Broad keywords don't rank on landing pages. Niche ones do.

This is a problem you will often come across: choosing the right kind of keyword.

A broad keyword is a term like “digital marketing.” C-level execs and business owners often prefer and push for keywords like this. But broad keywords trigger informational SERPs. Google fills them with guides, explainers, and Wikipedia entries. Your service/product page will be extremely difficult to rank for a keyword like this.

A niche keyword, on the other hand, is a keyword like “digital marketing agency for ecommerce.” Google likes to rank commercial product pages, service pages, and pricing comparisons for a keyword like this. This is where landing pages can rank and convert.

Simple, but you’ll be surprised what a pain in the a** it is to convince clients.

Can we rank broad keywords in landing pages?

Yes, aggressive digital PR combined with solid technical SEO and decent content can rank any page on Google, and consequently on AI LLMs. Enough high-authority backlinks will override Google’s intent classification.

For example, a cybersecurity vendor publishes original research on breach response times. Wired covers it and ZDNet runs a piece. Three industry publications are now linking to their product page. Soon, that page starts ranking for “endpoint security,” a keyword where the SERP is otherwise all educational guides. The content may only be vaguely relevant, but with enough backlinks, you can game Google. It’s been the fundamental flaw with Google’s algorithms since it first came about.

However, backlinks are expensive. Super-expensive. They are also slow. They require something a writer actually wants to cover. Either that or thousands of dollars.

For the vast majority of landing pages, targeting a niche keyword (with some volume) where you can win is the better play.

What do niche, intent-matched keywords actually look like?

Keyword	Intent	What usually ranks	Will your landing page rank?
“recruitment”	Informational	Guides, job boards, Wikipedia	No
“recruitment agency for fintech”	Transactional	Agency service pages	Yes
“accounting software”	Informational	NerdWallet, PCMag, Investopedia	No
“accounting software for small business”	Transactional	Product pages	Yes
“endpoint security”	Informational	CrowdStrike guides, Gartner	No
“endpoint security for remote teams”	Transactional	Vendor pages	Yes
“personal injury lawyer”	Informational	Directories, legal guides	No
“personal injury lawyer Austin”	Transactional	Law firm landing pages	Yes

This pattern repeats in every industry. The broad keyword has volume but informational intent. The niche keyword has a fraction of the volume, but every search is a potential buyer.

Long-tail keywords convert at roughly 2.5x the rate of broad terms. Someone who typed 2 words is browsing. Someone who typed 4+ words is probably desperate.

Desperate people convert better.

How do you check if your keyword has intent mismatch?

Search it in incognito mode. Look at the top five organic results.

If it’s all guides, listicles, and god forbid, Wikipedia, you can conclude that Google thinks the intent is informational. A commercial service/product page will rarely rank for an informational query.

All product and service pages? Transactional. Your landing page fits.

The SERP is the intent classifier. Don’t worry about Ahrefs’ label, Semrush’s score, etc. Those are just algorithmic guesses. Only the live SERP tells you the truth.

How does RAG architecture handle keywords?

AI answer engines retrieve pages and split them into chunks (fancifully, this is called RAG architecture). In other words, AI ignores most of your page and just chunks each heading (H2) and its content as a standalone content bundle. The model then evaluates each chunk independently against other chunks it has gathered from other websites.

Staying with recruitment company example:

On a commercial intent page, your H2 sections rarely cover what recruitment is, types of recruitment, how agencies work, and industry trends. Each chunk on a landing page is shallow and, for the most part, promotes your product/service. The RAG retriever compares your shallow chunk against a dedicated guide that goes deep on the same sub-topic. The random internet guide wins.

Niche keywords limit the scope. Your H2 sections cover hiring timelines at Series A, what specialized recruiters can do for fintechs, contract vs. retainer models, etc. Each chunk goes deep on one specific thing. AI LLMs like that, and so does Google now.

What is a good H2 section for AEO?

One question per H2. Answer in the first sentence. Simple.

Q: How long does a specialized recruitment agency take to find startup talent?

Bad: “Hiring in the startup ecosystem involves many variables that can affect timelines.”

Good: “Series A fintech companies typically fill technical roles in 6-8 weeks with a specialized recruiter.”

The RAG retriever pulls a chunk, checks if it answers the query, and decides in milliseconds. If your answer is buried in sentence three, a competing chunk with the answer in sentence one gets cited instead.

There is also the query-fan-out strategy, which I will link here when it’s ready.

How to choose landing page keywords in four steps

1. Search the keyword in incognito. If the top five are informational, it will be somewhat challenging to rank your promotional landing page organically.
2. Narrow until the SERP turns commercial. Add your vertical, your use case, a location, or a modifier like “solution” or “for [specific audience].”
3. Structure every H2 for RAG and QFO. One question. Answer in sentence one. Each section standalone.
4. Digital PR only for your most strategic keywords. Get that ROI.

Here is a little .pdf checklist I built using Claude for this, it’s not perfect but it’ll get the job done: Landing_page_KW_checklist

Target what you can win and structure it so machines can read it. Combine it with decent on-page content. Simple AEO/GEO/SEO strategy.