Before you start
- The website you're researching (your own or a competitor's)
- Free accounts on Similarweb + Ahrefs / Semrush (free tiers cover most of this)
- 30-45 minutes for a complete discovery cycle
- A simple spreadsheet to consolidate findings
The playbook
7 steps
Start with traffic-overlap on Similarweb
Similarweb's Competitors tab is the fastest first pass. Paste the URL, scan the top 5-10 results. These are sites whose audience meaningfully overlaps with the target - your starting candidate list. The free tier limits depth but gives you the breadth pass for free.
Expected outcome
5-10 traffic-overlap candidates captured in your spreadsheet.
Layer in keyword-overlap from Ahrefs or Semrush
Different tools surface different competitors because they measure overlap differently. **Ahrefs Site Explorer → Organic competitors** lists sites ranking for the same queries. **Semrush Competitive Research → Organic Competitors** does the same with different weighting. Run both if you have access; either alone is incomplete. Add new names to your candidate list.
TipSort by 'common keywords' descending, not 'authority'. A small site ranking on 200 of your terms is a sharper competitor than a giant site ranking on 50.Expected outcome
Keyword-overlap competitors merged into your candidate list, deduplicated.
Pull paid-search competitors
If you have Google Ads access to the target, **Auction Insights** is the ground truth - it shows the domains you actually compete with at auction. If not, **SpyFu → enter URL → Competitors → PPC tab** approximates the same. Paid-search competitors are usually the most operationally aggressive members of the set.
Expected outcome
Paid-search competitors added with 'source: ppc-overlap' tag.
Mine the SERP for the target's top 10 commercial keywords
Take 10 commercial-intent keywords the target ranks for. Search each in incognito Google. Document the top 10 organic results + top 4 paid ads per query. Brands appearing repeatedly across queries (3+ of the 10) are core competitors. One-off appearances are tangential.
# Example workflow: # 1. Get top 10 commercial keywords from Ahrefs Top Pages # 2. Open each in incognito Google # 3. Tally brand appearances in top 14 results (10 organic + 4 ads) # 4. Brands at 3+ appearances = core competitorsExpected outcome
SERP frequency tally identifying core vs peripheral competitors.
Check 'alternatives to <target>' and review-site mentions
Google '<target> alternatives' and '<target> vs'. Open G2, Capterra, Reddit threads, and any 'best of' listicles. The brands explicitly compared to the target are the competitive set buyers actually consider. This step finds the recent entrants and niche players the tool-based methods miss.
Expected outcome
Mention-graph competitors captured from review sites and listicles.
Verify paid-social activity per competitor
For each candidate, do a 30-second check in **Facebook Ad Library** and **LinkedIn Ad Library**. Active paid spend = operationally serious. Dormant = either bootstrapped/profitable (rare) or in a slowdown (common). Tier serious operators above dormant ones.
Expected outcome
Per-competitor activity flag (active / dormant / unknown).
Synthesize into a tiered output
Three tiers. **Tier 1 (3-7 brands)**: appears in 3+ discovery methods AND active in paid. Weekly monitoring. **Tier 2 (3-7 brands)**: appears in 2 methods. Quarterly audits. **Tier 3 (everything else)**: 1 method, low priority - Google Alerts only. The tiered output is what your team actually uses; an unstructured list of 20 brands gets ignored.
Expected outcome
Final 3-tier competitor set ready for weekly, quarterly, and passive monitoring cadences.
Shuttergen
Discovery finds them. Shuttergen helps you outship them.
Once your competitor set is tiered, the next move is shipping creative that beats theirs. Shuttergen runs a live creative audit on your Tier 1 set and generates variants tuned to outperform.
Pitfalls
What goes wrong
Relying on a single tool
Each tool measures overlap differently. Similarweb finds traffic overlap, Ahrefs finds keyword overlap, SpyFu finds paid overlap. The intersection is the real set; the union is what you should research.
Skipping the SERP mining step
Tools index slowly. New entrants and niche players show up in manual SERP searches months before they show up in tool databases. Manual mining is non-optional.
No tiering on the output
A 20-brand flat list is intelligence overload. Team doesn't know who to watch weekly vs quarterly vs passively. Tier the list or it won't get used.
Quarterly re-discovery skipped
Competitive sets shift every 6-12 months. Discovery is not one-and-done. Re-run quarterly to catch new entrants and exits.
Limits
When this playbook won't work
- Sites under 6 months old - tools haven't indexed enough traffic/keyword data
- Hyper-local single-city service businesses - international tools underindex them
- Very small TAM B2B - overlap signal is too thin
- Affiliate-dominated niches where the visible URL isn't the operator
What 'website competitor' actually means in 2026
Definition has gotten messier. A website competitor isn't just 'another company selling the same product.' It's any site competing for the same audience attention, search intent, ad inventory, or buyer consideration set. A SaaS competitor might lose more eyeballs to an unbundled adjacent tool than to a direct competitor. A DTC brand might lose more buyers to a content site running affiliate links than to a head-to-head DTC.
The implication: multi-method discovery is non-optional. Single-method discovery (e.g. 'who else sells my product?') misses 50% of the competitive set. The methods in steps 1-6 above each surface a different slice; the union approximates reality.
The output should drive weekly action, not annual decks. A competitor list that gets refreshed annually and reviewed once a year is a decorative artifact. A tiered list with weekly monitoring on Tier 1 turns competitive intelligence into shipped strategic moves. The discipline is the difference.
Internal: find-competitors, competitive-analysis-software, competitor-monitoring-tools.
Discovery finds them. Shuttergen helps you outship them. Once your competitor set is tiered, the next move is shipping creative that beats theirs. Shuttergen runs a live creative audit on your Tier 1 set and generates variants tuned to outperform.
Free vs paid: when each is the right call
Free is enough for the first discovery cycle. Similarweb free tier + manual SERP mining + ad-library checks find 70-80% of the competitive set. Sufficient for a single-shot research project.
Paid earns its weight when discovery is ongoing. If you're refreshing quarterly, monitoring weekly, and feeding the data into creative/positioning decisions, paid tools (Ahrefs, Semrush, Similarweb Pro) speed up each cycle by 3-5x. The breakeven is usually around 'this is a recurring workflow' rather than 'this is a one-time question.'
The most underused free tool is Google itself. Manual SERP mining and 'alternatives' searches catch recent entrants months before they appear in paid tool databases. For competitive intelligence on emerging categories, manual beats paid every time.
Internal: keyword-monitoring.
FAQ
Frequently asked
What's the fastest way to find website competitors?
Are free competitor research tools good enough?
How many competitors should I research?
Do tools find the same competitors?
How often should I refresh the competitor list?
What if my tools find no competitors?
What's the difference between SEO competitors and business competitors?
Related
Keep reading
Resource
Find competitors
Sister keyword variant.
Resource
How to find competitors of a website
Long-form variant of this playbook.
Resource
Competitor monitoring tools
Tools for ongoing monitoring.
Resource
Competitive analysis software
Software roundup.
Research
Anatomy Of Good Meta Ad Library
Apply discovered list to a creative audit.
Discovery finds them. Shuttergen helps you outship them.
Once your competitor set is tiered, the next move is shipping creative that beats theirs. Shuttergen runs a live creative audit on your Tier 1 set and generates variants tuned to outperform.