But it's not a binary choice between a) excluding everyone without LinkedIn and b) spending an infinite amount on search. Surely, there are affordable ways to evaluate engineering candidates other than their LinkedIn profiles. For example, looking at resumes alone is often sufficient to exclude 90% of applicants, simply because 90% of applicants list no relevant skills or experience.
I agree and I experienced some few exceptions to seldo thinking. For example, friends who are incredible capable and extremely focused on their field. It works for them because they are known from conferences and specific software groups.
No, at different levels it is the approach everybody takes. The concept of search cost[1] always existed in economy. You can think that an answer to your question is in some obscure web site but if that web site is in the #1000 position in Google, very few people will find it.
Regarding LinkedIn, I think if you want to increase the probability of being hired in US (and other countries where LinkedIn is popular) you must be there. You can argue against it like I argue about needing to have a Skype, or a PayPal account to receive payments: everybody is there, network effects, etc.
- [1] http://en.wikipedia.org/wiki/Search_cost