Almost every guide on how to look up keywords on a website answers the wrong question. They hand you a list of tools and say “paste the URL, get the keywords.” But there are two completely different things hiding inside that request, and mixing them up is why most people walk away with a spreadsheet they can’t act on. One question is what is this page trying to rank for — the intent baked into its title, headings, and copy. The other is what does this page actually rank for today — the live queries sending it traffic. These are not the same list, they rarely overlap cleanly, and the method you use depends entirely on which one you’re after.
The two questions hiding inside one search
Before you open a single tool, decide which problem you’re solving. A page’s target keywords live on the page itself — the phrases the writer optimized for, visible in the title tag, the H1, subheadings, image alt text, and repeated body terms. You can read these directly, for free, on any public URL. A page’s ranking keywords live in Google’s index and change daily. You can only estimate those with a tool that maintains its own keyword-to-URL database, or read them exactly if you own the site and have Search Console connected.
Why does the distinction matter so much? Because a well-optimized page targets maybe five to ten phrases, but a page that has aged into authority often ranks for hundreds — most of them long-tail variations the author never explicitly wrote. If you only read the on-page signals, you miss 90% of what’s actually driving traffic. If you only pull the ranking data, you miss the strategic intent behind the page. Serious competitor teardowns need both.
Method 1: Read the page source for target keywords
The fastest way to see what a page is trying to rank for is to read its own markup. Open any page, right-click, and choose “View page source” (or add view-source: before the URL). Then scan four places in order of weight:
- The title tag (
<title>) — the single strongest on-page signal of the page’s primary target. - The H1 and H2 headings — the topic structure and secondary phrases.
- The URL slug — often the cleanest statement of intent, because slugs get chosen deliberately.
- The meta description — not a ranking factor, but it tells you the phrases the author wanted to win.
The old trick of reading the meta keywords tag is dead — Google ignored it over a decade ago and most sites stripped it out. Don’t waste time there. Instead, note which phrases repeat naturally across the title, the first paragraph, and the subheadings. That repetition, not a hidden tag, is the real fingerprint of target intent.
Method 2: Search Console — the only ground truth for sites you own
If the site is yours, stop guessing. Google Search Console‘s Performance report shows the exact queries that produced impressions and clicks, with real position and click-through data straight from Google. Nothing else on this list is a fact; this is. When people ask how to look up keywords on a website they control, this is the first and best answer, full stop.
Two honest limits keep it from being perfect. Search Console samples and anonymizes: queries with very low volume get hidden to protect privacy, so your true long-tail is always larger than what’s shown. And it only reports the last 16 months. Still, for a site you own, one export of the Queries tab beats any third-party estimate, because it’s measured behavior rather than a modeled guess.
Method 3: Site explorers for what a page actually ranks for
For any site you don’t own — a competitor, a page you admire — you need a tool that has already crawled the web and built its own map of which URLs rank for which queries. Paste the domain or a specific URL into a site explorer (Ahrefs, Semrush, and similar tools all do this) and it returns the estimated organic keywords, each with a search volume, difficulty score, and the position that URL holds.
This is where reverse-engineering gets powerful. Sort the ranking keywords by traffic share and you instantly see which two or three phrases carry the page, versus the dozens of incidental long-tail terms it happens to catch. SEO Rocket runs this on live Ahrefs index data inside a chat window — you ask which keywords a competitor URL ranks for and it returns the pool with volume, difficulty, and the actual ranking position, so you’re reading real crawler data, not a spun estimate.
Method 4: Content gap analysis for the keywords you’re missing
Looking up one page’s keywords tells you what it wins. The higher-leverage move is asking what a set of competitors collectively rank for that you don’t. That’s a content gap analysis: feed in four or five rivals, and the tool intersects their ranking keyword sets against yours to surface the phrases where they all show up and you’re absent.
These gaps are usually the best-qualified keyword ideas you’ll find, because they’re already proven to drive traffic in your exact niche — you’re not guessing at volume, you’re copying a validated demand signal. SEO Rocket’s competitor gap analysis does this across multiple domains at once and hands back a prioritized list rather than a raw dump, which is the difference between “here are 4,000 keywords” and “here are the 30 worth writing about first.”
