Email is somehow STILL getting slept on right now.
Yes, social is great.
(I have a YouTube channel with 500k subs. So I get the value)
But those followers are rented. One algo change and your "audience" is gone overnight.
The email list is the only audience you own. There isn't even really a close second.
More in the clip.
Boring but true: Sites that perform well in AI answers almost always have sound SEO fundamentals.
I'm talking about stuff like: URL structure, quality content, real backlinks.
Full explanation in the clip:
AI changed B2B content.
Generic "what is X" articles are pretty much dead. Claude answers those questions now.
What still works is content matched to a specific decision stage:
-educational content with unique data AI doesn't have
-comparison pages for people weighing options
-case studies for people about to buy
At Exploding Topics we build all three.
Here's what each one looks like.
Rand Fishkin's team at SparkToro analyzed millions of devices to see how people search now.
The finding that stood out to me: Google's share of search is significantly smaller than just a few years ago.
And more people are searching outside of Google: Amazon, YouTube & ChatGPT.
As Rand put it in the report: "search is a behavior, not a channel."
Here's my takeaway for anyone doing SEO right now.
Here's a huge AI SEO mistake I see a lot of SaaS sites making.
LLMs need complete, ACCURATE information if they're going to use that information as part of an AI answer.
Which means you don't want your site full of hyped-up marketing copy like, "We're redefining the CRM category."
Basically, just be explicit about what your product does, without a lot of hype or vague promises.
For example, at Exploding Topics, we have several dedicated pages outlining who our product is for and what we do.
We even have a few pages dedicated to our product's key features.
And if you don't have this kind of information clear on your site, LLMs will use other sources to fill in the gaps, including speculation, nasty Reddit reviews, and guesses based on what your competitors do.
And that stuff is not usually very flattering for your brand.
But when you have content on your site that describes exactly what your brand does and who it's for, you're likely to get in front of people searching for your category in AI search.
I'm very excited to be speaking at @nypl to discuss how to rank in SEO and LLMs.
Register here: nypl.org/events/programs/202…
2
7
1,456
Exploding Topics = early trend detection. Backlinko = actionable SEO strategies.
Neither of those associations is an accident.
The first step of ranking in LLMs is choosing ONE hyper-specific concept for your brand.
(And it can't be a broad category like "CRM software.")
It has to be specific enough to show up across hundreds of sources naturally.
Here's how to do it:
LLMs really focus on WHO you are, not WHAT you publish.
This is super different than SEO.
ChatGPT and Claude basically build a mental model of entities: brands, people, tools, and concepts.
Show up alongside the same concept enough time? LLMs will associate YOUR brand with that concept.
That's why Exploding Topics became the default answer to "best tools for finding trends" and thousands of other AI prompts.
How it works in the clip.
A mistake I see a lot of founders make with content marketing is trying to scale too soon, too fast.
They decide that they need to publish a new YouTube video every week, a podcast every single day (or in B2B SaaS) thousands of articles that target long-tail keywords.
But that's not what we did at Exploding Topics. Instead of pumping out tons of content, we decided to focus on creating great content that stood out.
For example, our writers would spend about 20 hours creating detailed reports on trends in industries like e-commerce and health care.
These were deep dives, data-driven reports that were generally useful and unique.
Yes, these pages brought in traffic, but more importantly, they brought in high-quality leads, trials, and customers.
If I had just followed that generic strategy of creating hundreds of posts targeting long-tail keywords, it probably wouldn't have worked. Because today, scale isn't a competitive advantage anymore.
With AI, you can push a button and produce 500 articles or YouTube video scripts. The advantage comes from actually creating something that people want to see and share.
One page on our site has been cited by Forbes, Fortune, scientific journals... even the UK government.
It's not some multi-million dollar industry study.
It's a simple page that tracks how many users ChatGPT has, updated every time OpenAI drops a new number.
Here's how you can do something similar for your site:
The way people search has COMPLETELY changed.
Today, search is more like a loop through a bunch of different platforms.
-Google to understand the problem.
-YouTube to see solutions.
-Reddit to find out what's actually good.
-ChatGPT to decide.
More details in the clip:
3
8
1,764
Two keywords that Exploding Topics currently ranks for:
"how to find trending products" and "tools for product research."
They look pretty much the same.
But they're actually SUPER different.
The first person is still sort of learning about the space.
