If I imagine the coolest piece of software for searching for candidates on the Internet, here’s what I would like to have available. (Please note, I said searching, not sourcing; sourcing has more to it.)
The Dream Software would search for “distributed profiles”. By a “distributed profile” I mean professional information about a person collected from many sources where she has some presence, such as the evidence of her skills, experience, location, current employer, certifications, education, contact info, or any other info.
Do you cross-reference? Picking a promising profile in one place and cross-referencing elsewhere (at least Googling!) is always a good idea, because we want to be both productive and considerate and only contact “the right” people. Now, imagine all that cross-referencing work done for us ahead of time! That would be dream come true.
My dream software would find talent using data distributed across many networks and sites, and only help validate the info. I would also be able to pick the easiest way to contact the person. Say, I may find a super-C++ Engineer on his blog and send him a message on LinkedIn, where his profile has very little info besides the title and the company (and where I’d never find him by keyword searching). If the software also knew about my memberships and highlighted the links where I can easily send an introductory message… that’s probably too much to ask at this point.
It’s a hard task to design and develop this software. Here are the two major challenges, as I see them.
1. Developers have to choose between crawling the web and sending queries to sites and networks and cross-referencing on the fly. Crawling is gradual due to the vastness of info and expensive due to storage needs; sending queries is limited and slow.
2. Gluing together pieces of a distributed profile is always a challenge if the person’s name is not unique. An easy way to identify two profiles as belonging to the same person is an email address, but this doesn’t scale well.
At this point I know of two packages (both paid, am not affiliated) that do things along these “dream” lines; each chose to crawl the web:
— I’ll write in some detail about each in the upcoming posts. I would be very interested in hearing from you if you are a user of either, or if you know of another piece of a dream software.