These days, most major HRTech vendors offer some form of automated ‘flight risk analysis.’ Imagine what a competitor might do if they could change your basic retention policy by changing the data that drives your intelligent tools.

Machines are significantly easier to fool than people. The machine always assumes that the data is clean. It rarely has the sense or capacity to question data quality. It’s up to the user to figure out if the. Machines are predictable. People, less so. 

These days, most major HRTech vendors offer some form of automated ‘flight risk analysis.’ That is, there is a standard calculation based on some combination of internal and external data. It gives you an estimate of the likelihood that employee X will be leaving soon. Many tools include this flight risk analysis as a component of the individual employee profile. Anyone with the right permissions can see the machine’s estimate of some particular employee’s attrition likelihood.

Let’s imagine that you find out the Sara Smith has a 70% likelihood of leaving in the next 90 days. And, the company policy is that we make decisions about anyone whose score is 70% or higher. What do you do?

The first choice is nothing. Sara is a well-known pain in the butt and we’d be happy to see her go. We decide that we won’t do anything to try to retain her. The moment we make that decision, it becomes clear from body language that she is a short timer (organizations are really good at picking up on clues about people’s position in the status hierarchy). Her ability to get things done is compromised. It becomes a self-fulfilling prophecy.

The second choice is to do something. Retention bonuses, reassignments, new responsibilities, deep conversations, pleading, promotions, and good talking to are a few of the available options.

Of course, as soon as the organization starts rewarding people in order to keep them around, everyone will want in on the act.

With this in mind, we asked a group of talent oriented professionals on Facebook the following question: “Your employer is using a flight risk algorithm to evaluate the likelihood of your departure. What do you do differently? How will you try to game it?”

Here was the best response (thanks, Andrew Gadomski)

1. Post your house for sale. They use MLS as an indicator
2. Sign up for job board alerts, even if they are bogus, and do so by logging into monster, CareerBuilder etc using your LinkedIn profile
3. Add competitors and like businesses to who you follow on LI, and start liking those posts. Increased engagement overall is good
4. On LI, change your settings on how you can be contacted, make your email and phone public, make your changes public and able to be seen by others, and change your headline. Also, update your profile pic and background.
5. Do these activities within a few months of your recent joining or promotion anniversary. That’s another signal for flight (well documented actually)
6. Go on a tear by updating your connections, removing and adding (all social really)
7. Go into Twitter and start following competitors and like businesses like crazy but not their actual handles, their career handles
8. Follow job boards on Twitter
9. Follow about 100 recruiters to your Twitter where the handle says recruiter, staffing, headhunter etc

^^ that’s all to game it and have the flags go off so you get engaged prior to review time so you can sweat your supervisors for more cash on your merit increase. You can also do the same when you apply internally for a job and really really want it.

Basically, if you install a system like this, you must expect that employees, particularly those who are most likely to be flight risks, will game it.

There are many bits of interface problem and opportunity that come along with increasing the use of smart tech in HR applications. People are really good at learning to conform to and exploit new systems of measurement and evaluation. Expect to see Glassdoor style operations that offer insider guidance about how to game your systems.

This is the least nefarious data-hack that HR is going to encounter. Imagine what a very aggressive competitor might do if they could change your basic retention policy by changing the data that drives your intelligent tools. How would you tell? Do you have any security?