Yandex source code leak: what it reveals about ranking factors
Short answer In January 2023, Yandex's internal code leaked, including lists of ranking factors from 2022. It confirmed what search engines had long said: links, matching words, clicks, freshness and original content all count. Some leaked factors, like a .com bonus, are Yandex quirks Google denies. For a small business, the lesson is the basics, not a new trick.
Want help applying this to your business? Talk it through on a free 30-minute call.
Book a callWhat leaked, and is it real?
Yes, the code is real, but it is old. In late January 2023, a 44.7 GB archive of Yandex code appeared as a torrent on a hacking forum. The leaker said it was taken in July 2022 and held all of Yandex's code except its anti-spam rules (BleepingComputer). Yandex, Russia's largest search engine, confirmed the code came from its internal repository but denied a break-in:
"Yandex was not hacked. Our security service found code fragments from an internal repository in the public domain, but the content differs from the current version of the repository used in Yandex services." (Yandex statement, quoted by BleepingComputer)
BleepingComputer reported that Yandex blamed a former employee. Yandex said it was running an internal investigation.
Two numbers float around:
| Number | What it is |
|---|---|
| 1,922 | One file of ranking factors, shared first. Of these, 244 were marked unused and 988 deprecated (replaced), leaving about 690 that might still work (Dan Taylor, quoted by Search Engine Land) |
| 17,854 | All ranking-factor files in the code combined, found by Michael King's deeper review (Search Engine Land) |
A ranking factor is any measurement a search engine uses to decide which page shows first. A long list doesn't mean each one matters much. Most get tiny weights.
Which leaked Yandex factors also apply to Google?
The ones Google confirms on its own pages. Google never comments on Yandex, but it publishes how its own systems work. Here are the leaked factors that got the most attention, next to what Google says:
| Leaked Yandex factor | What Google says | Applies to Google? |
|---|---|---|
| PageRank: a score from the links pointing to a page. Yandex's 17th heaviest-weighted factor (King) | PageRank "continues to be part of our core ranking systems" (Google ranking systems guide) | Yes |
| Penalty when over half of a site's links use sales-style keyword anchor text (King) | Buying links or trading them for ranking is link spam (Google spam policies) | Yes, in spirit |
| Text matching (BM25, TF-IDF, BERT: methods that score how well page words match a search) | "The most basic signal that information is relevant is when content contains the same keywords as your search query." (Google: ranking results) | Yes |
| Clicks and time on page, partly from Yandex's own analytics tool (King) | Google uses "aggregated and anonymized interaction data to assess whether search results are relevant" (Google: ranking results) | Yes, in aggregate |
| Duplicate text pushed down; original authors credited (King) | Google runs deduplication and original content systems (Google ranking systems guide) | Yes |
| Bonus for a .com domain; bonus for not being .ru (King) | The domain ending "only matters if you're targeting a specific country's users," and even then is "usually a low impact signal" (Google SEO Starter Guide) | No |
| Bonus when the domain name matches the search words (King) | Google's exact match domain system stops giving "too much credit" to such domains (Google ranking systems guide) | Mostly no |
| Penalty for any ads on the page, and more for Yandex's own ads (King) | Google says results "are not influenced by whether a business, individual, or organization buys ads" (Google: ranking results) | No evidence either way about ads on the page |
| Factors that appear to boost certain news agencies (King) | Nothing comparable in Google's public documentation | Can't act on it |
The pattern: where the leak matches Google, it matches what Google already tells site owners. The surprises are mostly Yandex-only.
What do the leaked factors mean for a small business website?
Nothing new to do. The useful factors point to the same short list Google publishes:
- Earn real links. A few links from local and trade sites beat many bought ones. Keep anchor text natural: your business name, not "best plumber Dallas" every time. How: backlinks for a local business.
- Use the customer's words. One page per service, titled the way people search. Details: on-page SEO for service pages.
- Write what only you can. Copied or spun text gets filtered out by both engines. Use your own photos, prices and job details.
- Update pages when facts change. Google weighs freshness only for searches where new information matters, like news or a new product (Google ranking systems guide). When to update, merge or delete: content decay.
- Make the click worth it. Both engines watch how searchers react. A page that answers the question keeps people from bouncing back to the results.
Myths about the Yandex leak to ignore
- "Buy a .com to rank better." That was a Yandex weight. Google says the domain ending barely matters. Keywords in the domain: exact match domain SEO.
- "Running ads boosts your organic ranking." The leaked weights did the opposite: pages with ads lost points. Google says paying for ads doesn't change your organic ranking.
- "Yandex penalizes content older than 10 years." The age factor gives 1 to a page dated today and 0 to one 10 or more years old. Its weight was negative, which King read as Yandex leaning toward older pages, not newer ones. Either way, Google's freshness rule is about the search, not a date cutoff.
- "These 17,854 factors are how Google ranks." They are Yandex's, from 2022, and Yandex said the code was already out of date when it leaked.
Example
A roofing company deciding on a domain
A Toronto roofer reads that Yandex gave .com domains a bonus and considers moving from a .ca domain. Google calls a country ending like .ca "a strong signal" that a site is meant for that country (Google Search Central: multi-regional sites). Moving would trade that signal for a Yandex quirk. They keep the .ca.
A dental office with 40 old blog posts
The office hears "old content gets penalized" and plans to re-date every post. Google's own self-check asks whether you are "changing the date of pages to make them seem fresh when the content has not substantially changed" (Google Search Central: helpful content). They update the 5 posts that get traffic and have wrong prices, and merge the rest.
A landscaper offered 100 links for $200
A seller cites the leak to prove "PageRank still works." Links do count, but paid links are link spam under Google's rules. The landscaper asks two suppliers to list them on their dealer pages instead.
FAQ
Was Yandex hacked?
Yandex says no. It said fragments of its internal code appeared in public and it saw no threat to user data. It blamed a former employee (BleepingComputer).
Does Google use the same ranking factors as Yandex?
Some of the same kinds: links, word matching, interaction data, freshness, original content. Google confirms those on its own pages. The exact weights in the leak are Yandex's and don't transfer.
Did Google's ranking factors leak too?
Internal Google documents surfaced in May 2024. Google confirmed they were real and warned against "making inaccurate assumptions about Search based on out-of-context" information (Search Engine Roundtable).
Should a US or Canadian business optimize for Yandex?
Only if you sell to Russian-speaking markets where Yandex is used. For customers in the US and Canada, work on Google and Bing.
Different for you?
- Your traffic dropped and you suspect an algorithm change: the leak won't explain it. Start with traffic dropped after a Google core update.
- You got a manual action for links: see Google manual action.
- You want the basics done every month without learning the factor lists: SEO agent.