> The results keep getting "refined" so as to suit the popular 80% of queries, while getting much worse for any technical or obscure queries.
The fundamental, unavoidable problem is that the cost of providing high-quality results on the long-tail of possible searches tends to grow faster than the revenues that can be earned from those increasingly rare, obscure, long-tail searches. Any search service seeking to maximize profit, like Google or DDG, ultimately always evolves to perform less and less well on the long tail of possible searches.
The search service we all wish we could have -- a service seeking to maximize the quality of individual searches, no matter how obscure -- may not be feasible as a profit-maximizing business.
The fundamental, unavoidable problem is that the cost of providing high-quality results on the long-tail of possible searches tends to grow faster than the revenues that can be earned from those increasingly rare, obscure, long-tail searches.
I think even two years ago, Google searches had far more depth and yet Google was quite profitable (then the searches were still biased but now stuff is simply gone). Sure, if someone looked at the marginal profitability of every single search result, it would look like what we're seeing. But there was a time when good indexing of stuff that didn't turn a profit by itself was done as a service to attract people to Google and/or to improve the Internet generally. That time has passed, clearly but it was a decision.
I agree. In all likelihood, the decisions were made gradually to improve overall efficiency without losing search volume, but the unintended consequence was to degrade search quality, at first gradually in subtle ways, and then suddenly in very noticeable ways.[a] It's possible no one in the company's executive team has noticed the loss of quality. In fact, they may not think anything's wrong even now.
The fundamental, unavoidable problem is that the cost of providing high-quality results on the long-tail of possible searches tends to grow faster than the revenues that can be earned from those increasingly rare, obscure, long-tail searches. Any search service seeking to maximize profit, like Google or DDG, ultimately always evolves to perform less and less well on the long tail of possible searches.
The search service we all wish we could have -- a service seeking to maximize the quality of individual searches, no matter how obscure -- may not be feasible as a profit-maximizing business.