Search Intent Optimization: A How-To for Auditing and Repairing Pages You Already Own
A repeatable procedure for classifying existing URLs by SERP evidence, spotting intent drift, and deciding which pages to rewrite, merge, redirect or leave alone.
Search Intent Optimization Is an Audit-and-Repair Job, Not a Content Job
Most advice on search intent optimization assumes you are about to write something new. In practice, the more painful problem is the inventory you already have: hundreds of URLs published over years, briefs written by people who have since left, pages edited incrementally until they serve no single intent clearly. Your organic performance is flat not because you lack content, but because a meaningful share of your URLs answers a question nobody asked.
This procedure is deliberately upstream of everything else. It is not about AI answer surfaces, citation tracking or tooling. It is about one decision per URL: does this page match the intent behind the query it targets, and if not, what do we do about it? The framing we use is intent drift - the slow slide of a page that was once correctly aimed.
Intent drift is invisible in rank trackers. A page can hold position 6 for two years while progressively answering less of the query it ranks for.
Who This Procedure Is For: The SEO Lead Who Inherited a Bloated Page Inventory
You will get the most out of this if you are an in-house SEO, content lead or agency account manager who has taken over a site with a mixed inventory of blog posts, category pages, service pages and location pages, and you suspect misalignment rather than thin content is the cause of the plateau. It assumes you can export a URL list and read a SERP. It does not assume budget for a rebuild.
Prerequisite 1: Pull a URL Inventory You Can Actually Sort
Start from a crawl or your sitemap, not from a keyword tool. Export every indexable URL with its title tag, primary H1, word count and the query it appears to target. Sort by organic impressions over the last 90 days so you are working on pages with evidence of demand. A 4,000-URL export is normal for a mid-sized ecommerce or property site; you will not audit all of it, so rank it.
Prerequisite 2: Build a SERP Evidence Sample Before You Label Anything
This is the step most teams skip, and it is the one that makes the rest defensible. For each query, open the result page and record the top five to ten organic results by page type, not by brand. Manual sampling for 50 to 100 priority queries is enough to expose patterns; you do not need an API to start. Record what Google is rewarding, then label.
| Query | Sampled page types (top 5) | Intent label | Owning URL | Current page type | Gap | Action |
|---|---|---|---|---|---|---|
| best espresso machine for small kitchens | category, category, listicle, product | Commercial | /blog/espresso-guide | Blog post | Blog absorbing clicks a category should own | Rewrite as category + 301 |
Step 1: Classify Each Query by the Page Types Actually Ranking
Assign one dominant page type per query from this set: category, product, listicle, comparison, service, location, FAQ, forum thread, or video. Where the sample is split - say three comparison posts and two product pages - record the split and treat the query as contested rather than forcing a clean label. Contested SERPs are where your current page type most often loses.
Step 2: Fast-Pass the Intent Taxonomy Using Google's Own Know / Do / Website Framing
You do not need a taxonomy debate. Google's own quality rater documentation scores pages against whether the query is a know query (information), a do query (action or transaction) or a website query (navigate to a specific destination). Map every query to one of those three, then to a sub-type using the modifiers below. This takes minutes per query and kills the vague 'informational vs transactional' argument that stalls workshops.
| Modifier | Sub-intent | Archetype it implies |
|---|---|---|
| near me | Do (local) | Location page with NAP and hours |
| vs | Know (commercial investigation) | Comparison page with a verdict |
| for [audience] | Do | Segment landing page |
| under R[price] | Do | Filtered category or budget hub |
| without | Know | Constraint-led explainer |
| alternatives to | Know (commercial) | Honest alternatives list |
| how to | Know (procedural) | Step guide with real steps |
| is X worth it | Know (evaluation) | Verdict page with trade-offs |
Step 3: Score Every URL for Intent Drift, Not Just Keyword Targeting
For each URL in scope, score 0-2 on four questions: does the H1 name the query's outcome; does the first screen answer the sub-intent; does the primary CTA match the intent type; and does the body contain evidence (specs, prices, real steps) the ranking pages contain? A total of 6-8 is aligned. Three or below is drift, and drift compounds - pages with three or below are almost always the ones a previous team patched by adding a paragraph in 2023.
Drift often comes from legitimate edits: a seasonal banner, a new lead-capture form, a rewritten intro for a campaign. Each change is reasonable; the sum is a page that no longer completes the query.