Appen, an Australian data services company that Google contracted with for a large number of its third-party search quality raters, was notified by Google that its contract is ending on March 19, 2024. Appen said it had no prior notice and the cancellation would result in a loss of $82.8m of revenue at a gross margin of 26% for the company.
What search quality raters do. Google’s quality raters assess the quality of the Google search results. They do not directly influence the search results, and quality raters cannot downgrade or upgrade a specific site in Google Search.
Search quality rater guidelines “are used by our search raters to help evaluate the performance of our various search ranking systems, and their ratings don’t directly influence ranking,” according to Google. “The guidelines share important considerations for what content is helpful for people when using Google Search. Our page on how to create helpful, people-first content summarizes these concepts for creators to help them self-assess their own content to be successful in Google Search,” the company added.
What it means. Appen is one of a few sources that Google uses to contract quality raters. It seems, based on the almost $83 million revenue, that Google contracted Appen for a couple of thousand raters. Google has written it has about 16,000 search quality raters, so those employed by Appen represent a significant portion of the total quality raters contracted.
Is Google going to cancel its contracts with its other partners that provide search quality raters? We are not sure. Is Google going to replace these quality raters with other human quality raters from those other partners? We are not sure. Is Google going to replace all the quality raters eventually with AI instead? That is what many are speculating. We, again, do not know.
Dawn Anderson, one of our brightest contributors here, speculated these changes would happen months ago. And now with many from the Google Ads team being laid off and replaced with AI. Google following a similar playbook for assessing quality makes sense.
Why we care. Human search quality raters have been a part of Google’s process for improving search quality for well over a dozen years. Is this move a big pivot from using humans for this process to using AI for this process? It is hard to know.
Will this result in poorer search quality in the future? Again, hard to know but time will tell.
Hat tip to Cyrus Shepard for spotting this:
