Who will bring ML to Search?


In a previous post I asked “How to bring ML to Search?” In this post I want to discuss which companies could likely bring machine learning to search engines. Some ML techniques are already used for Web search, but search engines fall behind modern possibilities by far. When I speak of ML in this post, I also mean related NLP techniques.

The benefits of bringing ML to search are many: better spam fighting, a replacement for PageRank, higher quality and more varied results on the first page and search term disambiguation, among others.

So who are the players in the search arena? First there is Google, still the #1 as of this year, but gradually losing its foothold. Next come other incumbent Web search engines, like Bing and DuckDuckGo, followed by regional offerings of similar quality. Another big group is the many startups worldwide which have created, or are in the process of creating, search engines from scratch. Finally there is the possibility of entirely new entrants which haven’t been founded yet.

So what is keeping each of these groups from developing and launching ML features in search? Google most likely doesn’t want to disturb its dominant position by upending the market. Web search is very profitable for Google, and they aren’t forced to really innovate. Other large, incumbent players would benefit much more from introducing ML capabilities by taking over the #1 spot. Bing seems to be stuck too much on emulating Google to really innovate. DuckDuckGo is working mostly on privacy features, which is a different avenue than going after ML.

Startups would benefit greatly from ML features, by standing out from the crowd fairly early on. They are probably held back by the great cost, compared to their size. They would have to pay for development of ML features and pay for additional compute capacity, both from venture funding. A newly founded startup in the search space would have to develop both the search engine infrastructure, such as for crawling, and ML features on top of this.

I believe that whoever manages to integrate modern ML techniques into Web search, whether current or future companies, has a realistic chance of taking the #1 spot in this segment, which would make the company very valuable. And that is my startup idea for today.