Method 5: Free reconnaissance with search operators
No budget? Google’s own operators do surprising work. Use site:example.com to see every indexed page on a domain, then site:example.com "topic phrase" to find which pages target a specific theme. Google’s autocomplete and the “People also ask” and “Related searches” blocks reveal the real query variations searchers type around a topic — free, first-party demand data straight from the source.
Operators won’t give you volume or difficulty numbers, so treat them as reconnaissance, not measurement. They’re for mapping the shape of a competitor’s coverage and discovering intent variations, which you then validate with a tool that has real metrics attached.
How the tools actually estimate (and why the numbers disagree)
Understanding the mechanism keeps you from over-trusting any single figure. Third-party site explorers don’t have Google’s data. They build a giant keyword database — often hundreds of millions of tracked terms — then periodically check where each URL ranks for each term and record it. Your “estimated organic keywords” is really “the subset of our tracked keyword universe this URL currently ranks in the top 100 for.” Two tools give different numbers because their keyword databases and crawl frequencies differ, not because one is lying.
Two practical consequences follow. First, brand-new or very low-volume queries may be invisible to every tool simply because nobody added them to the database yet. Second, estimated traffic is a model — volume multiplied by an assumed click-through curve for that position — so read it as directional, not literal. When a decision matters, cross-check the tool’s estimate against Search Console (if it’s your site) or against the raw SERP.
A worked micro-example: reverse-engineering one competitor page
Say a competitor’s article outranks you for a topic you care about. Here’s the ten-minute pass. First, view-source: their URL and note the title tag says “best budget standing desks” while the H1 says “affordable standing desks under $300.” Already you know their target: two phrasings of the same commercial intent. Second, drop the URL into a site explorer. It reports the page ranks for 140 keywords, but sorting by traffic shows just three — “cheap standing desk,” “budget standing desk,” and “standing desk under 300” — drive 80% of the traffic. The other 137 are long-tail scraps.
Third, you run a gap check and discover they also rank for “standing desk for small spaces,” a phrase your page never mentions. That single gap is your action item: it has real volume, it’s adjacent to your existing content, and a competitor already proved it converts in your niche. That’s the whole point of learning how to look up keywords on a website — not to admire the list, but to find the one or two phrases worth acting on this week.
Honest caveats: where this breaks down
Three failure modes trip people up. The first is the market mismatch: if a tool defaults to the US index and the site’s audience is in Singapore or the UK, the keyword list can look near-empty or wildly off, because you’re querying the wrong country’s index. Always set the target market before you trust the numbers. The second is chasing volume over intent — a page ranking for a high-volume term that never converts is worse than one ranking for a low-volume, high-intent phrase. Read the SERP to confirm the intent matches yours. The third is treating estimates as gospel: rankings jitter daily, and a single snapshot means little without a trend line. Track movement over weeks, not one afternoon’s spot check.
Frequently asked questions
Can I see a website’s keywords for free?
Partly. On-page target keywords are free to read via page source and search operators. Ranking keywords with volume and difficulty require a tool with its own index — most offer a limited free tier or trial, and SEO Rocket includes a free tier plus AI keyword research on real Ahrefs data at around $50/month for full access.
How accurate are third-party keyword tools?
Directionally reliable, precisely imperfect. They estimate from their own tracked keyword database and a modeled click-through curve, so treat the numbers as comparative rather than exact. For a site you own, Google Search Console is the only true source.
What is the difference between target and ranking keywords?
Target keywords are the phrases a page was optimized for — visible on the page. Ranking keywords are the queries the page actually appears for in Google, which usually number far more and skew long-tail. You need both views for a complete picture.
How many keywords should one page target?
Optimize each page around one primary keyword and two to four close variations that share the same intent. The dozens or hundreds of long-tail terms a strong page also ranks for come naturally from thorough coverage — you don’t stuff them in manually.
Turning a lookup into a plan
Knowing how to look up keywords on a website is step one; the value is in what you do next. The workflow that holds up: read the page source for intent, pull ranking keywords from a site explorer, run a gap analysis against three to five rivals, validate the shortlist against real volume and market, then write pages that beat the weakest competitor already ranking. SEO Rocket chains those steps in one chat — keyword research on live Ahrefs data, competitor gap analysis, a validation-gated AI writer, and rank tracking — the same playbook proven across 1,000,000+ ranking pages. Look up the keywords, then act on the two that matter, and you’ll be ahead of everyone still admiring their spreadsheet.
