What SEO actually is, and why it got harder in the age of AI

SEO has a branding problem. Most people picture spammy keywords and shady tricks. Here is the honest version: SEO is the work of getting the right people, the ones already searching for what you do, to find you instead of a competitor. That is the whole job. Everything else is detail.

It starts with a keyword, which is just a phrase a person types into a search box. But not all keywords are equal, and that difference is where the entire business case lives.

The money math: why volume is not value

Every keyword carries a rough dollar value, and you can estimate it with one line of math:

search volume × your share of the clicks × your conversion rate × what a customer is worth = revenue

The trap is chasing big numbers. A keyword with a huge search volume can be worth almost nothing, and a small one can be worth a fortune. The cleanest proof is the price advertisers are willing to pay per click, because nobody bids real money on a search that does not lead to a sale. Here is live data on three searches:

“chatgpt” is one of the most searched phrases on earth, and almost worthless to optimize for, because the people typing it are not buying anything. “crm software” gets a sliver of that traffic, but every click is a business shopping for software, so advertisers pay fifty times more for it. An opportunity is a buyer, not a big number. Good SEO chases value, not vanity volume, and knowing the difference is most of the skill.

The funnel, and what an SEO actually manages

People do not search once. They move through a journey, and the same person searches differently at each step:

Top of funnel is research: “what is a CRM.” Middle is comparison: “best CRM for small teams.” Bottom is intent to buy: “CRM software pricing.” Each stage is a different keyword, a different page, and a different job.

That is three keywords for one product. A real business has thousands, and the priorities shift constantly, this quarter you are pushing one product, next quarter another. The SEO’s actual job is to manage that entire moving map: which searches matter, which pages serve them, and where the money is being left on the table. It is closer to portfolio management than to writing meta tags.

Why it is hard: the black box

Here is the part nobody likes to admit. Google does not tell you how its ranking actually works. It offers a friendly framework, E-E-A-T (experience, expertise, authoritativeness, trust), as the lens it claims to judge content by. That is useful, but it is the guidance, not the machine. The real system is a stack of many algorithms layered on each other, and a leak of Google’s own internal documentation in 2024 suggested it weighs far more than it publicly admits.

Why the gap? Because Google’s incentive is not your success. Its incentive is your unique data and a great experience for the searcher, so it can keep its near-monopoly on search and the ad revenue that comes with it. “Make good content” is real advice, but it also conveniently means “give us more of the unique data we rank and monetize.”

You cannot get good at this by following the referee’s press conferences. You get good with data, craft, and a community of practitioners who test what actually works.

So real SEO is a mix of disciplines: a bit of code, marketing, writing, analytics, project management, and conversion sense. And it leans hard on a community, the practitioners and researchers who crowdsource what is genuinely moving rankings, because no single person, and certainly not Google’s PR, has the full picture.

Why it got harder: AI ate the top of the funnel

Then the ground shifted. People started asking ChatGPT the questions they used to type into Google, and Google, scared, rebuilt its results page around AI answers of its own. The research stage, the top of the funnel where most searches happen, increasingly gets answered without a single click to anyone’s website.

The scale is not subtle. “chatgpt” alone pulls roughly 94.6 million US searches a month now. That is where a huge share of the questions went. The informational click, the bread and butter of old-school SEO, is being absorbed by the answer itself.

The game changed underneath everyone. It is no longer only “rank the blue link.” It is be the source the AI answer cites. If ChatGPT, Perplexity, or Google’s AI Overview names your business as the answer, you win the moment, even when there is no click to count. That is a different craft, and most of the industry is still catching up to it.

So is SEO dead? No. It changed.

People literally search “is SEO dead” (a small but telling 1,800 times a month). The honest answer: no, but the easy version is. What works now is the same fundamentals pointed at a new reality:

Relevance still means content that genuinely answers the search. Authority, the links and mentions that make the rest of the web vouch for you, matters more than ever, because it is also what makes an AI engine trust you enough to cite you. And the target moved up: the goal is to be the cited source, optimized for buyers, not the tenth blue link optimized for a vanity keyword.

That is the work I do, and the method I write down and keep current as the engines change. It got harder, which is exactly why it is worth doing well.