You Keep Asking Us About AI Rankings. Here Are the Answers.

The same questions are asked each time differently, but always the same thing underneath.

How do I rank in AI?

That’s the big one. The one they all want answered. And the honest answer is that you don’t. Not the way you think you do. But we’ll get to that.

I decided to collect them. The greatest hits. The questions we hear so often they’ve worn grooves in our brains. So I’ll give them proper answers. Not the ones people want to hear, but the real ones.

“Can I rank number one in ChatGPT?”

This is always the first thing. They’ve been number one on Google for something, or they’ve been trying to be, and now there’s this new thing, this AI thing, and they want to know where they sit.

The answer is that there is no number one. There’s no list, no positions, no little blue link sitting at the top waiting to be clicked. That’s not how any of this works.

An LLM generates a response every single time. The same question asked twice might produce two completely different answers. There is no podium, and there never was.

What you’re really after is consistent mention and consistent recommendation. You want to be the name that keeps coming up when someone asks the AI a question in your space. You want to be part of its vocabulary, part of how it thinks about your category.

That’s very different from ranking. And it requires very different thinking.

“Does SEO still matter then?”

It matters more than ever, which is annoying for the people who were hoping they could skip it.

LLMs pull from web content. Heavily. The stuff they were trained on was web content. Your web content.

If your site is well structured, authoritative, technically sound and genuinely useful, all the things good SEO has always been about, you’re feeding the machine exactly what it needs to trust you and cite you.

Traditional SEO isn’t dead. It’s the foundation that LLM visibility gets built on top of. Not a separate discipline, not a relic. The floor beneath the new floor.

“How does the AI decide what to recommend?”

Good question. They never like the answer, though.

It depends on the model, but broadly, there are three things at play.

1. Training data is the first one. What the model learned from the open web. If your brand existed in high quality, authoritative content when the model was being trained, you’re in there somewhere. Buried, maybe, but in there. SO your past SEO efforts will have paid off here.
2. Then there’s retrieval. Tools like ChatGPT, Gemini and Perplexity search the web in real time. So your current content matters. Your freshness matters. Your structured data matters.
3. And then consistency. If your brand is mentioned the same way, in the same context, across multiple trusted sources, the model notices. It builds confidence around you. It starts to associate you with that topic.

None of this is a mystery. It’s just slightly different from what we’re all used to.

“Can I pay to show up in AI answers?”

Not yet. Not in any meaningful way.

There’s no fully rolled out Google Ads equivalent for AI. You can’t buy a slot. ChatGPT Ads are in Beta, Perplexity is experimenting with sponsored results,and others will follow eventually, but right now AI visibility is almost entirely organic. Earned, not bought.

Which means the brands investing now, before the ad products arrive, are building something their competitors can’t just outspend them on later. That’s a rare thing in marketing. A genuine first mover advantage that actually holds up.

I tell every client the same thing. The window for this is open now. It won’t stay open forever.

“How do I track this? How do I know if it’s working?”

This one’s fair, and the honest answer is that the measurement side is still catching up.

Any tool that purports to “track AI rankings” is lying to you.

There are tools developing and popping up every month. But the most reliable method right now is systematic prompt testing. You build a set of queries, the questions your customers actually ask, and you run them across ChatGPT, Perplexity, Gemini, Copilot, etc. Regularly. On a cadence.

Then you track whether you’re mentioned, how often, what the sentiment looks like and whether the information is accurate.

It’s manual and it’s imperfect, but it works. And it tells you things that no dashboard can tell you yet.

“What’s the difference between AI Overviews and ChatGPT?”

People mix these up constantly, and I don’t blame them.

Google’s AI Overviews sit on top of the search results you’re used to. They’re connected to Google’s index, so the SEO work you’ve been doing has a fairly direct line into whether you show up there.

ChatGPT, Perplexity, Copilot, etc, those are different animals entirely. They draw from their own training data and their own retrieval systems. The connection to your traditional SEO is real, but it’s less direct. More mediated. You’re influencing the outcome through a longer chain.

You need strategies for both. They overlap because strong authoritative content is always the common denominator, but they’re not the same game, and you shouldn’t treat them like they are.

“Is AI going to replace Google?”

It’s not replacing Google, but it is reshaping what search means.

We’re moving from click-based discovery to answer-based discovery. People used to search, scan the results, click a link, read a page. Now they ask a question and get an answer. Right there. No click, no visit, no page view.

That should either terrify you or excite you, depending entirely on how you respond to it.

The brands that become the source, the name the AI cites, the authority it paraphrases, those are the brands that win in a zero click world. Everyone else watches their traffic erode and wonders what happened.

“So what do I actually do? Right now. Today.”

I love this question because it’s the only one that really matters.

Get your brand information consistent across the web. Everywhere. Wikipedia, industry directories and authoritative publications. If the AI finds conflicting information about you, it trusts you less. It’s that straightforward.

Create content that answers questions. Real questions, the ones your customers type into search bars and speak into their phones. Answer them clearly, completely, with genuine expertise.

Build topical authority. Not keyword volume but real authority and depth. Be the source that other sources reference.

Use structured data and schema markup. Give the machines what they need to understand what you do and why you matter.

And get mentioned. Earned media, expert quotes, third-party references. Every mention on a trusted site is a signal that tells the AI you belong in the conversation.

None of this is magic. All of it is work.

“How long before I see results?”

That depends entirely on which flavour of AI you’re talking about.

For tools with real time search, things like Perplexity, AI Overviews, ChatGPT… improvements can show up in weeks. Maybe a couple of months. You make changes, the retrieval system picks them up.

For what the LLM actually “knows,” its baked-in training data, that’s a longer game. Model updates happen on cycles, months apart, sometimes longer. You’re planting seeds that won’t bloom for a while.

But the brands that start now will be embedded in the next training cycle. And the one after that. Compound returns of the kind you simply cannot catch up to once you’ve fallen behind.

“Is this just hype? Do I really need to worry about this?”

This is the last one they ask, usually while putting their coat on, trying to sound casual about it. Like the answer doesn’t really matter to them either way.

It is not hype.

AI-assisted search is growing fast, especially among younger users and in B2B research. The people making purchasing decisions are increasingly asking an AI before they ask Google. Before they ask anyone.

If your competitors show up in those answers and you don’t, that’s market share you’re losing in a channel you can’t buy your way back into. Not easily, not quickly.

The question isn’t whether to invest. It’s how soon. And the answer to that one is always the same…

Now.

These are the questions we hear every week. If you’re asking them too, you’re not behind. You’re paying attention. That’s the first step. The second step is doing something about it. Get in touch if you want an AI visibility agency.