predictive recruiting

Will HR use predictive data to cross the creepy line?

on March 31 | in Human Resources, Recruitment, Social Recruiting, Software Development, Sourcing, Start-ups | by | with 1 Comment

I love exploring new software tools and sites – particularly those that can be repurposed for things other than their original intention and oftentimes for better value. This time round, I’m talking about a site originally targeted at recruiters, but which I think potentially offers better value to HR people – if they’re willing!

Signal is a beta project, coming out of the Joberate team in Finland. Lead by serial entrepreneur Aki Kakko, the origins of the project lay, I believe, back in an unconference discussion about predictive data at a Tru event. One short hack project later and a prototype was being kicked around. Development continues with both commercial models and potential crowd-funding under discussion.

What does it do? It looks for social signals. Any recruiter will tell you it’s easy to spot the profile of someone looking for their next job – the gathering of LinkedIn Recommendations, the sudden uptick in the people they follow on Twitter. What’s harder is following a whole heap of people, then reaching out at just the right time to snag the best hires. Signal is designed to monitor a pool of people designated by the recruiter, then nudge the recruiter to get in touch… at just the right moment. The algorithm seems a little twitchy to me at present, but usage and experience will help the team improve on this quickly.

So why should HR be looking at this, instead of recruiters? Recruiters value tools in terms of cost of hire – what part the tool plays, how much it saves them, etc. I believe in this space, HR actually have more at stake. If you set signal to watch a pool of your top 200 EXISTING employees – and reach out to them before that crucial resignation letter happens….

Perhaps it’s more applicable to smaller firms (and potentially those tech startups whose employees are often so much more visible online). After all, lose one person from a workforce of ten thousand, it might not have much impact – 0.01% of your workforce? Losing one person from your workforce of fifty? 2%. Very different. How about the value of the loss? Lost productivity, training investment, time of hire and induction for replacement, general disruption and so on.

So here’s the sticky question: does this amount to spying on your existing employees? Signal uses public data to generate it’s indicators, available to both the internal HR team and to all external recruiters… but does a move like this break the bond of trust that must surely exist between employer and employee? Is there a way HR teams could use this in a responsible manner and not break that bond… or should employers accept this is too far, that it crosses the creepy line… and that recruiters will always have the upper hand?

Let me know your thoughts in the comments – I’ll be catching up with some of the folk behind the product next week in Helsinki at HRTechTank and TruHelsinki!

  • Laimonas

    Great article James! I would like to add that it’s also
    applicable to large enterprises. Current US employee turnover rate is 3.2% (http://data.bls.gov/timeseries/JTS00000000TSR)
    and just imagine how heavily it effects large enterprise with let’s say 10 000
    employees. We are speaking about 320 employees thinking about leaving the company. Turnover cost employers as much as 150% of departing employee’s
    salary. That figure can rise up to 250% for managerial positions. Knowing the turnover rates and costs, we can see how valuable preventing this from happening is.

« »

Scroll to top
HELLO WORLD