Two pages on the same site, both optimised for "email automation," both published eighteen months apart. You know the drill. Google alternates between them for weeks, then seemingly picks one at random, and the other sits in the "Submitted but not indexed" pile like a forgotten gym membership. That exact situation ate three afternoons of my life before I finally built the spreadsheet that fixed it.
A keyword map is the document that decides, once and for all, which page owns which query. Not a keyword list. Not a content calendar. A map. One row per target, one owner per row, no ambiguity. Get it right and your internal linking, your content briefs, and your on-page edits all start pointing in the same direction instead of competing with each other.
Key takeaways
- A keyword map assigns one primary keyword to one page — no shared ownership, ever.
- Build it from your existing URLs first, then layer in the keywords you want to rank for.
- Search intent decides the page type. Get that wrong and no amount of optimization saves you.
- For sites over 200 pages, work in silos or funnel stages — never in one giant flat list.
- A map is a living document. Mine gets reviewed every quarter, without exception.
- If two pages target the same query, you have cannibalization. Fix it with a merge or a redirect, not with hope.
How to build a keyword map for your site: the four-column method
Forget the fancy templates for a second. A keyword map needs exactly four columns to work. Everything else is decoration.
| Column | What goes in it | Why it matters |
|---|---|---|
| URL | The exact slug of the page | Prevents duplicate rows and dead targets |
| Primary keyword | One query per row, no exceptions | Kills cannibalization before it starts |
| Search intent | Informational, commercial, transactional, navigational | Determines whether the page should exist at all |
| Status | Live, planned, needs rewrite, merge candidate | Tells you what to do on Monday morning |
You can add a secondary keyword column, and I usually do. But if you add it, pick one semantic variant, not five. Pages that try to rank for everything rank for nothing. I learned that the hard way on a client blog where I lazily mapped seven related terms to a single article — it crawled around position 30 for the primary term for months until I split it into three pages.
Why search intent comes before keyword volume
Here's a trap I see constantly: people sort their keyword list by monthly volume, then assign the biggest terms to the biggest pages. That's backwards.
Intent decides the page format. If the query is informational ("what is keyword mapping"), you need a guide or a definition page. If it's transactional ("keyword mapping tool"), you need a product or pricing page. If it's commercial ("best keyword mapping tools"), you need a comparison article. Assigning a high-volume transactional term to a blog post guarantees you'll never convert the traffic you get.
The order is: intent → page type → primary keyword. Never the other way around.
What should be the first step in creating a keyword map?
Crawl your existing site first. Not your keyword list — your site.
This is the step most guides skip, and it's the one that saves you the most work. Before you plan anything new, you need to know what you already have. Export every URL from your sitemap, or run a crawler over the domain and pull the full URL list into a spreadsheet. On my own project, that first export gave me 134 URLs, of which I'd genuinely forgotten the existence of about 20.
Then, for each URL, write down the primary query that page is currently targeting — not what you wish it targeted. Use Search Console, look at the queries each URL actually receives impressions for, and pick the dominant one. Where three or four queries are tangled up, that's a signal the page is unfocused.
Now you have a baseline. The keyword research comes second, because you're filling gaps, not starting from zero.
Should you map keywords to pages, or pages to keywords?
Both, and in that order. Pages to keywords first (what do my existing URLs already own?), then keywords to pages (what am I missing, and where does it go?). Trying to do it in one pass is how you end up with a spreadsheet nobody reads.
Building the map without burning out
I'll be honest: the first keyword map I built took me about six weeks of on-and-off work. It was too much. The second time, on a different site, I did it in four days by working in clusters instead of row by row.
Cluster by topic first. Group your keywords into themes of five to fifteen related terms — "email automation basics," "email automation tools," "email automation setup." Each cluster becomes one page (or one page plus supporting articles). This is faster than mapping individual keywords, and it naturally prevents overlap because you're thinking in topics, not strings.
Once clusters exist, ask one question per cluster:
- Does an existing page already serve this topic? If yes, that page gets the cluster. Rewrite or expand it if needed.
- Is there a page that sort of serves it but is drifting? Split it or refocus it.
- Does the topic deserve its own page? Build one.
- Do two pages both serve the cluster? Pick the stronger one, redirect or merge the other.
- Is the topic too thin to justify a page? Fold it into a broader cluster's page as a section.
The merge-or-redirect decision
When two pages compete, you have three options and only three. Merge them into one longer page (best when both have decent content). Redirect the weaker one to the stronger (best when one is clearly dominant). Or rewrite one to target a completely different cluster (best when both have unique value and you just misassigned them).
What you don't do is leave both live and hope Google sorts it out. It won't. It will just keep flip-flopping and neither page will build authority.
Handling large sites with hundreds of pages
Past about 200 URLs, a flat keyword map stops being useful. You can't hold 800 rows in your head, and neither can the person who inherits the document after you.
Segment by silo or funnel stage. For a SaaS site, that typically means: awareness pages, comparison pages, feature pages, and conversion pages. Each gets its own tab or section in the map. For an e-commerce site, segment by category — never map individual product pages unless they're strategic.
Then prioritise. Not every page needs a primary keyword. On a 600-page site, I usually find that 60 to 80 pages are doing the real ranking work. Map those carefully. For the rest, assign broad topic ownership and move on. You can always drill deeper later.
Keyword mapping tools worth using
You can build a keyword map entirely in Google Sheets, and honestly, for sites under 200 pages, that's what I'd recommend. It's free, it's flexible, and you'll actually understand your own data because you typed it in.
Specialised tools add value when you're dealing with scale. They crawl the site, pull query data, and auto-suggest mappings. Useful. But be careful — automated mapping tools frequently assign keywords based on volume alone, which is exactly the mistake you're trying to avoid. Treat their output as a first draft, not a finished document.
For templates: a free spreadsheet template with URL, primary keyword, intent, and status columns is all you need. Anything fancier is usually someone's attempt to sell you software.
Keeping the map alive
A keyword map built once and never touched is worse than no map at all, because it gives you false confidence. Pages get added. Keywords shift. Competitors appear.
My rule: review quarterly. Fifteen minutes per silo. Check that every live page still has a primary keyword, every planned page got published, and no two pages have drifted into overlap. That last one is the sneaky failure. Two pages can start out cleanly separated and slowly converge as their content expands — usually because a writer added "one more section" without checking the map.
Add a last reviewed date column. It sounds fussy. It isn't. When someone asks in a meeting why a page targets a keyword nobody remembers choosing, you'll be glad you have it.
The sites that rank well consistently aren't the ones with the most sophisticated SEO stack. They're the ones where every page has a clear job, and everyone knows what that job is. The map is how you make that true — one row at a time.
