The news is just in that Microsoft is announcing that it will delete full IP addresses from their Bing search engine log files after 6 months. In July 2009, when Microsoft and YAHOO announced their search partnership, we predicted that real competition in the search arena between Google and Microsoft could lead to privacy benefits for consumers. Today, we are pleased to see Microsoft stepping up to the plate to delete one of the most sensitive and potentially identifiable elements in search log-files.
There is really no need to keep IP addresses long term to improve search results, when the primary geographic data can be immediately derived for use in search query analysis. At the same time, an IP address can be used by governments or legal adversaries as a key to a user’s identity. Search engines do need to keep IP addresses for a limited time to help with security and anti-fraud measures.
Here is a Microsoft chart showing the new Bing search data retention periods:
Yahoo’s current policy is to delete IP addresses at 3 months, with a small subset kept for 6 months for security uses. Google partially deletes IP addresses at 9 months by lopping off the last octect of the address.
Today’s announcement doesn’t affect practices around other cross-session identifiers, such as the hashed cookies maintained in search log files, so there is still work to be done in this area by Microsoft and Google. But although a cookie enables potential research that can assemble enough data about some random selection of users who have provided enough clues due to their queries, it is the more sensitive IP address that is the key to being able to force the identification of any individual users. Kudos to Microsoft for this important search privacy advance.