🧠AI SEO 2024 | Embrace Artificial Intelligence | James Dooley AI🧠

Kasra Dash: Welcome back to another video. Today we’re going to be talking about AI, SEO, and our strategies. I’m joined with James Dooley, as we can see over there. Pleased to have you on, James. I think you should probably kick this one off – what is your content strategy for using AI?

James Dooley: Obviously going into 2024 now, AI is becoming a massive part of everything we’re doing within our agencies.

We’ve made certain that we’ve not laid any member of staff off – it’s there to supercharge every single member of staff within the business. The output on loads of different things is incredible. On the content side of things, we’re both investors in AutoBlogging. Some people could say we’re biased because we’re investors, but I only invest in brands or companies that I believe are the best – and if they’re not the best at the time, I want to make certain they become the best. I’ll keep going and going until we improve quality control. AutoBlogging now, with Godlike Mode, is incredible. Even the bulk mode is really good. That was a painful two–three weeks of going back and forth, but the prompts now are amazing, and the images are great as well.

Kasra Dash: So on AutoBlogging, you’re mainly using it for supplementary supporting documents, right?

James Dooley: Yeah, mainly for the supplementary, supporting blog content. And then we’ve got Topical Titan coming out and Content Veggie, which are premium versions of AutoBlogging – so that’s where the human editing side of AI content comes in, with more advanced prompts.

We’re now using that for money pages, which I honestly never thought I’d do. Roll on six weeks and I’m using it for money pages.

Kasra Dash: One question people will probably have is: are you using AutoBlogging right now for content on money sites, or just on non-money parts of the site?

James Dooley: I’m using it on money sites, yes – but mainly for blogs. For actual key money pages, you might have an editor spending more time refining what AutoBlogging outputs.

Godlike Mode is a one-off generation rather than a bulk mode. The prompts in there to write a 2,000–3,000 word article can be up to 10,000 lines of prompts behind the scenes. The amount of OpenAI usage is huge – I think it’s about 50 times more OpenAI usage now than we were doing two months ago, because Scott’s been absolutely battering, verifying, and improving it. Now it’s got to a point where, with GPT-4 – and obviously GPT-5 coming out this year – it’s just insane. I can’t believe how much better GPT-4 is compared to 3.5. It’s only going to get better again. And to be fair, it’s not just that AI has got better – we have got better at prompts. Every single one of the team has been trained to be better at AI prompts – prompt engineers. As they’re doing that, and we’re prompting in the correct manner, the output is incredible.

Kasra Dash: So that’s your content strategy. When it comes to images, what are you using there?

I mean, myself, I’m just using DALL·E within ChatGPT and stuff like that, but Midjourney’s got a lot better. You’re absolutely crushing it with YouTube thumbnails and any images you’re doing in the background and BL. You’ve tested everything – what’s your favourite now?

James Dooley: If I’m out on the fly, I love using DALL·E. It’s really easy – you can even get it on your phone now, and it’s part of the ChatGPT interface.

If I’m doing something a bit more bespoke – for example e-commerce product images, and I want them to be a bit more unique – I’ll use Adobe Firefly .You can change the background, add a wood texture, and make the image really nice. You know yourself – unique images rank way better than the same stock image that’s been plastered all over eBay, Amazon, and everywhere else. Making images a bit more unique helps a lot.

Kasra Dash: One question I’ve got for you – you’re a firm believer in DALL·E. I see you using it almost every day. Do you ever try to make that DALL·E image a little bit more unique by adding it to Canva, adding some text and stuff like that?

James Dooley: Yeah, all the time. I always put it into Canva, and if I can, I’ll put some wording over the top, if it makes sense. I’ll add a heading or the logo on there and things like that.

I’m a massive believer, like you are – the split testing we did on unique images was incredible. The results we got from you using Adobe Firefly where you took a stock product image and made it unique were out of this world. So, with what you can do now with AI, one important thing you’ve just mentioned is you still have the human interaction. Whether it’s you doing it or a member of staff, you’ve still got that staff member – it’s not like you’ve just got rid of all your graphic designers. If anything, it’s speeding up the process. Instead of them being able to output 10 images a day, they might be able to output 50 images with prompts.

