A rival's website is the one place where their whole strategy is on public display: what they sell, what they charge, who they sell to and how fast they are. Competitor website analysis is reading that display properly — page by page, against one checklist — instead of glancing at a homepage and guessing. Done once, it tells you where you are beaten and where you are ahead. Done every quarter, it tells you what changed.
The short version: pick three to five rivals, audit the same six pages on each, measure traffic, speed, technical health, content and the tech underneath, score it all on one sheet, and convert the biggest gap into one 30-day fix.
Step 1: Pick the rivals and the pages that matter
Start from the competitor analysis framework: three to five rivals your customers actually compare you with, not the biggest name in the industry. Then, for each rival, list the same six pages to audit:
- Homepage — the positioning sentence, the offer, the first call to action.
- Pricing page — the tiers, the anchor price, what is included and what costs extra.
- Bestseller or main category page — the product customers compare first.
- One top content page — their most-linked guide or blog post.
- Checkout or contact page — how many steps stand between a visitor and a purchase.
- About page — the story, the team, the proof (years in business, clients, awards).
Auditing the same pages on every rival is what makes the comparison honest — one rival's homepage against another's pricing page is a shop window against a till. Write the six URLs down first, and give yourself about 20 minutes per rival.
Step 2: Estimate their traffic and visibility
You will never see a rival's real analytics, and you do not need to. What you need is a rough size and a direction: are they bigger than you, growing, and where does their traffic come from? Similarweb's free checker answers that for most sites: type in a domain and it shows an estimated monthly visit count, the recent trend, engagement figures, the channel mix and the top countries — without an account. The free web lookup allows about three searches a day, and the browser extension is free with no account.
Two honest limits. First, every third-party traffic figure is a modelled estimate, not a measurement — built from panel data, not from the rival's analytics account. Second, for sites under roughly 5,000 monthly visits the free checker often shows nothing at all. That tells you the rival is small, which is itself useful information. Record the estimate with its date and a one-word verdict — bigger, similar, smaller — and move on.
For visibility, run the cheap checks that need no account: search site:rivaldomain.com to see roughly how many of their pages Google has indexed, and check whether they appear for the money keywords your customers type.
Step 3: Measure speed and Core Web Vitals
Speed is the one area where a free tool gives you numbers as good as a paid one. Google's PageSpeed Insights is free: paste in any URL and it returns real-user field data from the Chrome User Experience Report plus a Lighthouse lab test. The field data is what Google ranks on; the lab score is for diagnosis.
Google judges three Core Web Vitals at the 75th percentile of real visits, separately for mobile and desktop. The "good" thresholds are:
| Metric | What it measures | Good |
|---|---|---|
| LCP (Largest Contentful Paint) | How fast the largest visible element appears | 2.5 seconds or less |
| INP (Interaction to Next Paint) | How quickly the page visibly responds to a click, tap or key press | 200 milliseconds or less |
| CLS (Cumulative Layout Shift) | How much the layout jumps around while loading | 0.1 or less |
INP replaced First Input Delay as a Core Web Vital in March 2024. Test the rival's homepage and bestseller page on mobile first — most of your customers are on phones, and mobile is where the gaps usually are. Note the three numbers per page, per rival, with the date.
Step 4: Crawl the site like a search engine does
A website that looks fine in a browser can be broken underneath: dead links, duplicate titles, pages Google cannot reach, orphan pages. Screaming Frog's SEO Spider crawls a site the way a search engine does and lists every URL with its status code, title, meta description and headings. The free version crawls up to 500 URLs per crawl with no sign-up — enough for almost any small business site — and a licence costs $279 per user per year if you outgrow it.
Run the crawl on each rival's domain and pull out five numbers: page count, error pages (404s and server errors), pages with missing or duplicate titles, pages with missing meta descriptions, and images without alt text. A rival with 40 pages and 12 broken links has a site nobody maintains; a rival with 400 indexed pages and clean titles is a content machine. Cross-check the count against the site: search from Step 2 — a big gap between crawled and indexed pages usually means thin or duplicate content Google chose not to index.
