over time. So in this example, it looks like Easel AI have recently generated loads of pages and that coincides with their traffic spike. Usually that kind of traffic doesn't last too long. And then I can also download that as a PNG and that will show me uh traffic and pages in a designed format too. Uh so coming back to the table I can also see traffic share and keyword coverage here across the whole category. I can look at sites that are rising or reversing down or falling or flat and I basically can just assess the whole traffic impact. So essentially, Growth Scanner is where I can see traffic trends across tons of sites all in one place and use it as a launchpad for digging deeper into the strategies of winning uh winning sites. So yeah, hopefully you found that interesting. And now I'll show you my next use case uh which is the refresh engine. And the whole point of this is that so uh you can fill topic gaps uh with unique AI content. So um it's actually built on HRS's AI content helper and it helps you fill those gaps uh with unique and relevant insights from subject matter experts. So it's based on my saved posts. So um this is the refresh engine and you basically choose an article to update that you suspect has declined due to a content gap. Sorry about that. And then you add the URL um and add a focus keyword and then click create and run an audit. I've chosen programmatic SEO. And it's just going through the competitors here. You can select your competitors. Uh and then you hit run the audit. And as you can see, we're scoring 76 out of 100 here. Um um but when it comes to pros and cons, we have a topic gap there. So competitors are beating us there. So, if I scroll down to suggested actions, I can see ideas of how I can fill that gap. But I don't want to just fill it with generic AI content to improve the score. I want to generate it from scrapbook. And scrapbook is basically where I save my LinkedIn posts, articles I've written by subject matter experts that I found interesting and so on. So, I can generate the draft from the scrapbook. And wherever you see a square bracket there, that represents a post. So, this is from Chris Long and it's about um thin content in programmatic content. So, if you click accept, it will drop you down there and you it will show you where it's added in that section of content and you can move it around manually if it's not quite in the right place and you can vibeed and um then you can hit rescore with AI content helper to see what impact that gap has had. So, as you can see, we've added three points here uh to our overall score. Um, and it's added in links to those posts. And then you can also check which individual parts of the uh content gaps that it's filled. Um and so this is just showing that I've moved up uh I've moved that content to a different place. And now I've hit publish to WordPress. Um and this essentially publishes straight to WordPress. Agent A can publish straight to WordPress. And then it will show you highlighted where that content has been put within the blog post. And then you can preview it. and it's added in all of the short codes that we use at HRES. And when you're happy with that, you can just remove those yellow um uh yellow placeholders. Sorry, I think the video sped up really quickly there, so I was really trying to fill the time, but hopefully you found that interesting. Um and yeah, I'll hand back over to uh Constants now. Oh, Constance. was talking while muted. My bad. So, uh I I was just saying that like for the use cases that you guys see with agent A, they were built with AI, but honestly, it's not like specific to just this one platform. They can be built as long as you saw the workflows that were happening today or even like Matteo showcased his GitHub. So you can go to the link after afterwards when we share the slides to see what he's built and then ask your AI platform of choice to build something similar to do the uh like you can just reference the GitHub link and um build something similar for your own so you guys can have the same workflows happening and they're all using a number of these are using like um functions from the HS platform they're just piped in via connector or via API. So anyway uh as we go ahead both of you guys have shown some absolutely phenomenal use cases. First in Href specifically and now also in AI. So uh specifically agent A in this case, but like we're talking about AI use cases. Now it's time for the audience to decide whose use cases was more powerful. It's time to place your votes. You have 30 seconds starting from now. Here we go. Wow, these battles are so tight. It's so cool. Really really not sure who's going to win today. Okay, the last votes are coming in. Are you ready? I'm closing the poll in three, two, one. Poll closed. All right. So, let's check here. The winner is for All right. I'm just going to quietly say we're not going to talk about who um the winning votes for round two. We're going to reveal at the very end just so you guys don't cheat know who the who the winner is going to be. Okay. First round is Okay. All right. Let's move on. Final round. This is it. Both our marketing titans today have everything powered up to deliver their final strike against their opponent. Each fighter will be presenting their best, most powerful use case. And once again, Louise, I know you just presented, but it will be time for you to share your final attack. Please go ahead and take it away. Okay, here we go. Okay, so this is my final knockout punch. hopefully. Um, I've mentioned it before. It's scrapbook. Um, and it's about I basically fuse it with Ryan Law, my manager's one-click blog pipeline to weave in my posts that I've saved that I found really interesting from subject matter experts, research studies, um, yeah, articles, and I built it in agent A. I actually started building it in Lovable and I've moved it over to agent A. And yeah, this is it. So, this is Scrapbook. It's an app that saves my LinkedIn posts, um, my favorite LinkedIn