The complete guide to competitor teardowns
By Owen Fisher·Updated 24 July 2026·14 min read
A competitor teardown is a structured study of a rival's website: what pages they have, what each page says and shows, and what that means for your own product. This guide covers the five pages worth studying, how to capture clean evidence, and how to present findings, with or without a tool like Stravue.
What is a competitor teardown?
A competitor teardown takes one rival's website apart, page by page, to understand the decisions behind it. Not their traffic, not their revenue: the site itself. What they lead with. What they charge. How they explain the product. What they ask for at signup. It is the cheapest competitive research there is, because everything you need is public.
Teardowns earn their keep when they end in a deliverable. A designer uses one to argue for a redesign direction. An agency uses one to open a pitch. A product marketer uses one to answer "what did they change this quarter?". In every case the output is the same: annotated screenshots, in order, with a point of view attached.
What a teardown is not
- Not analytics. You cannot see a rival's conversion rate, and a teardown should never pretend to. It reads intent from design decisions, which is a different and honest claim.
- Not user testing. Watching real people use a competitor's site is a separate method with separate costs. A teardown is your expert read, not a usability study.
- Not scraping. A teardown studies public marketing pages. It does not log in, harvest data or go behind paywalls.
The five-page teardown
Most competitor sites answer five questions, on five kinds of page. Study these five and you have the shape of their whole strategy. This is the five-page teardown, and it works on almost any B2B or consumer software site.
| Page | The question it answers | What to look for |
|---|---|---|
| Homepage | Who is this for and what is the promise? | The headline's audience, the first visual, the primary call to action, what is above the fold on a phone |
| Pricing | How do they charge and who do they want? | Tier names, the featured plan, what is gated where, free trial or demo, annual discount framing |
| Product or features | What do they lead with and what is buried? | Feature order, screenshots versus illustrations, which capability gets a whole page |
| Signup flow | How much friction stands before value? | Fields asked for, steps to first value, credit card up front or not, what the empty state teaches |
| Proof | Why should anyone believe them? | Logos, case studies, review badges, security pages, named customers versus vague claims |
Capture each of the five on desktop and on a phone. Half of most sites' traffic is mobile, and the phone version often tells a different story: shorter headlines, cut sections, a different call to action. If the two versions disagree, that disagreement is a finding.
Desktop
Phone
How to run a teardown, step by step
- 1
Write the question first
One sentence, before you open a tab. "How does their pricing page justify the jump to the top tier?" beats "look at their site". The question decides which pages matter and when you are done.
- 2
Capture before you judge
Grab clean full-page screenshots of the five pages, desktop and phone, before forming opinions. Sites change without notice, so capture first and the evidence holds still while you think.
- 3
Annotate what you see, in place
Pin notes directly on the screenshots. "CTA repeats four times", "no pricing until the footer", "trust logos above the fold". A note on the pixel it describes beats a paragraph in a separate document.
- 4
Separate observation from interpretation
"The featured tier is the most expensive" is an observation. "They are steering buyers upmarket" is an interpretation. Keep both, but label them. Your credibility in the room rests on the audience knowing which is which.
- 5
End with three moves, not thirty notes
Finish the teardown by answering your opening question and proposing at most three actions. A teardown that ends in a to-do list of thirty items changes nothing.
How to capture the evidence
The capture step is where teardowns die. Doing it by hand means opening every page, screenshotting section by section, stitching the pieces, then repeating everything at phone width. For a five-page teardown of one competitor, that is an hour of tab work before any thinking starts.
| Method | Good for | The catch |
|---|---|---|
| Browser dev tools | One page, occasionally, free | Manual per page, no phone-plus-desktop pairing, cookie banners land in the shot |
| Screenshot extensions | A handful of pages, free | Still one page at a time, quality varies, nothing is organised afterwards |
| Screenshot APIs | Developers automating captures | You write and maintain code, then still need to assemble the deck yourself |
| Stravue | Teardowns that end in a deck | Paid, from $39 a month, and it deliberately leaves the analysis to you |
Stravue's version of this step: paste the competitor's address and what you're after, and it finds and captures the right pages, full-page, desktop and phone, with cookie banners and chat widgets removed. The screenshots land on a board ready to pin notes on, and the deck exports to PDF, PowerPoint or a share link. See how the capture step works or the worked example of a full teardown.
How to present a teardown
Nobody reads a forty-row spreadsheet of observations. The format that survives contact with a meeting is a short deck: one screenshot per slide, your annotation on the image, one takeaway line underneath. Ten slides is plenty. Full structure and slide order are covered in how to present competitor research.
- Open with the question and the answer. Slide one states what you set out to learn and what you found.
- Show, then tell. The screenshot carries the claim; the caption carries the meaning.
- Pair desktop and phone when they disagree. The mismatch is often the most quotable slide.
- Close with the three moves. Every recommendation points back at a slide someone just saw.
How often should you tear down competitors?
A deep teardown of your top three rivals twice a year is a reasonable baseline for most teams. Between teardowns, watch the pages that move: pricing, homepage, feature announcements. Checking by hand every Monday does not survive a busy quarter, which is why monitoring competitor changes is worth setting up once and letting run.
Common teardown mistakes
- Touring instead of answering. Screenshotting everything with no question produces a scrapbook, not a study.
- Copying the leader. "They do it, so should we" ignores that their choices serve their strategy, segment and scale, not yours.
- Trusting memory. "Their pricing used to be simpler" is a guess without a dated capture. Evidence beats recall.
- Desktop only. If you never captured the phone version, you studied half the site.
- Presenting notes instead of a narrative. Findings that are not ordered into an argument get politely ignored.