Step 5: Read their content and find the keyword gaps
Traffic and speed tell you how big and how fast a rival is. Content tells you what they are winning on. Open their top content page and ask four questions: what keyword is it built around, what does it promise first, what does it cover that your page does not, and what does it link to next? Then compare the two with the competitor content analysis method — the same topic, side by side, scored on depth, freshness and proof.
Look for three kinds of gap. Coverage gaps: topics they rank for that you have no page for. Depth gaps: pages where they answer the question properly and yours stops halfway. Freshness gaps: their guide is dated this year and yours is two years old.
Do not copy their content. Copy the gap: the topic they cover and you do not, the question they answer and you leave open. One new page that closes a real coverage gap is worth ten rewrites of pages that were already fine. A rival's review themes often point at the same gaps — review analysis shows you what buyers praise and complain about.
Step 6: Look underneath — tech stack and change over time
A site's technology tells you how fast it can change. A rival on a modern stack with a blog they update weekly can react to your moves in days; a rival on a ten-year-old template with a blog last updated in 2023 cannot. Free tools such as BuiltWith or Wappalyzer list the platform, analytics, chat widget and payment provider a site runs. You are looking for one answer: is this rival's site actively maintained, or is it a brochure that happens to rank?
Then check what changed. The Wayback Machine (web.archive.org) keeps dated snapshots of most public sites, so you can see what a rival's homepage said six months ago: a new price, a new guarantee, a rewritten headline. A rival that changes its pricing page twice a year is testing; one whose site has not changed in two years is coasting.
Step 7: Score it on one sheet and pick one fix
Raw findings do not decide anything. Put every rival on one audit sheet and score each cell from 0 to 10, with yourself scored the same way. A starter sheet:
| Check | You | Rival A | Rival B | Rival C |
|---|---|---|---|---|
| Traffic (estimated monthly visits) | 5 | 8 | 4 | 6 |
| Mobile speed (Core Web Vitals) | 6 | 9 | 5 | 7 |
| Technical health (errors, titles) | 7 | 8 | 6 | 5 |
| Content depth on money topics | 5 | 7 | 8 | 4 |
| Pricing clarity | 6 | 5 | 7 | 8 |
| Checkout / contact friction | 7 | 6 | 5 | 9 |
Three rules keep the scoring honest: score the evidence, not the reputation; score yourself with the same standard; and use ranges where you are unsure. Feed the result into your SWOT: a gap where a rival beats you on something customers value is a weakness with a deadline, and a check where you lead is a strength you can say out loud, with the number attached.
Then pick exactly one fix for the next 30 days — the gap that scores highest on two questions: how much do customers care about it, and how closable is it in a month? "Our mobile LCP is 4.1 seconds and theirs is 1.8" is a perfect first fix. "They have 400 pages and we have 40" is not a 30-day fix; it is a content programme for the quarterly plan.
Mistakes that quietly ruin a website audit
- Auditing one page and calling it the site. The homepage is the shop window; the pricing, checkout and content pages are where customers decide. Audit the same six pages on every rival.
- Treating an estimate as a measurement. Similarweb figures are modelled. Use them to rank rivals and spot trends, never to quote a rival's exact traffic.
- Trusting the lab score and ignoring field data. A Lighthouse 95 on a fast office connection says nothing about your customers on budget phones. The Core Web Vitals Assessment box in PageSpeed Insights uses real-visitor data — that is the verdict.
- Auditing once and framing it. A single audit is a photograph; only the second one tells you whether anything you did worked.
- Copying the rival's site instead of closing the gap. Their layout and copy are answers to their costs and their customers. Steal the insight, not the design.
Cadence: how often to audit a rival's site
Run the full seven-step audit quarterly for the three to five rivals that matter, and re-check the one metric you are actively fixing monthly — speed is the usual candidate, because it moves fastest. Re-audit a single rival immediately when they relaunch, reprice, or suddenly outrank you. Between audits, keep the sheet: each rival's six numbers, the date collected, and one line on what changed.
Start this week with the smallest useful version: three rivals, six pages each, one afternoon. The first sheet will contain at least one number that surprises you — and that surprise, converted into one 30-day fix, is the entire return.