| --- |
| license: other |
| license_name: http-archive-derived |
| license_link: https://huggingface.co/datasets/pracadata/website-technologies#source-and-licence |
| language: |
| - en |
| pretty_name: Website Technologies, Monthly (derived from HTTP Archive) |
| size_categories: |
| - 100K<n<1M |
| task_categories: |
| - tabular-classification |
| tags: |
| - web |
| - technology-detection |
| - tech-stack |
| - http-archive |
| - websites |
| - market-intelligence |
| - tabular |
| configs: |
| - config_name: tech_adoption_monthly |
| data_files: data/tech_adoption_monthly/*.parquet |
| default: true |
| - config_name: technology_catalog |
| data_files: data/technology_catalog/*.parquet |
| - config_name: tech_by_domain |
| data_files: data/tech_by_domain/*.parquet |
| - config_name: tech_changes_monthly |
| data_files: data/tech_changes_monthly/*.parquet |
| --- |
| |
| # Website Technologies, Monthly (derived from HTTP Archive) |
|
|
| **Which technologies popular websites use, month by month**: CMS, e-commerce platform, analytics, tag manager, CDN, JavaScript frameworks, payment and marketing tools, with the version when the page exposes it. Detection is done by the [HTTP Archive](https://httparchive.org) with a real browser, on desktop and mobile, for the origins in the Chrome UX Report (16,119,182 origins in the October 2026 crawl, 3,619 distinct technologies). **This repository is a free sample of that data**; build of 27 October 2026, refreshed monthly. Latest crawl: **October 2026**. |
|
|
| > **The full data is on the [Apify Store](https://apify.com/pracadata/website-tech-stack)**: lookup of any domain in the crawl, **reverse search** ("which sites use Klaviyo?", at any popularity level), and the **complete change history** between monthly crawls. Pay per result, no live crawling, MCP-ready for AI agents. This dataset is the "try before you pay": same schema and same quality, narrower slice. |
|
|
| ## Read this first: what the sample contains |
| | Config | What it is | Rows | |
| |---|---|---| |
| | `tech_adoption_monthly` | **Complete.** Technology × category × popularity bucket × crawl month: how many origins use it and what share. Aggregate, no domains. Crawls 2026-08 to 2026-10 | 86,442 | |
| | `technology_catalog` | **Complete.** Technology name, categories, first and last crawl seen, origins in the latest crawl | 3,619 | |
| | `tech_by_domain` | **Slice.** Origin × technology for the **top 10,000 most popular origins** (CrUX buckets up to 10k), **October 2026 crawl only** (6,763 origins) | 121,423 | |
| | `tech_changes_monthly` | **Slice.** Changes between 2026-09 and 2026-10 for the **top 10,000 origins**, only changes **both desktop and mobile agree on** (`confidence = both_clients`) (2,540 origins) | 6,016 | |
|
|
| - **Not in this sample** (Apify only): domains below the top 10k (the crawl has 16,119,182), older crawls of `tech_by_domain`, the full change history, changes seen by only one client, and the reverse index. A missing origin here means nothing: it is just outside the slice. |
| - **Adult and gambling websites are left out of the two by-domain tables** by a keyword filter on the host name (612 origins in the top 10k in this build). It is best effort: a keyword filter misses some sites and wrongly drops a few others. The aggregates and the catalog are not filtered. |
| - **Home page only.** Detections come from each origin's root page. A technology used only on inner pages (checkout, blog) can be missed. |
| - **About 4,000 technologies** (the HTTP Archive's open-source fork of Wappalyzer). Commercial databases track far more. No detection does not prove that a site does not use a technology. |
| - **Detections flicker.** A technology can be detected in one crawl and not in the next without the site changing (A/B tests, consent banners, lazy loading, timeouts). In our measurements, even among changes that desktop and mobile agree on, 9% to 12% of `added` and `removed` changes reverse in the following month. `removed` means "not detected in this crawl", not "uninstalled". |
| - **One crawl per month.** "Left the crawl" (an origin that is missing from a month) is not "removed the technology": origins missing from either month never appear in `tech_changes_monthly`. |
| - **`rank_bucket`** is a Chrome UX Report bucket: `1000` is the top 1k origins, then 5k, 10k, 50k, 100k, 500k, 1M, 5M, 10M, 50M. In `tech_adoption_monthly` it is **cumulative** (the 10k bucket includes the 1k and 5k origins). An origin is `https://www.example.com/`, so `example.com` and `www.example.com` can be two origins. |
