I’ve been working with a business partner developing a plug-in for Microsoft Outlook as a bit of a learning/side project/commercial enterprise and an idea arose for a second app idea – completely unrelated to the plug-in.
Mobility and talent sourcing/utilisation are huge issues for organisations so we did a bit of digging around as part of preliminary research – what is out there and what kind of platform would work best. Could make a good app – internally – for organisations, particularly in professional services.
So we considered Google search; LinkedIn search and Advanced People Search; SharePoint; Facebook; and an online dating site – the former for obvious reasons, the latter based on the buzz their algorithms generated in the tech and online psychology arenas.
The clear winner in terms of speed and accuracy was LinkedIn – likely its combination of dense user-generated data, user-generated natural language search courtesy of the “endorsement” feature; and its search engine.
The other methods performed OK – Facebook performed poorly as expected, for the obvious reason of generally smaller sets of people and its not being a professional network. Google was obviously slightly out of scope but did deliver third party resources as well as plenty of links to LinkedIn (the key being add “Linkedin” to the search terminology). SharePoint was hard to judge due to access issues, but we assumed would perform somewhat less well than LinkedIn, within and organisation, due to challenges in getting categories and tags correct and getting users to complete profiles etc.
The surprise was the online dating site, which we thought would be a match made in heaven (pardon the pun). There was a lot of buzz several years back about the powerful algorithms this multi-million dollar industry was using. The app idea had been based on using that kind of back end as a basis but make the focus finding people and skills rather than ‘dates’. A brilliant idea: just repurpose it! But based on the performance (and we think we found out why – for reasons like this) it’s amazing the species can procreate, let alone use the tool for skill seeking. [It was an interesting experience … but that’s a story for a different day.]
Still – there are learnings to be had that apply to all social media. In passing, Shel Holtz recently put up a brilliant blog posting about conventional wisdom being anything but when it comes to social media and many of these things rang true in this very unscientific study.
The long and short of it is that, blindingly obvious as it might sound, if you want to find a specific set of skills within a defined universe and even geography – regardless of the platform you are using – you can’t rely on one way to find them. The devil is in the user profiles – a combination of tags, categories but most importantly user-generated, natural-language data.