To find keywords competitors are using, you have two routes: read their pages manually, which is free and slow and surfaces only what they targeted deliberately; or pull their organic keyword list from a tool, which shows every term they actually rank for — including the accidental wins that are often the most valuable finds. Serious work uses the second and sanity-checks with the first.
The harder problem is not extraction. It is filtering. A mid-sized competitor might rank for 40,000 terms, and 39,000 of them are noise. This guide covers both halves: getting the list, then cutting it down to the 30 keywords you should actually write for.
Step 1: Pick the right competitors
Your search competitors are frequently not your business competitors. The sites outranking you for “commercial kitchen equipment financing” might be a bank, an industry magazine, and two blogs — none of which your sales team has ever heard of.
Build the list from the results themselves. Take your five most important commercial keywords, search each one, and record which domains appear repeatedly in the top ten. Three to five recurring domains is your real competitive set. Add one aspirational site — bigger than you, same topical territory — because its keyword list is a map of where the topic goes.
Check the country index while you are at it. If you sell in Australia and your tool is querying the US index, every number you pull will be wrong and most of the competitor set will be irrelevant.
Step 2: Pull each competitor’s organic keyword list
A site explorer returns every keyword a domain ranks for, with position, estimated volume, difficulty, and the URL that ranks. This is the raw material.
Two views matter more than the full list. The top pages report shows which individual URLs bring in the most estimated traffic — that tells you what the competitor’s business actually runs on, which is more informative than any single keyword. And positions 1 to 3 filtered by volume shows you what they dominate, which is what you will have to displace.
One honest caveat: every number here is an estimate. Volume figures are typically modeled twelve-month averages, and position data comes from periodic crawls, so a competitor’s “estimated traffic” is directionally useful and precisely wrong. Use it to rank opportunities against each other, never as a forecast of what you will earn.
Step 3: Run a content gap across all of them
This is the highest-value step, and it is where manual research cannot compete. A content gap analysis compares your domain against several competitors at once and returns keywords they rank for that you do not, with each rival’s position shown in its own column.
The column layout is the point. A keyword where all four competitors rank in the top ten and you rank nowhere is a topic the market has validated and you have simply not covered. A keyword where one competitor ranks at position 47 and nobody else appears is probably an accident, not an opportunity.
Filter the output in this order:
- Keep terms where two or more competitors rank in the top 20. Multiple rivals ranking means the topic is real.
- Drop anything with a competitor’s brand name in it. You will not rank for their brand and you should not try.
- Drop volumes below your relevance floor. For most B2B niches that is around 20 to 50 monthly searches; for consumer, higher. Very low volume terms are still worth writing when intent is sharp, but they should not lead the plan.
- Sort by difficulty ascending within each intent group. You want the easiest term with real commercial value, not the biggest number on the sheet.
SEO Rocket runs content gap across up to five competitors with per-rival position columns, which is the format that makes this filtering fast. Whatever tool you use, insist on seeing each competitor’s position separately — a merged list hides the signal.
Step 4: Check what they targeted on purpose
Now do the manual pass on their five most important pages, identified from the top pages report. Three things to read:
- The title tag and H1. Whatever phrase appears in both is almost certainly the deliberate target.
- The URL slug. Slugs are usually written from the target keyword and rarely changed afterwards.
- The first 100 words and the H2s. These reveal the supporting terms and the subtopics they judged necessary.
Comparing intentional targets against actual rankings is where the good discoveries live. When a page ranks well for a term that appears nowhere in its title or headings, the competitor stumbled into it. Those terms are frequently under-served — you can build a dedicated page and beat a page that was never really trying.
Step 5: Separate opportunity types
Do not treat the surviving list as one queue. Split it three ways, because each needs different work and delivers on a different timeline.
Gaps are terms where competitors rank and you have no page at all. These need new content. Slowest to pay off — expect 8 to 16 weeks in a competitive niche before a new informational page settles at its true position.
Overlaps are terms where you both rank but they are ahead. These need improvement, not creation: expand the existing page, tighten the title, add the subtopics their page covers and yours does not, and build a few links. Fastest wins on the whole list, especially where you sit in positions 5 to 20.
Defenses are terms where you rank ahead but a competitor is climbing. Worth knowing about before you lose them. Load these into rank tracking and watch the trend rather than the daily reading — movement of two or three positions between checks is normal noise, not an attack.
Step 6: Turn the list into a build order
Cut to 30 keywords for the quarter. More than that and nothing ships properly.
Order them by a simple rule: commercial value first, difficulty second. Start with the terms where a buyer is at the end of the query — comparison terms, “best” terms, and problem-plus-solution phrasing — because those convert while you wait for the informational pages to mature. Then work down the informational list from lowest difficulty upward.
For each keyword, before writing, benchmark the weakest site on page one rather than the leader. Record its referring domain count and page length. If result number eight ranks with 19 referring domains and 1,300 words, that is your target. This single habit is the difference between a plan that finishes and one that dies in month two.
Save the final list into a keyword pool your writer and rank tracker both use, so you assign each term once and never retype it. Then check the plan against your own Search Console data — for pages you already have, your impression and position data beats any third-party estimate, and it will occasionally tell you that a term you were about to chase is already half-won.
What this method will not tell you
Three limits worth stating. You cannot see paid search data this way, so a competitor bidding hard on a term will not show up in organic exports. You cannot see conversion — a competitor may rank beautifully for a keyword that earns them nothing. And you cannot see their strategy, only its residue, so a page they published last week will not have rankings attached to it yet.
Use competitor keywords as evidence of validated demand, not as a plan handed to you. The pages that win are the ones that cover what the market proved people want, then add the thing none of the current results have.
