My profile on Linked-In is only 85% complete. That’s largely because I haven’t gone to the trouble yet to ask someone to write a recommendation for me. It didn’t seem worth it to me, because as Daniel Tunkelang notes, the recommendations–solicited and okayed by the beneficiaries–often read like the effluvia of a mutual admiration society.
I was talking about this a while back with Reid Hoffman, the founder of LinkedIn. His profile features 49 recommendations. The first three say he’s a ‘brilliant strategist’, ‘fantastic to work with’ and has ‘tremendous raw intelligence.’ Others pile on more praise. So I asked Reid what value these recommendations have.
‘Did any of them mention that I was well organized or a good manager?’ he asked. He said this while we sat in his office, which was piled with papers, with empty Amazon book boxes strewn on the floor and cables winding this way and that. His point was that the recommendations can provide insights if you’re clever enough to look for what’s missing.
This may seem like a round-about way to read recommendations. But it would be easy enough for algorithms. Imagine charts that represent all of the qualities of a professional, …
My profile on Linked-In is only 85% complete. That’s largely because I haven’t gone to the trouble yet to ask someone to write a recommendation for me. It didn’t seem worth it to me, because as Daniel Tunkelang notes, the recommendations–solicited and okayed by the beneficiaries–often read like the effluvia of a mutual admiration society.
I was talking about this a while back with Reid Hoffman, the founder of LinkedIn. His profile features 49 recommendations. The first three say he’s a ‘brilliant strategist’, ‘fantastic to work with’ and has ‘tremendous raw intelligence.’ Others pile on more praise. So I asked Reid what value these recommendations have.
‘Did any of them mention that I was well organized or a good manager?’ he asked. He said this while we sat in his office, which was piled with papers, with empty Amazon book boxes strewn on the floor and cables winding this way and that. His point was that the recommendations can provide insights if you’re clever enough to look for what’s missing.
This may seem like a round-about way to read recommendations. But it would be easy enough for algorithms. Imagine charts that represent all of the qualities of a professional, from creativity and teamwork to organization and leadership skills. If programs could mine the qualities mentioned in recommendations, they could create heat maps for each of us, each one showing where we’re hot–and where we’re not. Might be something for the R&D folks at LinkedIn.
