Home / Guides / How to check what technology a website uses
Guide · Technology checks
How to check what technology a website uses
By SearchWebCode · Updated November 18, 2026 · Detection · 2 min read
Start with one page and a specific question. Identifying a frontend framework, an analytics tag and a hosting provider are separate tasks. A useful result is an evidence log that shows what you observed and what remains unknown.
Read the original page source
Look for generator declarations, public asset paths and script URLs. Keep the surrounding tag: a real asset reference is more useful than a technology name inside an article.
A WordPress plugin path, a Next.js asset or a tracking script can suggest different layers of the page. Follow the WordPress plugin check or Next.js detection steps when those clues appear.
Inspect resources and document headers
- Open DevTools → Network before reloading.
- Select the document request and read its response headers.
- Inspect script and stylesheet requests, including their URL and initiator.
- Repeat the relevant interaction if a feature loads later.
A header belongs to the selected response. A third-party asset's headers do not describe the host of the main document. Use the Cloudflare check for delivery evidence and the HubSpot hosting explanation for CMS-versus-integration distinctions.
Classify the evidence
This illustrative log describes a fictional page. It does not assert that the markers identify every version or architecture:
| Observation | Layer | What to record |
|---|---|---|
A /_next/static/ stylesheet request | Frontend framework clue | The requested URL and response. |
A static.klaviyo.com script | Integration clue | The tag and request context. |
A cf-ray document response header | Delivery evidence | The particular response and check date. |
| A plugin path inside an HTML comment | Source mention | The comment, without calling it an active plugin. |
| No direct evidence of a database | Backend | Unknown. |
A tag may be stale, copied or loaded only under conditions you did not trigger. A runtime check can strengthen an observation but does not expose every server component. Leave an unknown field empty rather than filling it from a guess about a familiar framework.
Investigate an unfamiliar script
Search the full hostname or distinctive path in the vendor's documentation. The explainers for HubSpot scripts, Meta Pixel code and Zendesk widget assets demonstrate this approach. Identify the requested feature before inferring the website's overall platform.
Find other websites using the footprint
Copy a confirmed marker into source search. For example, /_next/static/ finds stored references to that asset path. It is a discovery query, not a live installation test.
Inspect several matches, keep capture dates with the evidence and use the reviewed-list workflow before exporting. If you compare detection products, hold the task and evidence standard constant using the comparison checklist.
References
Chrome documents Network inspection and resource search. Technology-specific guides above link to their vendors' references.
Source counts checked 2026-10-14. They describe stored homepages, not live installations. How the counts are measured.
← All guides · Browse technologies · Search and export pricing
Frequently asked questions
Can I identify a complete backend from public source?
Usually not. Browser-visible evidence exposes only part of the system. Record unsupported backend components as unknown.
Does a technology name anywhere in HTML count as installation evidence?
No. It may be a comment, tutorial or copied example. Keep the matching context and check relevant requests or runtime behavior.