posts, um, and all my favorite articles that I've read. And then media. So, that's PDFs, images, and, uh, YouTube videos which get transcribed. And I essentially grab LinkedIn posts from my LinkedIn saved folder. So if I open the post, I will use um Obsidian's web clipper extension which grabs the post content and comments and saves it and then I can save it as markdown and then I basically go back to scrapbook upload that markdown uh open it and then as you can see hopefully it should load there. So there's that post by Chris Long. Um and then for articles I just paste in the URL and for the media I upload it. And from there I can go ahead and tag those posts. So each one gets relevant tags and then I can search in the search bar. So this is me searching for Mark Williams Cook. I can also search topics. So for example, say I wanted to search for the topic of markdown. Uh I can see posts on markdown. And this is the interesting part. Then I can start selecting for ideas. So I'm going to select this markdown post by Mark Williams Cook and this one by Lily Ray. And then I can head over to the URLs tab and I can search markdown there. And I've got this article from Jonno Aerson, which is great. And then from there, I can either generate headline ideas or I can send these examples straight to Ryan's blog pipeline that it's Ryan's blog pipeline, but I've amended it to my own uh specification. And then uh I'm just going to look at headlines for now. And it essentially gives you a creative headline and an SEO headline. And the creative headline is uh focused on generating unique angles. And the SEO headline is based on relevant keywords sourced from HF's keywords explorer. Uh, and you can go through all of those options and see which ones you find most interesting. And when you know what you want to write about, then you can just hit the pipeline button. And it essentially roots you to my version of Ryan's blog pipeline. Uh, and it will add in the information. It will include the target keyword, the my author style, and the article synopsis. Uh, and then it will supply the post that you chose. So there's Mark's uh LinkedIn post and then beneath that there's Lily Ray's LinkedIn post and then Jonno Aerson's article there. And from there you just hit start pipeline. Um and it will take you through each stage of the pipeline there from keyword research to research references and then down to preview. And this is what a preview looks like. So it will essentially generate the content. Um it will generate diagrams as well. And then um you should be able to see the Lily Ray quote has been added in here. And then further down it's actually embedded a LinkedIn post from Mark Williams Cook. Um and then right at the end you can see Jono Aerson uh and his article has been referenced and written about. Um, and yeah, that's essentially how I weave in interesting examples and actually unique content and posts and studies and research into my AI generated content to make it fresh and uh, actually worthwhile reading and not just like the homogeneous AI slot that we sometimes see online nowadays. And there we go. That's my last use case. That is Louis's final punch. And honestly, I really really like what she's built with Scrapbook so much that I've also used it myself. It's just like knowledge management is just so powerful when you want to create amazing content out there. And that what will Matosh have held back so far to make his knockout punch? Mat, now is your time to take it away. The final presenter for today's showdown. All right. Thank you. Okay, there we go. Uh the uh last one would be to train agent A on great content so it can do more for you. So this is also uh the second part of nurturing your agents um where the uh SOT the source of truth will be the the u knowledge that the agent has. Now this is about skills. So, uh, skills are usually, you know, uh, already pre-made. Um, when you find them online, they are, you know, made to tested and and and, um, created out of an idea. U, but you can actually create a skill just from from anything interesting that you saw online. So, for example, if you saw a cool guide online, you don't you can just show it to your agent, in our example, agent A, and it will create skills out of that um out of that content. So, you don't necessarily need to have like a, you know, welldesigned um uh and structured skill. The agent, the AI can figure it out uh itself. And this is basically how I keep up with everything that Louise uh publishes. So like I said, Louise, I said I I will have a surprise for you. Um so for example, this uh amazing uh guide on how to moni monitor your YouTube video mentions is just amazing. It's great, but it's too much clicking. So I mean, look at that. It's just super long. and you know, click here, click there. Um, I'm too lazy for that. So, so what I do, I just uh enter a prompt like this. I need you to check out this article and develop a skill. Sorry about that. And develop a skill called YouTube mentions. Whenever I invoke it, ask me what specific analysis I need. And that's what it did. Um, let's try to actually try that. Um, so if you haven't seen agent A, this is how it looks. Um, you have your chat in the center right here and then your whiteboard with everything that the agent creates for you, apps, reports, um, um, articles, stuff like that. So, I can now that I've um showed my agent how to develop that skill, I can go and say uh need my YouTube skills. Uh now, please works. I think it will right there. We have it just as planned. So, uh, right now you can see that the agent asked me to, uh, pick an analysis that I want to run, um, analysis straight from Louis's, uh, blog post. Um, yeah. So, that is how you, you know, achieve come closer to that, you know, idea of having like a super smart assistant, having a second brain and all of that. you can just train the agent from content from you know smart people. Um yeah thank you. That's it. Okay and that is it. That is the final knockout from Mush. All right. Let