I keep reverting back to the team: if you don’t embrace AI, AI is going to take your job. It’s really as simple as that. You’ve got to embrace it and get it working for you. AI is your VA.

If you can get AI as your virtual assistant, it can supercharge your output – whether it’s content, images, videos, whatever. The output they’re doing is four, five, six, sometimes ten times more. On the content side, they can do ten times more content in a single day than they were doing 18 months ago.

Kasra Dash: I was reading a blog recently and they basically said ChatGPT is almost like your personal assistant – and you should use it as your personal assistant, definitely.

Right – that’s the image side of stuff. Now let’s talk a little bit more about video, because I know you’ve been using AI for video. You’ve literally been sending me loads of examples. What’s your go-to AI video tool?

James Dooley: The go-to video tool at present is InVideo AI. It’s absolutely incredible. You literally give it a prompt –

Kasra Dash: Is it an app or is it a website?

James Dooley: It’s both, but I’ve been using the app. I’ve just been saying, “Can you create me a three to four minute video on this concept?”

The stuff it comes back with is out of this world – really, really good. I’ve got three or four high-level videographers, and I sent it through to them. They were like, “Who’s doing these videos?” I said, “Me – I did it in two minutes.” They didn’t believe it. InVideo did everything, including the voiceover. I can change the voiceover, change the script, choose whether I want the voice from the UK, California, Miami – whatever you want, you can change the voiceovers. What I’m looking to do in 2024 is, as a massive believer in embracing AI, I’m going to go down the route of faceless video brands and scale that out. I’m going to create a brand that does faceless videos for others. Obviously we own a lot of rank and rent sites. These people – carpet cleaners, roofers, plumbers – they don’t really give me many videos to put on the site. But I want effective frequency – I want to be seen on social media, ranking videos, sharing on Twitter, Facebook, etc. So it’s going to make it easier for them to scale on the video front. I’m going to create some faceless video brands this year as well. And for you to talk on the image front – we might be looking, you and I, at founding and setting up an AI image company as well. Then moving on from that is the topical maps. Why don’t you tell them about how you’ve been using AI for topical maps as well, because we’re creating a new brand coming out called Topical Titan. It’s going to come out in the next six weeks, and what that can do – semantically relating topics and splitting different intents – is unbelievable.

Kasra Dash: Let’s take it back a step, right, because you’ve just dropped in “semantically”. You’re using big words – I don’t like big words!

Basically, the whole purpose of Topical Titan – and I think you’d agree – is: when you get a topical map for your SEO website, usually they just spit out 300 or 400 or 1,000 articles. Sometimes it can be a bit of a mess and you’re like, “Where do I start?” The whole purpose of Topical Titan is that we first do the actual topical map of your website – but then we say, “Whoa, whoa, whoa – let’s just hone in on this subject.” We’re planning to make it a lot easier for people to understand. There are guys like Koray out there where, when you speak to him, it’s like speaking to a human calculator. But the purpose of Topical Titan is for somebody looking to set up a new affiliate website or a new SEO site in general – we’re trying to make the topical map more understandable. We’re going after a lot of easier keywords first, so you’re essentially winning on secondary keywords and then slowly building up to rank for a monster keyword like “male running shoes” or “female running shoes”. There are a few of us involved in that, but as James mentioned, it’ll be beta launched to a few private people in the next six weeks.

James Dooley: I liked that you were talking about it then. I’m going to drop in a few big words like “semantically related” – that is the correct term. Some people say we’re overcomplicating it, but we’re not.