| - **A technology can have several categories** and appears once per category in `tech_adoption_monthly`: filter by `category` before summing. |
| - **Versions are best effort.** `versions` comes from what the page exposes; some values are wrong. `version_major` keeps only a leading integer. |
| - **Monthly, with lag.** The HTTP Archive crawl starts on the second Tuesday of each month and takes 1 to 2 weeks; this dataset is built after the month's data is final. Crawls run from US Google Cloud locations. |
|
|
| ## Examples |
| Share of the top 10k origins on Shopify, month by month (DuckDB): |
| ```sql |
| select crawl_date, round(100 * share_of_origins, 2) as pct, origins |
| from 'hf://datasets/pracadata/website-technologies/data/tech_adoption_monthly/*.parquet' |
| where technology = 'Shopify' and category = 'Ecommerce' and rank_bucket = 10000 |
| order by crawl_date; |
| ``` |
| Top-10k origins that switched on Klaviyo in the latest month (both clients agree): |
| ```sql |
| select origin, rank_bucket, new_versions |
| from 'hf://datasets/pracadata/website-technologies/data/tech_changes_monthly/*.parquet' |
| where technology = 'Klaviyo' and change_type = 'added' |
| order by rank_bucket, origin; |
| ``` |
| The full stack of one site in the latest crawl (any site of the top 10k): |
| ```sql |
| select technology, categories, versions, detected_desktop, detected_mobile |
| from 'hf://datasets/pracadata/website-technologies/data/tech_by_domain/*.parquet' |
| where origin = 'https://www.shopify.com/' |
| order by technology; |
| ``` |
| ```python |
| from datasets import load_dataset |
| ds = load_dataset("pracadata/website-technologies", "tech_adoption_monthly") |
| ``` |
|
|
| ## Source and licence |
| - **Derived from the HTTP Archive** (<https://httparchive.org>) monthly crawls, accessed through the public Google BigQuery dataset `httparchive.crawl.pages`. **Every row carries this attribution in the `source` column** ("Derived from the HTTP Archive (October 2026 crawl)"). |
| - **Technology detection** is done by the HTTP Archive's open-source fork of Wappalyzer (<https://github.com/HTTPArchive/wappalyzer>). Only the detection **results** are published here; the detection rules, icons and descriptions are **not** redistributed. Technology names are trademarks of their owners and are used only to identify them. |
| - **Popularity buckets** come from the Chrome UX Report (Google), as included by the HTTP Archive. |
| - **Licence: none published by the source.** The HTTP Archive publishes no formal licence for its data. Its maintainers have said in public that "people are free to use the data however they want" and that attribution is appreciated ([discussion](https://discuss.httparchive.org/t/what-are-the-rights-licensing-for-data-in-http-archive/1924)). We do not claim a licence for this dataset: the licence field is `other` and means exactly that. We have asked the HTTP Archive for written confirmation and have not received an answer yet. Use it with attribution and at your own judgement. |
| - **Not affiliated with or endorsed by** the HTTP Archive, Google or Wappalyzer. |
| - **Removal requests.** Write to `lucashpissaia@gmail.com` with the origin or domain. It goes into the opt-out list (`transformacao/seeds/optout_tech_stack.csv`) that every build applies before anything is published, so it disappears from the next build; copies already downloaded cannot be reached. If the HTTP Archive asks us to stop, this repository goes private and stops updating. |
|
|
| ## Privacy |
| This dataset describes **websites and the software they run**, not people. It has only the root URL of each origin, its registrable domain, a popularity bucket and the detected technologies: no page content, no inner URLs, no cookies, no WHOIS, no e-mail addresses and no phone numbers, and it will not be enriched with contact data. An origin can still be a personal website: use the removal address above. |
|
|
| ## How to cite |
| ```bibtex |
| @dataset{website_technologies, |
| title = {Website Technologies, Monthly (derived from HTTP Archive)}, |
| author = {pracadata}, |
| year = {2026}, |
| url = {https://huggingface.co/datasets/pracadata/website-technologies}, |
| note = {Derived from the HTTP Archive monthly crawl (httparchive.crawl.pages); detection by the HTTP Archive fork of Wappalyzer; rank buckets from the Chrome UX Report} |
| } |
| ``` |
|
|
| ## Terms |
| - Provided **as is, without warranty**, with the limitations above. Cite the HTTP Archive as the source if you publish results. |
| - Questions or corrections: open a discussion on this repository. |
|
|