me go ahead and share my It's the final round of voting. What a battle folks. We had like one person share about how to bring in the best content. uh around the web into their like second brain and the second person uh showcased about how to bring in the best skills uh from around the web into their you know like AI powered workflows. So really it's like two sides of the same like uh new workflows that I think all marketers can definitely benefit from in the age of AI. Uh but now the fate of today's showdown rests in your hands. It's time to cast final vote. Whose final use case hit the hardest? You have 30 seconds starting But you guys are just [laughter] making making it making it so close to the finish line. I've don't we've never had such a close battle between uh two people on the Yusuke showdown. Okay, I'm gonna wait for five more seconds so that people can fill in their votes. Four, three, two, one. [gasps] What do I do now? so the voting is now closed. Ladies and gentlemen, we have witnessed tonight nothing short of extraordinary. Both fighters held nothing back today in today's use case showdown and they demonstrated why they make unbeatable marketing content that has helped us build the audience here at HRS for years to come. uh I say here that only one can emerge victorious and one has barely marginally emerged victorious. I mean by like a number of votes. So thank you for everyone who has contributed a score because it could not be closer today. The audience you guys have spoken. The winner is Louise Linhan. And with that knockout blow, the battle is decided. [clears throat] Your winner Thank you. [laughter] Congratulations, Louise. You've become the third H showdown champion. How does it feel? Please share a few words. Oh, it feels great. And now I'm gonna have a bit more money in my pocket to take this image to my hairdresser and get [laughter] beautiful hair because AI makes great. [laughter] Any any words for Medouch? Um, well, I mean, the fact that we were so neck and neck, it just shows that, you know, we were both either equally great or equally bad, but I'm going to go with equally [laughter] great. I think you guys are equally great. It's awesome. No, really. You guys make such awesome stuff about how to do marketing differently. I think it's really good. Anyway, go ahead. No, that's all I had to say. Um, yeah. Okay. Um, great. That's all right. And Mat, um, you fought hard tonight, but you were like edged out just by a number of votes. I mean, it's just showing how amazing like the value you guys both shared tonight. Do you have any final words for Louise here? Yeah, I think you know well deserved that uh the scrapbook app. I think like half of the team at have that right. Yeah, I mean I I have a scrap of and that's and that's that's always inspired by you know Louis's idea. So yeah, great stuff. Thank you. Awesome. And as pledge Mat will be donating $500 to Louisa's choice of charity and she'll be announcing that or at least Mateo will be announcing that on social media so look out for that soon. Let's give a round of reactions for Louise and Mat. All right. All right. Now, let's wrap up here. Oops, that was saved the materials one. [laughter] All right. Before I close out this event, a few announcements. If you want to relive the magic, I will be uploading uh today's showdown on YouTube very soon. So, I mean, it's not exactly a live stream, but um you will see the replay and you'll be set. Everyone will be emailed the replay uh whoever they have registered. So you guys can rewatch again like right after uh this uh use case showdown ends. You watch it and share it and send it to anyone who needs to know what uh needs to see what went down tonight. And uh secondly, when you close out this webinar, you will receive a feedback form. Please take a moment to share your thoughts about what you saw today. Your feedback helps us make events in the future and showdown um rounds in the future even better. And who knows, it might influence just who we bring into the ring in the future. So, if you also have people you like to pitch to see here on stage today, please also share that in your feedback. We all like to um take those votes in. It will also help us pressure those people to come. Ha. All right. And before we leave, of course, we did mention a lot of uh different use cases from agent 8 today. So, if you're already a subscriber at Hrefs, you would already see a free trial. So, you can go to hs.comagent-a or scan this QR. Wait, yes, scan this QR code. Uh, and you you at the top right corner, if you're already an HFS subscriber, you'll see like a free trial that you can start. Uh, if you've already used a free trial, of course. I mean, it's it's a one month free trial that goes once. Uh, otherwise, you can give it a try. And, um, we are on social media, I think, giving away lots of agent a free trials every so often. So, please do watch out for that as well for our next giveaway. We're really trying to get as many marketers as possible to try our platform. So, please watch out for that as well. And finally, I want to thank uh take a moment to thank our incredible uh competitors today, Louise and Mouch. Thank you for giving us your all. The audience will be walking away with battle tested strategies they can use to create like amazing content and not AI slop and build amazing marketing automations. And of course I would like to thank uh you everyone in the audience today for joining us. I hope you've come away with insightful ways you can use HFS and agent A or just any like AI platform of your choice. Please look forward to the next live session and use case showdown. And thank you so much for making our community the way it is. All right, guys. This has been the HF's use case showdown. It has been a pleasure to be your host today. Until next time, keep building, stay hungry, and we'll see you in the arena. Goodbye. Byebye. Thanks, guys. Thanks for joining. Bye everyone. Okay. All right, guys. See you guys later. Take care, everyone. Bye.