TLDR: Keyword analysis is the process of evaluating search terms by their search volume, competition and intent, so you can prioritise which keywords are worth targeting. It turns a long list of keyword ideas into a clear plan for your pages, your blogs and your wider SEO strategy.
What is keyword analysis in SEO?
Keyword analysis is the process of evaluating search terms to understand how much demand there is for them, how competitive they are and what the person searching wants. The core metrics are search volume, keyword difficulty and search intent. Together, they help you decide which keywords are worth targeting on a page, in a blog or across your wider SEO strategy.
It’s worth separating analysis from discovery. You’ve got all your keywords in the discovery phase, in your research phase, and then it’s just kind of like, okay, well, what are we actually going to do with those keywords? Keyword research finds the options. Keyword analysis decides which ones earn a place.
Why does keyword analysis matter for SEO?
Keyword analysis is foundational. SEO now has three strands: technical, content and, more recently, GEO (generative engine optimisation), which sits separately but stays closely aligned. Keyword research underpins all of them, because it tells you what people are searching for.
If you don’t have that information, you’re going into it blind, good keyword analysis guides everything that comes next. It helps you prioritise, find new topics and areas to cover, and understand how your ideal customer searches for the products or services you provide.
It also stops you chasing the wrong numbers. A keyword with huge search volume looks attractive, but if you have no realistic chance of ranking for it, it isn’t an opportunity. Volume always has to be balanced against competition.
And it comes down to time. Every page or blog takes resource to create, so you need to justify why it’s worth doing. It’s not always about the return, but typically it is about a return and a KPI towards that, you want to see that it’s doing something for the business’s bottom line at the end of the day.
How to do Keyword research for SEO
How you approach keyword research depends on how mature the account is. Most businesses will already have a list of topics they want to write about, so let’s use a blog as the example.
Start with a brainstorm before you open any tools. Ask yourself what someone might be searching for around this topic. This gives you your seed keyword, the core term everything else builds from.
Next, search for it on Google. Before going near an SEO tool, look at what already exists and what’s ranking. Then put your seed keyword into a tool like Semrush or Ahrefs to uncover variations, question-based searches and different ways people search that you hadn’t thought of.
For a blog, pick around three to five keywords. From there, build out your structure, your headings and how much content you need by looking at what currently ranks for those keywords.
That’s also the point to ask whether you can realistically compete. If you work in financial services and HSBC’s blog is ranking for your target keyword, you’re unlikely to outrank it, because HSBC has far more authority. It’s about being realistic on what you can achieve. In that case, a long-tail keyword (a longer, more specific search phrase) is often a better option, and it’s an increasingly useful way to gain visibility in AI-driven search too.
How to analyse search intent, volume, and keyword difficulty
Search intent is what it says on the tin. It’s about what the person is looking for and what kind of search they’re making. Informational and commercial are the two main types, but you’ll also see navigational and transactional intent labelled in platforms like Semrush.
Informational searches come from people looking for an answer, and often start with a question word, like “how long do running shoes last?” Navigational searches come from people trying to reach a specific website, such as “Nike login”. Commercial searches come from people further down the marketing funnel who are comparing options before they buy, like “best running shoes for beginners”. Transactional searches come from people who are ready to buy, such as “buy ASICS Gel Excite 11”.
Intent tells you which type of page to target a keyword on. Commercial and transactional keywords typically belong on product or category pages, while informational keywords belong on blogs. Semrush and Ahrefs will label intent for you, but it’s always worth checking what’s ranking in Google too.
Search volume is your search demand, usually shown as average monthly search volume. It’s an indicator of how many people are searching for that term.
Keyword difficulty is how competitive a keyword is. If your site has strong site authority, you’re more likely to rank for difficult keywords. A brand-new website, on the other hand, isn’t realistically going to rank for something like “running shoes” when Nike, ASOS and other big players have more authority, more links and far more history behind them.
How to spot SERP features before you commit to a keyword
A SERP feature is anything on the search engine results page that isn’t a traditional list of website links. There are lots of them, and Google changes them all the time.
Which features appear depends on the intent behind the search. Commercial queries are more likely to show paid ads or sponsored listings. Local searches like “coffee shops in Leeds” usually bring up the map pack. Informational searches, like what to buy your mum for Christmas, are more likely to trigger an AI Overview.
Checking SERP features before you commit to a keyword tells you how people are searching, what information they want and the format they want it in. If videos are ranking, for example, that’s a strong signal about how to present your own content.
SERP features also create extra opportunities for visibility. AI Overviews are tricky to navigate because there’s no clear rhyme or reason to which websites get cited, but the current thinking is to structure content so it’s more likely to be quoted. That means answering the question in your first paragraph, using bullet lists and writing in a question-and-answer style, because that’s what AI Overviews and LLMs tend to pull from. People Also Ask boxes are another useful source for shaping your FAQs and headings.
There is a flip side. All of these features compete for your users’ attention, which can mean lower traffic, particularly with zero-click searches, where people get their answer without ever visiting a website. SERP features bring challenges as well as benefits, but visibility still matters, so they’re always worth checking before you commit to a keyword.
How to use LLMs for SEO keyword research
LLMs like ChatGPT and Claude are useful at the start of keyword research, but they’re not a replacement for it. We advise that you shouldn’t use them for anything more than generating ideas and formatting.
Where they shine is discovery. Put your seed keyword into an LLM and ask for variations or question-based keywords, and you’ll get a list far quicker than typing it all out yourself.
The catch is that, on their own, LLMs like Claude can’t give you reliable search volumes, because they don’t have access to live search data. They’ll often suggest very niche terms because they think that’s how people search, when people probably don’t. Connected to a tool like Semrush or Ahrefs through an integration (such as an MCP), an LLM can pull in real keyword data, but the numbers still come from the tool, not the LLM. Either way, treat its suggestions as a starting list and check them against Semrush or Ahrefs to see what people are really searching for.
LLMs are also helpful for categorising keywords, especially if you prompt them well. Once you have your list, you can ask ChatGPT or Claude to group keywords by the topic areas on your website. Some of our keyword lists run to more than 1,000 terms, so this saves serious time. They can do a reasonable job of categorising by intent too, but it isn’t perfect, so you’ll still need to QA the results manually.
What they won’t do is the analysis. LLMs can help you spot patterns and ideas you might not have thought of, but prioritisation and next steps need real data and human judgement. Use them at the start, maybe use them to categorise, don’t use them for the full thing, and then make your own decision at the end.
How Modo25 can help with keyword research and SEO
Keyword research is the foundation of every successful SEO strategy, and it’s what our team of SEO specialists do every day. We use LLMs to speed up the research stages, which frees up more time for the work for our clients: analysis, prioritisation and turning keywords into content that ranks.
Whether you’re an agency that needs specialist support or a business looking to build a keyword strategy from scratch, we can help. And if you’d rather build that expertise in-house, our supported in-housing model means we’ll get your keyword research working properly, then train your team to take it forward themselves.
