How to Spy on Your Competitors' Keywords

Most people start competitor keyword research by listing the businesses they compete with for customers, then plugging those names into a tool. That's usually the wrong list. The furniture store two blocks away might be your real-world competitor, but if a national retailer and three affiliate blogs own every page one of ranking for "best ergonomic office chair," those are the sites actually fighting you for the click. Search competitors and business competitors overlap less than people assume, and starting from the wrong set wastes the rest of the research.

This is a walkthrough for finding the right competitors, pulling the keywords they rank for that you don't, and turning that list into something you'll actually act on, rather than a spreadsheet that sits unopened after the first week.

Find your real search competitors first

Forget your mental list of rivals for a minute. Open Google (incognito, so personalization doesn't skew results) and search the 10 to 15 keywords that matter most to your business, the ones closest to a sale or a signup. Write down whichever domains show up repeatedly across that list, not just once.

A domain that appears for two of your terms is probably a tangential player. A domain that shows up for eight of them is a real search competitor, whether or not you'd ever think to compare yourselves in a board meeting. It's common to find that your actual SERP rivals are a mix of direct competitors, a few much bigger brands that happen to rank adjacent to your category, and two or three content sites or comparison blogs that out-publish everyone.

If you'd rather skip the manual tally, Ahrefs' Competing Domains report (inside Site Explorer) and Semrush's Competitors report both do this automatically: plug in your domain and they'll surface the sites with the highest keyword overlap, ranked by how many shared rankings they have. Both are paid features, but they save the fifteen-keyword manual exercise above once you already have a subscription. Either way, the output should be four to six domains, not twenty. More than that and your gap analysis turns into noise.

One more filter worth applying: separate the sites that compete with you commercially from the ones that just rank well on the topic. A software review aggregator might out-rank your SaaS product for "best project management tool," but it's not selling project management software. You still want its keywords in your research (it's a useful proxy for topic coverage), just don't mistake it for a competitor whose whole content strategy you should copy.

The free methods, and what they're actually good for

None of the free options give you a side-by-side keyword list the way a paid gap tool does. What they give you is enough signal to prioritize before you pay for anything, or to get by entirely if your budget is zero.

Google Search Console, used correctly. GSC doesn't show competitor data, full stop. It only reports on queries where your own pages already get impressions. What it's good for here: pull every query where you rank positions 4 through 15, sort by impressions, and you've got a list of keywords you're already partially competing for. Take that list to Google and manually check who holds positions 1 through 3 for each one. That tells you exactly who's beating you and on what, for the queries closest to being winnable, without paying for anything.

The `site:` operator. site:competitor.com keyword tells you whether a competitor has a page targeting that term at all, and often which one. It's blunt and doesn't give volume or ranking position, but it's fast for a quick check ("do they even have a page for this?") before you spend a tool credit confirming the same thing.

Manual SERP review. Search your priority keywords in an incognito window and actually read what's ranking: title tags, the angle each page takes, whether it's a listicle, a tool, a long guide, or a product page. This is the step people skip because it's slow, and it's also the step that tells you things a keyword tool can't, like whether the top five results are all outdated and due for a refresh, or whether Google is clearly rewarding a specific content format for that query.

Google Ads Keyword Planner. You need a Google Ads account to access it, but you don't need an active campaign to open the tool. Without ad spend, though, volumes show as broad ranges ("1K-10K") instead of exact numbers; running even a small campaign gets you tighter figures (Google Ads Help, Keyword Planner). It won't show you which competitor ranks where, but it's a legitimate free source for checking whether a keyword idea has real demand before you build a gap list around it.

Ahrefs' free Backlink Checker. Not a keyword tool, but worth mentioning here because it's genuinely free and relevant to the same research: enter any domain and see its top 100 backlinks and basic authority metrics, no subscription required (a free account is now needed to see the full numbers, per Ahrefs' 2026 update to the tool). Useful for sizing up whether a competitor's ranking strength comes from content, links, or both.

Stack these four together (Search Console plus manual review plus Keyword Planner plus a few site: checks) and you can run a credible first pass on competitor keywords without a credit card. It's slower, and it won't scale past a handful of competitors and a few dozen keywords. That's the trade-off.

Keyword gap analysis, the core technique

Once you're ready to pay for speed, this is the technique the paid tools are built around, and it's simpler than the dashboards make it look. You put your domain next to one or more competitor domains, and the tool returns keywords sorted into a few buckets:

• Keywords competitors rank for and you don't rank for at all ("Missing" in Semrush, folded into Content Gap in Ahrefs)
• Keywords you both rank for, but they outrank you ("Weak")
• Keywords only you rank for (useful for seeing what you'd lose if you deprioritized a page)
• Keywords none of you rank for yet, surfaced from the tool's broader keyword database as adjacent opportunities

The "Missing" and "Weak" lists are where almost all the useful work happens. Missing keywords are net-new content opportunities. Weak keywords are often faster wins, since you already have some relevance signal for the term and might just need a stronger page, better internal links, or a content refresh rather than something built from scratch.

