Reference-led, rights-aware

How to copy a website design without copying what is not yours

Rebuilding your own site on a new stack is ordinary work. Lifting somebody else's words, images and identity is not. The line should be clear before a copier touches the URL.

Quick answer

Only copy a site and its assets when you own them or have permission to reuse them. For third-party inspiration, extract broad ideas and familiar layout patterns, then write new copy, use original or licensed assets, choose distinct branding and build the implementation yourself. When rights are unclear, stop and get permission or legal advice.

Reviewed for material changes on August 13, 2026

Start with permission, not the screenshot

The clean cases are straightforward: your company owns the existing site, the client has authorised the rebuild and confirmed its rights, or a licence expressly allows the use you plan. Put that authority in writing. Name the domain, materials, permitted purpose and any assets that are excluded. A message saying 'use this for inspiration' is not automatically permission to republish every photograph and paragraph.

If the reference belongs to somebody else, do not feed it into a copier to produce a near-duplicate. Study the underlying problem it solves — navigation, hierarchy, content sequence, responsive behaviour — and create a distinct expression. Publicly visible does not mean free to reproduce, and downloading a page does not transfer its rights.

  • Owned: confirm the company owns the code, copy and commissioned assets, not merely the domain.
  • Client work: obtain written authority and ask the client to identify third-party licences.
  • Template or stock work: read the licence for redistribution, client transfer and derivative use.
  • Third-party inspiration: take notes on ideas and patterns; do not create a substitute copy of the site.
  • Unclear: pause and ask the rights holder or a qualified lawyer.

Separate the design idea from the protected expression

Copyright generally does not protect a bare idea. Familiar conventions such as a navigation bar, hero, testimonial row or three-column price table are building blocks used across the web. That does not make the exact page fair game. The particular text, photographs, illustrations, video, music and source code can be protected, and an original selection or arrangement can matter too.

The U.S. Copyright Office says a website itself is not typically treated as a copyrightable work, while the text, artwork, photographs and other content on it may be copyrightable. It also says website or page material may qualify as a compilation or collective work. Other countries differ, and copyright is not the only right in play. Logos, names and source-identifying design can raise trademark or passing-off questions even where a simple layout is commonplace.

The practical rule is stricter and easier to operate: use a third-party site to understand a pattern, not to reproduce its recognisable expression. Change the information architecture where your content calls for it, develop your own design tokens, write every sentence anew and use assets with a clear provenance.

Make a rights and component inventory

Before rebuilding an authorised site, crawl it manually or with an approved tool and list the pages, components, behaviours, fonts, icons, photographs, embeds and downloads. Add an owner and licence column. This catches the stock image bought by an old agency, the font licensed only for the former host and the customer logo approved for one case study but not every page.

Create a second inventory for the design system: colours, type roles, spacing, container widths, borders, radii, shadows, breakpoints and interaction states. The point is not to collect thousands of pixel coordinates. It is to identify the small set of rules that explains the site.

  • Page or component and its business purpose.
  • Content owner and approval status.
  • Asset source, licence and any expiry or channel restriction.
  • Desktop, tablet and mobile behaviour.
  • Forms, accounts, payments, search and other backend dependencies.
  • Third-party scripts, embeds and data flows that need fresh consent or configuration.

Rebuild the system, not a pile of screenshots

Start with semantic page structure and reusable components. Define the type scale and spacing rules, then build the header, footer, buttons, cards, forms and content sections. A screenshot can show one viewport; it cannot tell you the intended reading order, keyboard behaviour, loading state or what happens to a long headline on a narrow phone.

Use screenshots as a visual check, not as the implementation. Preserve text as text, controls as controls and headings in a sensible document order. Match the approved reference at representative widths, then test the widths between them. A faithful rebuild should be more maintainable and accessible than a traced bitmap.

Replace copy, assets and branding deliberately

For an owned migration, reuse only the material confirmed in the inventory. For an inspiration-led project, replace all brand names, logos, product screenshots, customer marks, copy, photography, illustration, icons and distinctive decorative artwork. Do not keep a competitor's typo, testimonial or legal footer as a placeholder; placeholders have a habit of reaching production.

Write from your own offer and evidence. Choose a distinct palette and type pairing, and adjust the component proportions and content sequence to suit the new information. Search the proposed name and marks before launch. The USPTO notes that marks can be confusingly similar in sound, appearance, meaning or overall commercial impression; exact identity is not required.

A visible page does not reveal the backend

A copier can infer interface and front-end behaviour. It cannot recover a private database, payment configuration, account rules, email automation, recommendation model or internal admin process from the public page. Recreate those systems from requirements and authorised integrations. Never copy credentials, analytics IDs, tracking pixels or private endpoints from source material.

List each interactive path and assign a real destination. Decide where forms go, how errors appear, what data is stored, when consent is required and who can access it. Run payments in test mode and use new production keys. A page that looks right but sends enquiries nowhere is not a copy; it is a broken photograph.

Test fidelity, difference and provenance

For an authorised migration, compare the old and new site at the same viewport sizes and with the same approved content. Check line wraps, spacing, focus states, forms, redirects and analytics. Record intentional differences so a stakeholder does not report them as defects later.

For an inspiration-led build, run a different review: could an ordinary visitor mistake the new site for the reference brand, and have any unlicensed words or assets survived? Keep the permission, asset register, licence receipts and approvals with the project. If a source has no recorded right, remove or replace it before launch.

  • Visual comparison at agreed desktop and mobile widths.
  • Keyboard, screen-reader labels and reduced-motion behaviour.
  • Forms, validation, redirects, checkout and error states.
  • Search metadata, canonical URLs and migrated redirects.
  • Rights review for every public word, image, font, icon and mark.
  • Written sign-off from the owner or authorised client.

Sources and checks

Product limits and prices came from the companies themselves:

Plans move. We date every check so you know when to verify again.

Straight answers

Questions people actually ask

Is it legal to copy a website design?

There is no blanket yes. Rebuilding a site you own or are authorised to use is the clean case. With a third-party site, general ideas and common patterns are different from copying its text, images, code, distinctive arrangement or branding. Laws vary, so get legal advice when the planned similarity or commercial risk is meaningful.

Can I copy a website if I change the logo and colours?

Changing a logo and palette does not clear copied text, photographs, illustrations, code or a recognisable overall execution. Build from your own content and rights-cleared assets, and make the expression genuinely distinct.

Can I copy the layout of another website?

You can learn from commonplace patterns such as headers, cards and pricing tables, but avoid tracing a distinctive page section by section. The exact selection and arrangement may matter, and trademark or passing-off rules can apply beyond copyright.

Can a website copier copy the backend too?

No responsible copier can recover a private backend from a public page. Databases, accounts, payments, automations and internal rules must be rebuilt from authorised requirements and connected with new credentials.

Can I copy my own website to a new platform?

Usually, provided you own or have transferable rights to the code, copy and assets. Check agency contracts, template terms, stock licences, fonts and third-party embeds before assuming everything on the old domain belongs to you.

Does fair use let me copy a website for my business?

Do not treat fair use as a percentage rule or a general business-copying licence. The U.S. Copyright Office says fair use depends on all the circumstances and four statutory factors. Get specific legal advice rather than building a commercial launch around an assumption.

Skip the blank page

Already know which website you like?

Paste the link. Copycat turns it into an editable Next.js first draft in under 15 minutes.

Copy a website for $100