The way AI is able to parse different intents now from large language models – it can tell you, even if there isn’t a separate SERP result yet, that a search term needs a different page. Everyone just seems to reflect what’s already in the SERP and do copycat content. In 2024, a lot of people and podcasts have been talking about Information Gain. Topical Titan is going to be there to give you information gain at the topical map level – forget just Information Gain on a single page, adding a few words or FAQs. We’re going to find whole articles that are missing from your site, or things you’ve only touched as an H2 on a page that need their own page with their own search intent. It’s like topical mapping on steroids. It’s not going to be Koray-style breaking down every single micro detail for content briefs – it’s more for someone starting a new site: “Here’s 350 pages you need to publish.” They might not even have Ahrefs or Semrush – or even if they do, they might not fully understand clustering keywords into the right pages.

Kasra Dash: I just want to go back a couple of steps because some people might not know what Information Gain is. Correct me if I’m wrong here:

Let’s say LG releases a new 50-inch TV tomorrow called the “LG James Dooley 45”. If you have that on your technology website today and you’re the first website to have it – that’s information gain, right?

James Dooley: Absolutely right.

Kasra Dash: So it’s new articles like that, but also times where, for example, people might have a page in the gambling sector on iOS betting apps, but they haven’t created a page specifically for iPhone and iPad.

There’s enough information on iPhones and iPads – different models, resolutions, etc. – to justify different pages. You might even have a different order of recommended betting apps for iPad vs iPhone. A lot of people just do “iOS betting apps”. If the large language models inside Topical Titan say, “You can do a different page on this,” that’s not just using the SERP to decide it. Sometimes Keyword Cupid, Cluster AI, and other clustering tools can be a bit obsolete when you start using AI and understanding there’s a different intent between iPhone, iPad, or iOS as a whole.

James Dooley: And one thing you mentioned that’s super important: on-page entities. For example, the resolution on iPhone and the resolution on iPad are completely different.

You can’t really have one page ranking perfectly for both because you need certain entities on the iPad page and different entities on the iPhone page.

Kasra Dash: Exactly.

Right – AI for SEO in 2024. We’re big believers. We’ve got the entire team all-in saying, “This is the future, we need to embrace it, otherwise it’s going to replace our jobs,” but actually it’s going to *supercharge* us. So, topic: AI link building within SEO. I like this one. Talk a little bit about the link simulator we’ve built in-house with AI.

James Dooley: The link simulator lets us evaluate which links are pushing the needle for your competitors.

We can basically filter out all the noisy links. You look at any site – there’s a lot of noise. Take Forbes as an example. Let’s say they have 200,000 referring domains (I’m sure they have more). Not all 200,000 are actually pushing power or helping them rank. First, we eliminate the noise. We look at the links that are pushing the needle. From there, we eliminate the worst 20–25% of links – the most toxic ones – and we hone in on the best performing 75%. So out of 200,000 referring domains, we might get rid of 100,000 and be left with 50–60,000 that we know are actually pushing the needle. Then you don’t run the risk of building toxic links; you’re not building noisy or zero-value links. Google sees a lot of links as zero value. The team have got so good with AI on the link building side. Have you got anything else to add? There’s another tactic on crunching data with AI in link building.

Kasra Dash: The biggest thing we wanted to do is move away from just looking at high DR dofollow backlinks.

