Baidu Follows Google and Abandons Organic Search Query Data in HTTP

In June, China’s biggest search engine Baidu officially announced organic search queries would be encrypted and no longer be carried in the http header.

Google has long implemented a similar strategy, with Bing following not long after.

Now, it is Baidu’s turn to abandon the practice.

For years, many local webmasters have used organic keyword data (i.e. search query data) as the basis for search engine optimization (SEO) and analytics.

For example, SEO reports are all built around major metrics including organic keyword ranking, organic keyword sessions and organic keyword conversions.

The change means Baidu will no longer pass users’ search queries to web servers and/or other analytics tool vendors – but instead retain this precious data for itself.

Moving the referrer data to an https is Baidu’s way of controlling this data.

HTTP Referrer

The http referrer is an optional http header field which allows the client to specify the address (i.e.URL) of the document (i.e. the web page) in which the URL in the request was obtained.

This allows a server to ‘log’ the original URLs and generate lists of backlinks to web pages. 

baidu-organic-search-http-500
For example, a user clicks a link on a web page on website X, and is taken to a web page on website Y. For this session on website Y, the http referrer is the URL of page X. 

If website Y is your site, and website X is Baidu (a search engine), the http referrer of any session should contain Baidu’s domain URL with the user’s search query tagged on the URL.

Most web analytics tools / software (including Google Analytics) have been using the information in the http referrer to show you reports with traffic sources and keywords (for search engine sessions). 

For example, most tools would have captured the referrer’s information. If we look at the header box above, we can see that: 

Source = Baidu

Medium = organic

Keyword = 微云网页版

Note the header is http, not https. 

Cases to Understand Referrers

Regarding the referrers, let’s take a look at Google Analytics reporting process as an example. 

Under the “channels” report:

Direct Traffic – referrer is blank 

Referral – referrer contains a website URL that is not a search engine 

Social Media – referrer contains a website URL that is associated with a social network site 

Display – referrer can be any site, and the referring URL must be tagged with Google Analytics default tracking parameters e.g. utm_source, utm_medium, etc. Normally these types of tagged URLs are used for online display ads or any non-search engine ads. 

E-mail – referrer should be the mail server’s domain. The referring URL must be tagged with Google Analytics default tracking parameters e.g. utm_source, utm_medium, etc. 

Paid Search – referrer contains a search engine domain, and the referring URL must be tagged with the Google Analytics default tracking parameters e.g. utm_source, utm_medium, utm_term, etc. For example, utm_medium=cpc. 

Organic Search – referrer contains a search engine domain with no Google Analytics default parameters tagged on the referring URL. 

Others – the referrer can be any site, and the referring URL is tagged with Google Analytics default tracking parameters e.g. utm_source, utm_medium, etc., but the value that is assigned to the utm_medium cannot be categorized into any of the above “channel” types. 

Baidu Referrer Channels

Baidu is the referrer in the case of organic search, in which Baidu used to pass through search query data to all web analytics tools, including Google Analytics.

However, after this http referrer change, this will no longer be the case for Baidu organic searches.

Real Case Examples with Google Analytics Organic Data

This graph shows Google Analytics data for multiple websites, which have been reviewed between January and June 2015.

baidu-organic-search-graph 

All the websites show similar results. Between January and March, all Baidu organic sessions were showing “keyword” data.

By mid-April, some Baidu organic sessions started not to show the keyword data, but they did show up as (not set). This (not set) value means in an organic search query (keyword) data is not revealed in the Google Analytics reports.

April is around the time Baidu started to enforce search from Baidu’s secured site: https://www.baidu.com

baidu-organic-search-https-500

In May, this same pattern continued. 

By the last week of June, the percentage of (not set) keyword data had increased to a new peak close to 80 percent.

After Baidu has implemented the encryption to referrer information, for example, most tools would have captured:

Source = Baidu

Medium = organic

Keyword = [Nothing]

Webmaster Tool Options 

Zhanzhang – Baidu’s official Webmaster tool – still provides Baidu’s organic search query data, but with some limitations.

For example:

• Only the past 30 days of organic keyword data is available

• Only organic keywords with the highest search are reported

• It is difficult to associate an organic keyword to a specific landing page

Web Analytics Tool Options 

Some web analytics tools will no longer be able to provide organic search query data within the http referrer.

These are:

• Free global web analytics tools e.g. Google Analytics

• Free local web analytics tools e.g. CNZZ Tongji

• Paid global web analytics tools e.g. Adobe / Omniture Site Catalyst

• Paid local web analytics tools e.g. 99Click

• Your web server’s log files

However, Baidu Tongji (a free web analytics tool provided by Baidu) can still report on organic search query data.

Goodbye Keywords, Hello Landing Pages

Another recommendation is for local webmasters to stop relying on organic keyword data and start placing landing pages in the center of their SEO analytics.

Related reading

cma feature
Search Console Search Analytics
i_fought_the_law
google amp logo
Simple Share Buttons