Website SEO analysis: the 3-part framework

Full transcript

1,181 words · the narration as spoken in the episode.

Intro

Your organic traffic just dropped. Or it's flat, and you don't know why. Or you're about to build ten new pages and you have no idea if anyone will ever see them.

Three different problems. But they all come down to the same three questions: what is happening on my site, what does Google actually show for these searches, and what are my competitors doing that I'm not?

That's an SEO analysis. And in this series, we're going to do it properly: with real data, real APIs, and no guesswork. In this first episode I'll give you the framework we'll use for the next forty-nine videos.

Why most SEO analysis goes wrong

Most people start with a tool. They open a dashboard, look at a score, and fix whatever turns red. The problem is that a score is an answer, and you haven't asked a question yet.

Here are the three mistakes I see over and over.

First: looking at one data source only. Search Console tells you what happened on Google. It doesn't tell you what Google showed instead of your page. Keyword tools give you volumes, but not whether those searches are winnable.

Second: mixing up cause and symptom. "Clicks are down" is a symptom. The cause could be lost rankings, a worse title, a new feature in the results page that pushes you down, or just seasonality.

Third: no order. People jump straight to competitors, or to keywords, without first knowing their own baseline. Then they can't tell whether anything they changed made a difference.

So, a simple fix: three parts, always in the same order.

Part 1: Traffic analysis

Part one is traffic analysis. The question is: what is happening on my site, and where?

The main source is Google Search Console. It gives you four numbers: clicks, impressions, click-through rate, and average position. Alone, each of them can mislead you. Together, they tell a story.

High impressions and low clicks? Your page is visible, but the title or snippet isn't winning the click. Good position and falling clicks? Look at the results page itself, something else may have taken the space. Lots of pages at position eight to twenty? Those are your cheapest wins. They're sitting right at the edge of page one.

The second source is Google Analytics. Search Console counts clicks from Google Search. Analytics tells you what those visitors did, and, importantly, whether the traffic is real. Bot traffic can look like growth until you filter it.

What you get out of part one: a baseline, a list of pages that over- or underperform, and a short list of suspects. You don't fix anything yet. You write down what you've seen. In part three, you'll check these suspects against your competitors.

Part 2: SERP analysis

Part two is SERP analysis. SERP stands for search engine results page. The question is: what does Google actually show users for the searches I care about?

This is the part most people skip, and it's where the biggest surprises are. Because keyword volume tells you how many people search. The SERP tells you what they get.

Look at three things.

Who ranks. Is it big brands, small niche sites, forums, or marketplaces? If the first page is all forums, a well-made guide may beat them. If it's all giant brands, that's a different fight.

What type of page ranks. Guides, product pages, comparison lists, tools? That is Google telling you what searchers want. It's called search intent, and the results page is the most direct evidence you'll ever get.

What features appear. Featured snippets, "People also ask", videos, maps, AI-generated summaries. Every feature changes how many clicks the ten blue links can get.

You can do this by hand for a few searches. For fifty, five hundred, or a monthly check, you want an API. That's where DataForSEO comes in. It returns the live results page as structured data, so you can store it, compare it over time, and run the same analysis on every keyword.

And it's cheap. A single live results request typically costs a fraction of a cent. We'll go through the exact numbers and how to keep costs under control in episode two.

What you get out of part two: for each keyword, a clear picture of who you're up against, what Google rewards, and whether it's realistic to compete.

Part 3: Gap analysis

Part three is gap analysis. The question is: what are my competitors getting that I'm not?

A gap is the difference between where you are and where you could be, measured against someone who's already doing it. And there are three kinds.

The keyword gap. Searches where competitors rank and you don't appear at all. This is the classic one, and an API makes it quick: you pull the keywords each competitor ranks for and subtract your own. Treat the result as an estimate, and double-check the important ones against the live results.

The page-type gap. This one keywords can't show you. Maybe your competitors all have a comparison page, a calculator, or a location page for each city, and you don't have that type of page at all. You find it by looking at their URLs, not only their keywords.

The structure gap. How they link internally, how they organize categories, how deep their important pages sit. Two sites can have the same content and very different results because of structure.

Finding gaps is the easy part. The skill is ranking them. For every gap you ask three questions: how many people search for it, how winnable is it based on the SERP from part two, and what's it worth to you. That's how a list of five hundred opportunities becomes a plan with five items.

What you get out of part three: a prioritized list of what to build or fix first, and a reason for each.

How the three parts fit together

Notice the order. Traffic tells you where to look. SERP tells you what's realistic. Gap tells you what to do. Skip the first and you can't measure results. Skip the second and you'll chase keywords you can't win. Skip the third and you'll fix things nobody's asking for.

And it's a loop, not a line. After you ship a change, you go back to part one, wait a few weeks, and measure whether it worked. The last episodes in this series are about exactly that.

What's coming in this series

Here's the plan. We start with foundations: setting up your data sources and your workspace. Then search traffic analysis. Then demand and keyword research. Then SERP analysis. Then gap analysis. And we finish by putting it all together on a complete example site and automating the whole thing.

Every episode comes with the code or template we use, so you can run it yourself. I'll use a mix of made-up example sites and real data from my own sites, anonymized so we stay focused on the method and not on the specific keywords. You can follow along with any site you own.

Wrap-up

If you want the checklist for this framework, it's linked in the description. Next episode: what data you actually need, where to get it, and what it costs. See you there.