The second is comparing options and about to pull out their credit card.
In the clip I walk through both keywords and how we optimized for them.
1
6
1,679
There's exactly ONE source that you fully control: your own website.
And AI systems train on it more than a lotta people think.
Even when ChatGPT pulls information from elsewhere, it often cross-checks it against your site.
So if your website is unclear or incomplete, ChatGPT will rely even MORE on third party sources that you don't control.
Here's what we do about it at Exploding Topics, including the two tasks we do at least 2x/mo.
AI has made content a commodity.
What's left?
As it turns out, plenty:
1. Perspective
AI can only remix what it knows. It doesn't have original takes.
In fact, one of my best-performing recent LinkedIn posts (217,550 impressions) was my perspective on the future of SEO.
2. Experience
People want to hear your war stories.
Positive. Negative. It doesn't matter.
The key is that you share what HAPPENED.
3. Receipts
Real numbers from real projects:n"Here's what happened when I tried LinkedIn Ads", "How I got 100 IG followers overnight".
4. Taste
AI produces infinite content and code. Knowing which one is actually GOOD is scarce.
5. Relationships
AI can't grab coffee with an editor, DM a founder, or get you on a big podcast.
The biggest breaks in my career came from a human knowing me.
6. Skin in the game
Aka stakes.
AI can make a prediction. But only humans can bet money, time or energy on something.
👍🏻
We asked our customers a simple question: how did you find us?
The answers surprised the crap out of us.
A big chunk of our customers were investors, who found us searching for "alternative data providers."
This is a world we honestly knew NOTHING about.
So we built content for exactly those searches, and it brought in a pipeline of new customers.
Full story in the clip.
A lot of people romanticize the broke founder eating ramen.
My path was a little less dramatic...
When I started Backlinko, I had another site quietly paying the bills.
(A personal finance site making around $5k a month).
That boring little site gave me a huge advantage:
I didn't NEED Backlinko to make money right away.
In fact, I spent a full YEAR building my audience before launching a product.
Plus, I could experiment and fail.
Compare that to the ramen founder: they have 3 months of runway and desperate for everything to work.
Now, to be fair: some people genuinely like that pressure. The whole "burn the boats" mentality.
If that's works for you, more power to you.
But in my experience, desperation makes you worse at business.
Specifically, you have to make SHORT TERM decisions.
And the key to success in business is playing the long game.
So my advice for anyone starting out? Keep your job (or income source) as long as you can.
👍🏻
11
28
3,957
A survey of 1,000 B2B SaaS buyers: HALF now start their research in AI tools instead of Google.
87% say AI changed how they pick vendors.
So yeah, if Claude doesn't mention you, your buyer is probably gonna see your competition first.
(Even if they were already considering you)
The good news is that fixing it doesn't take any special skills.
It's mostly marketing fundamentals tweaked for AI SEO: a clear site, consistent third-party descriptions, honest comparison pages.
Overview in the clip:
I've built and sold two multi-million dollar businesses. And I'm not particularly smart.
What I am is a good copier.
Let me explain before you thumbs-down me into oblivion:
Almost nothing I did to build my two companies was an original idea.
My most famous SEO strategy (the Skyscraper Technique) is literally "find content that already worked and make something better".
I just gave a known technique a small twist and a unique name. And that was enough to stand out.
Even Exploding Topics wasn't my invention.
One day someone sent me a janky tool a developer had built. It was exactly what I'd been trying (and failing) to make myself.
So I bought it.
Then we improved it to the point where it was acquired by a public company.
Here's my problem with original ideas: they're risky. You're investing time and energy into something no one may want.
But when you copy... demand is already proven. The only question left is whether you can execute.
Meanwhile, many smart founders I know struggle to get their ideas off the ground. They're interesting. But no one really cares enough to buy.
So no, I'm not the guy who's going to invent a new category.
I'm the guy who notices what's already working and asks: what would the 10x better version look like?
Feel free to copy this approach too. I obviously won't mind, LOL.
👍🏻
Our newsletter grew to 155,000 subscribers and drives about 25% of our ARR.
It all started with... talking to customers.
When we asked where people they find trends, it wasn't mostly social media.
The conversation was happening in newsletters, on platforms like Substack. And we didn't have one.
So we started Exploding Topics Tuesday. And it transformed our SaaS business.
9
14
2,361



