Content and Keyword Research

This category covers the content side of SEO: choosing the searches a page should answer, writing that page so people and search engines understand it, deciding what to do with the pages you already have, and what AI answers in Google and ChatGPT change about that work (less than the hype suggests). It’s for marketing managers and business owners who write or commission their own site content and want it to earn search traffic without tricks.

Where to start

Read in this order; each step builds on the one before.

1. Keyword research: what keyword research is, short-tail vs long-tail terms, and sorting searches by intent (informational, navigational, transactional). Start here to decide which questions your pages should answer.
2. On-page SEO essentials: the page elements to get right once you know the topic, from titles, meta descriptions and headings to alt text, URLs and internal links.
3. How to build a URL inventory, collect Search Console data for each page and give every URL one decision (keep, update, merge, redirect or remove), with our own 17-page site as the worked example.

The first two are short overviews from our older blog. Read them for the vocabulary; the content audit post explains how we decided which older posts to keep and rewrite.

Common decisions and mistakes

Writing for a keyword list instead of a reader. Google’s SEO Starter Guide suggests anticipating the different words people use (its example: “charcuterie” vs “cheese board”), but also says not to worry about every variation, because its language matching can relate a page to queries it doesn’t spell out. Repeating a phrase to rank is keyword stuffing, a named violation in Google’s spam policies. In the GEO study, keyword stuffing gave little or no gain in AI answers either.

Burying the answer. State the direct answer in the first one or two sentences under a heading phrased the way people search, then explain. It helps the reader who skims, and a self-contained, sourced sentence is easier for an AI system to quote accurately. Our rule of thumb: if a sentence can’t be quoted on its own and still be true, rewrite it.

Publishing more pages instead of better ones. Google’s people-first content guidance asks whether you produce lots of content on many topics hoping some of it performs, or use extensive automation to do it. Mass-produced Q&A pages built only to be quoted by chatbots fit Google’s description of scaled content abuse, and its spam policies now explicitly cover attempts to manipulate generative AI responses in Search.

Pruning by age. A content audit is not a deletion exercise. Merge weak pages that compete for the same search into the strongest URL and redirect the rest; keep old pages that still answer their question. Removing pages helps only when they’re thin, duplicated or off-topic and can’t be improved.

Treating GEO as a new discipline. Google’s documentation on AI features says there are no additional requirements to appear in AI Overviews or AI Mode beyond the usual technical requirements, policies and best practices. The research on what gets a page cited is thin: essentially one paper, tested on 2023 systems, measuring how a source that was already in the answer got quoted, not how it was picked. In our view, GEO is best treated as an extra check on ordinary SEO writing, and anyone promising AI citations is promising something they can’t control.

Inventing numbers. Adding statistics and quotations helped in the GEO study, which is exactly why they’re tempting to fake. Use a number only when you can link the primary source that published it.

Getting help

If you’d rather have someone plan your topics, audit the pages you have and write fewer, more specific ones, that’s what our content marketing service does, with no promises about rankings or about which AI system will cite you.