The trap at this stage is treating the raw export as a to-do list. A gap tool comparing you against three competitors can return thousands of rows. Most of them are irrelevant, too difficult to be worth it this quarter, or variations of the same core topic. The filtering step (next section) is what makes the tool worth the subscription instead of being a very large spreadsheet.

Finding their top pages, not just their top keywords

A keyword list tells you what to target. Their top pages report tells you how they're winning. Both Ahrefs (Top Pages, inside Site Explorer) and Semrush (Organic Research > Pages) will show you a competitor's highest-traffic pages, sorted by estimated organic traffic.

Open the top ten or so and look for a pattern rather than copying any single one. Are they mostly long guides, comparison pages, calculators, free tools, or glossary-style definition pages? Is there a format Google seems to reward for this competitor specifically, a video embedded high on the page, a comparison table near the top, original data? That pattern matters more than any individual page, because it tells you what kind of content investment is actually working for that site in that niche, as opposed to a one-off post that happened to catch a trend.

Cross-reference the top pages list against the keyword gap list. When a page shows up as both a top-traffic page and the ranking source for several of your "Missing" keywords, that's a strong signal the page itself, not just the keyword, is worth studying closely: structure, internal links pointing to it, how deep it goes versus how deep you'd need to go to beat it.

What competitors are paying for

Organic keyword gaps are only half the picture if a competitor also runs paid search. Ahrefs and Semrush both show a competitor's estimated paid keywords inside the same competitive analysis tools (Ahrefs calls it Paid Keywords under its competitive analysis section; Semrush calls it Advertising Research). SpyFu built its reputation specifically on this: its PPC Kombat tool compares up to three domains and flags keywords your competitors are bidding on that you aren't, plus the reverse, keywords you're paying for that nobody else is bothering with.

This matters for organic strategy too, not just PPC budgets. A keyword a competitor is willing to pay for, consistently, over months, is a keyword with commercial intent they've validated with real spend. If that term also shows up as a "Missing" keyword in your organic gap report, it's a stronger signal than volume alone that building a page for it is worth the effort; somebody's already proven it converts enough to justify an ad budget.

The paid tools, compared

Four tools come up constantly for this work, and they're not interchangeable. Here's what each one actually does for competitor keyword research, and current pricing as of October 2026 (all vendors change prices often enough that you should check the live page before budgeting against one of these numbers).

A few notes the table can't carry. Semrush's Keyword Gap tool compares your domain against up to four competitor domains at once and separates results into Missing and Weak tabs, with filters for keyword difficulty and volume, plus CPC. Ahrefs' Content Gap is the deeper of the two for keyword-level work, letting you pull up to ten competitor domains into a single comparison, and its companion Competing Domains report in Site Explorer finds competitors you might not have thought to add, based on actual keyword overlap rather than your guess. We cover the broader pricing and feature trade-off between these two in our Semrush vs Ahrefs comparison; the short version is Ahrefs edges out on link and content-gap depth, Semrush wins on breadth if you're also doing PPC or content tooling in the same seat.

SpyFu's Kombat tool caps out at three domains per comparison instead of four or ten, but it's the only one of the three built with PPC as a first-class citizen rather than an add-on report, which matters if paid and organic gap analysis need to live in one place. Its pricing has moved around enough across 2026 that third-party trackers disagree on the exact current number; confirm it directly before committing to a plan.

Similarweb doesn't do keyword-level gap analysis the way the other three do. What it's for is estimating a competitor's total traffic, where that traffic comes from (direct, organic search, paid, social, referral), and how your audience overlaps with theirs. It's a useful sanity check layered on top of keyword-level data (does this competitor's traffic shape match what their keyword rankings imply?), not a replacement for it. The free tier gives a limited snapshot (a handful of metrics, one month of history); the paid Starter plan runs $149/mo billed monthly or around $125/mo billed annually, per Similarweb's plans page.

If you only buy one tool and your work is mostly organic content, Ahrefs or Semrush will cover more of this workflow end to end than SpyFu or Similarweb alone. If PPC gap analysis matters as much as organic to your team, SpyFu's combined view is worth a look specifically for that reason, and our roundup of free AI SEO tools is a reasonable place to stretch a zero-dollar research budget further before committing to any paid seat.

Turning the data into a prioritized content plan

Export the Missing and Weak lists, then filter before you do anything else. A workable first pass:

Cut anything with search volume under whatever floor makes sense for your traffic goals; for most small sites that's somewhere around 50-100 monthly searches, lower if you're early stage and building topical depth, higher if you need results fast. Cut anything whose difficulty score sits well above what your domain has historically been able to rank for, you can check this against keywords you already rank on page one for for a rough calibration. Then sort what's left by relevance to what you actually sell or publish, not by volume alone; a 2,000-search keyword adjacent to your product beats a 10,000-search keyword three topics removed from it.

What's left is your real list, usually a few dozen keywords rather than the thousands the raw export handed you. Group them by topic, not by keyword, since several "Missing" keywords usually map to a single page you need to build or fix, not five pages. For the Weak list specifically, separate "needs a better page" from "needs more authority pointing at an already-decent page"; the fix is different and conflating them wastes a content brief on a linking problem.