AI is able to look at relevance – using NLP categorisation, semantic relevance – on both a page and a domain level. It can quickly crunch those numbers. The SEO tools in 2024 are becoming expensive. If I use Ahrefs to look at DR and I’m analysing tens of thousands of domains, every filter is an extra credit. Then you go to Semrush for spam score – which on its own I don’t love – or LinkResearchTools for toxicity and power. AI wants data – the more it can crunch, the better. If we can export spam score from Semrush, DR and traffic from Ahrefs, and power/trust/toxicity from LinkResearchTools, we can merge all that and then just use Google Sheets with AI as a link simulator. We’ll have unlimited filters – no “you’ve used another credit” every time. Then, for every single link being placed through the four different link agencies we own, we track whether links are going up and down and run them against backlink profiles to see what’s working and what’s not. We’ve massively expanded our blacklist of guest post farms. I believe the only other person with a big blacklist like that is Matthew Woodward – he was the guy who gave me the idea of making sure those blacklisted domains aren’t in our database. Now we’ve cleaned that up and we have lots of subsections in our database for what works in which niches. Having AI calculate toxicity, trust and power together has saved us so much time and made the output far better for all our clients. AI link building is huge – it’s in all elements of SEO now. One other thing we’ve started using AI for in Search is anchor text ratios. The team are exporting three or four competitors and, within ten minutes, they know what anchor text ratio you can safely get away with. So if you came to us and said, “I want some links for FatRank and I want to rank for ‘best SEO agency London’,” we can very quickly scrape competitors in that area, look at your current backlinks, and create super safe anchor texts. A lot of SEOs struggle with that – they over-optimise certain anchors or under-optimise and never push hard enough. Another area giving insane results recently is AI parasite SEO. We’re using Godlike Mode with AutoBlogging to create the content, spending 5–10 minutes editing it, adding a couple of images and call-to-action buttons, then posting it on free sites like Medium and LinkedIn Pulse. Then we hit those posts with a few tier-2 and tier-3 backlinks from SearchBro or other sources and within 24 hours we’re ranking for some surprisingly competitive keywords. Even people with no budget can do a version of this – they can ask ChatGPT to rewrite articles with the right prompts, upload to LinkedIn Pulse, and if it’s long-tail, it can rank just with social shares. So using AI for SEO in 2024 – these are incredible times, the best I’ve seen. People talk about 2010 when they could spam Google; I think now is actually the best time. I’ve never seen 24-hour results like we’re getting.

James Dooley: Definitely.

So, with regards to using artificial intelligence, if someone has a white-label agency – social media, link building, content – and they came to you and said, “Kasra, I want you to be my AI mentor,” what message have you got for those companies, especially if their staff are worried AI will put their jobs at risk?

Kasra Dash: I’d probably run an AI audit on everyone – staff-wise – and see what they’re spending the most time on, similar to what we’ve done in the office.

For example, if you’ve got a videographer taking five hours a day to cut up shorts and reels, why not use something like Opus that auto-creates shorts? Instead of five or ten reels, they can do 50 in the same time. Audit all your staff – not in a “what are you doing?” way, but to evaluate where time is being spent. For instance, we’ve not even talked about citations – lots of that is now AI-implemented. A lot of the contact forms you fill out with business info can be automated. That would be my first step. What would be yours?

James Dooley: Likewise. I’d say get a Python developer as one of your first hires if you’re going heavy on automation. Mix a Python dev with AI and you can do citations, link building for directories, spinning business descriptions every single time to make them unique.

With AI, that can be done in seconds – unique content per citation, easier indexing, more clout. If I was an AI mentor, it would definitely be about getting every team member on board. That’s the hardest part at first because they think it’ll replace them. But actually it should be supercharging them. If they understand that AI is going to supercharge their output – and if they become AI prompt engineers – it just makes their job easier. They can spend more time with their kids, take holidays, whatever – as long as the output is making the business money. If they get better with prompts, they’ll have more spare time. So I think they definitely should embrace everything they can in AI. Once they buy into it, they’ll be wanting to learn it over weekends and nights – they’ll be like, “These times now are incredible for scaling.” For us, like you said, we’re probably creating too many brands at the moment – an AI image brand, an AI faceless video brand, AutoBlogging with AI, AI topical maps – all related to SEO and supercharging staff while scaling. If we don’t innovate, we’ll evaporate. Someone else will come up and take over, and I will not have that. So we’ve got to keep driving forward, stay ahead of innovation, and see where it takes us. It’s going to be a very interesting year.

Kasra Dash: Yeah, definitely. Have you got anything else to add?

James Dooley: Not really, no. I think that’s everything. It’s just good to have a general chat solely about AI and nothing else, because sometimes you go off on tangents. If you’re talking purely about AI, it’s about making certain we’re leveraging it to its full potential.

Kasra Dash: So that has been the video about AI and SEO. Thanks for coming on, James.

James Dooley: Cheers.