Google Sitelink Search has been getting some buzz of late. Simply put, it’s a use of structured data in the HTML code to allow Google users access to your on-site search from Google’s interface. It appears for navigational, branded searches with clear dominant intent.
Here’s an example, featuring YouTube:
Moz has recently written about it, observing that in their dataset the new search box has positive correlation to increased site traffic.
But that’s their dataset.
Google does make it possible to opt out of this search box. A publisher need only insert the following code to tell Google “nope nope NOPE:”
<meta name="google" content="nositelinkssearchbox" />
And some sites are clearly and deliberately opting out. One notable site? Amazon.
Yep, right there on line 276, plain as day.
Why might they do this? I can think of several reasons. While the sitelink box will direct people into the search page on the site for the query, it bypasses all home page and category level merchandising by Amazon. Those pages are important to Amazon.
It may also create issues for them with tracking search engine referral, or affiliate referrals.
The lesson here? Your own data is the most important data. The Moz crew may see benefits to the site link search box. Meanwhile, Amazon is clearly and explicitly opting out.