4 ms·

To take on Google people should take on a defined niche. Don't try and solve search, try to solve some subset of search. We've seen this work when people take

To take on Google people should take on a defined niche. Don't try and solve search, try to solve some subset of search. We've seen this work when people take on CraigsList and Ebay and other "unstoppable" tech companies. Don't attack head on. For example, I can think of a few popular types of searches that Google doesn't do super well: code search, product search, local search, genealogical search, real-estate search. Blekko is of course working on searching specific namespaces, but that's not what I mean. I mean taking on a single underserved domain and really making it perfect.

dizzystar 14y ago
>> We've seen this work when people take on CraigsList and Ebay and other "unstoppable" tech companies. Don't attack head on. This is good. I used to work with an industry-specific portal that was perfected to work with said industry. Google would never be able to touch this space. Despite being a smallish industry, there was two large players and few smaller players. The innovation even in such a small space was quite astounding. There are definitely some areas that are too specific for Google to really work with. It is very good as a general search engine, but if your time is dependent on getting information fast in a specific industry, Google falls flat. The main issue with this strategy is that you only have a small subset of the population and you have to an expert in many domains to get it right. thus, you'll never be as large or profitable as Google. Of course, you better have people to talk to on the phone. That'll kill this engine before it gets off the ground. The only way I can think of toppling Google is if you created an engine that really focuses on productivity and gaining market share from the people who really need information and who are willing to pay for said information and offer them a free version that is better than their industry-specific tools, then after gaining penetration and toppling some of their industry players, branch into focusing on finding cute cat pictures and the like. There is more than one way to gain mind-share. Finding a better way to find links to movie reviews is definitely not it.
anigbrowl 14y ago
Don't you run the risk of limiting yourself to a local maximum? When Google launched, Yahoo and Lycos were the dominant search engine. Google was so immediately, obviously better (in both quality of results and presentation thereof) that switching was a no brainer. If you're saying it's not worth taking them on head-on, then it's similar to arguing that their general product can't be improved upon. Of course, people said the same thing about the huge head start that incumbents had back in the 90s with building curates indices.
hnriot 14y ago
I'm not sure i agree with those as examples of things google doesn't do well. Product search works great for me, as does local. Ditto for real estate search. Maybe your experience has been different, but try adding the zip code to your search for local. I'm not entirely sure I know what exactly you're looking for in code search. For coding problems with code examples, google always finds the right SO page.

dizzystar 14y ago

>> We've seen this work when people take on CraigsList and Ebay and other "unstoppable" tech companies. Don't attack head on. This is good. I used to work with an industry-specific portal that was perfected to work with said industry. Google would never be able to touch this space. Despite being a smallish industry, there was two large players and few smaller players. The innovation even in such a small space was quite astounding. There are definitely some areas that are too specific for Google to really work with. It is very good as a general search engine, but if your time is dependent on getting information fast in a specific industry, Google falls flat. The main issue with this strategy is that you only have a small subset of the population and you have to an expert in many domains to get it right. thus, you'll never be as large or profitable as Google. Of course, you better have people to talk to on the phone. That'll kill this engine before it gets off the ground. The only way I can think of toppling Google is if you created an engine that really focuses on productivity and gaining market share from the people who really need information and who are willing to pay for said information and offer them a free version that is better than their industry-specific tools, then after gaining penetration and toppling some of their industry players, branch into focusing on finding cute cat pictures and the like. There is more than one way to gain mind-share. Finding a better way to find links to movie reviews is definitely not it.

anigbrowl 14y ago

Don't you run the risk of limiting yourself to a local maximum? When Google launched, Yahoo and Lycos were the dominant search engine. Google was so immediately, obviously better (in both quality of results and presentation thereof) that switching was a no brainer. If you're saying it's not worth taking them on head-on, then it's similar to arguing that their general product can't be improved upon. Of course, people said the same thing about the huge head start that incumbents had back in the 90s with building curates indices.

hnriot 14y ago

I'm not sure i agree with those as examples of things google doesn't do well. Product search works great for me, as does local. Ditto for real estate search. Maybe your experience has been different, but try adding the zip code to your search for local. I'm not entirely sure I know what exactly you're looking for in code search. For coding problems with code examples, google always finds the right SO page.