Assign each surviving topic a rough score: search volume, estimated difficulty relative to your site's current strength, and commercial relevance, each on a simple 1-5 scale, multiplied or summed depending on how much you trust any one input. The exact formula matters less than having one consistently applied scale, so two people building the same list land on roughly the same priority order instead of arguing from gut feel each time.

A worked example (hypothetical)

None of this is a real business; it's a composite built to show the mechanics. Say you run a hypothetical site called "DeskHaven," selling standing desks and desk accessories direct to consumers, doing maybe 8,000 organic sessions a month.

Running Competing Domains in Ahrefs turns up five sites with meaningful keyword overlap: two direct standing-desk competitors, one big-box retailer whose furniture section ranks broadly across the category, and two review/comparison blogs that out-publish everyone on "best X for Y" terms. You drop the big-box retailer from the gap comparison (too broad a keyword footprint to be useful) and run Content Gap against the remaining four.

The export returns around 1,400 keywords. After cutting anything under 50 monthly searches and anything with a difficulty score more than 15 points above what your site currently ranks for, you're down to about 90. Sorting by relevance drops out another 40 that are tangential (desk accessories you don't sell, office chair terms that belong to a different product line entirely). You're left with roughly 50 keywords, and grouping them by topic collapses that into about 14 distinct content needs: six net-new pages ("standing desk for small spaces," "best desk converter under $300," and similar), five existing pages that rank on page two and need expansion or a stronger structure, and three pages that need backlinks more than new content, since they already target the right term reasonably well.

That's the prioritized list. Fourteen items you can actually brief and schedule, built from a 1,400-row export that would have been useless handed to a writer as-is.

The limits of third-party traffic estimates

Every traffic number any of these tools shows you (organic traffic estimates, a competitor's "monthly visits," a page's estimated search traffic) is modeled, not measured. These tools don't have access to a competitor's analytics. They estimate traffic from a combination of ranking position, their own click-through-rate curves by position, and in some cases panel or clickstream data licensed from third parties. That means two different tools can show meaningfully different traffic numbers for the exact same domain, and neither is simply "wrong"; they're built on different models and different underlying data sources.

Treat these numbers as directional, useful for comparing two competitors against each other inside the same tool, less useful as an absolute figure you'd report to a client as fact. If a specific number matters for a real decision (a competitor's actual search footprint before a partnership conversation, say), cross-check it across at least two tools and expect the real figure, whatever it is, to sit somewhere in the range they both suggest rather than matching either exactly.

Keyword difficulty scores carry the same caveat. They're a useful triage signal, not a guarantee; a "medium difficulty" keyword can still take months to rank for depending on your domain's current authority and how aggressively the page is built, and the reverse happens too, where a nominally "hard" keyword falls faster than expected because the current top results are weak on a dimension the score doesn't capture well, like content depth or page experience.

Competitor data tells you where the gaps are. Turning those gaps into a plan is the next step, and our roundup of AI tools for SEO strategy planning covers the clustering and brief tools that speed it up.

While you're in the backlink reports, you'll probably spot junk links. Before you reach for the disavow tool, read how to disavow links and why you usually shouldn't.

Most of these reports live in Semrush; our Semrush review shows which plan you need for them.

FAQ

How many competitors should I compare against in a gap analysis?

Three to five gives you a workable, non-noisy list. Semrush caps you at four per comparison and Ahrefs at ten; going much past four or five in practice mostly adds irrelevant keywords rather than better ones, since you're widening the net rather than sharpening it.

Can I do competitor keyword research without a paid tool?

Yes, slower and at smaller scale. Combine Google Search Console's query data (for keywords you already partially rank for), manual SERP review, site: searches, and Google Ads Keyword Planner for volume. It won't give you a clean side-by-side export, but it's enough to prioritize before buying anything.

Why do Semrush and Ahrefs show different traffic estimates for the same site?

Both model traffic from ranking positions and click-through-rate assumptions rather than measuring actual analytics, and each uses its own data sources and formulas to do it. Different inputs produce different outputs; treat both as directional rather than exact.

Is SpyFu a good replacement for Semrush or Ahrefs?

For PPC-focused competitive research, SpyFu's Kombat tool is a genuine strength and arguably underrated for that specific job. For deep organic content-gap work across a large site, Ahrefs and Semrush generally pull from larger keyword databases and support more competitors per comparison, so most SEO-first teams reach for one of those two as the primary tool and treat SpyFu as a supplement, if they use it at all.

How often should I rerun a keyword gap analysis?

Quarterly is reasonable for most sites; monthly if you're in a fast-moving, competitive niche where competitors are publishing aggressively. Rerunning weekly rarely surfaces enough new signal to justify the time, since rankings and gap lists don't shift that fast outside of algorithm updates or a competitor's major content push.

For a broader look at what to automate in this kind of ongoing competitive tracking, see our roundup of SEO automation tools, and for the baseline technical work that needs to be solid before any of this content gets built, our technical SEO checklist covers what to fix first.