Just so everyone who reads this is clear on my views on “HR” as a term , pls read this https://www.linkedin.com/pulse/20140312204229-5106401-i-am-a-human-not-a-human-resource . For me – Nothing trumps “respect for individuals” when it comes to people .
I am a big fan of using analytics as a decision support mechanism in the corporate world . However , the importance of context is unusually high when it comes to using analytics to make decisions about people – hiring , career progression , firing or anything else of that nature . If we are not context sensitive , I can almost guarantee you that analytics about people will almost always lead to poor decisions .
People – unlike resources ( the R in HR) or Capital ( the C in HCM ) – are not fungible.
When we do analytics on cash , inventory etc every unit is the same and we can make good decisions on aggregated information . People are not the same – and at an aggregated level, the data is usually misleading .
Let’s say our Sales is going south and we need to find a list of sales reps to fire . The standard analytics is to find quota attainment of all reps , and take the bottom list and fire them . More experienced managers might give some allowance for forward looking pipeline to see if some of the people need to be given a second chance . If the decision to fire is taken at top management level – everyone in sales looks the same and finance department could just say “you need to let go of 5 telesales reps ” . The spreadsheet does not usually tell you all the different characteristics that make each Rep different ( like say only one of the reps is multilingual and that is super useful in real life) , whether there are mitigating circumstances ( 4 managers changed for telesales team in one year and each had a different strategy) , etc .
Unlike assets and resources , people care about other people
If you choose to discount your widgets heavily to do a fire sale , or throw away a few to scrap – the remaining widgets don’t get affected . It’s not the same with people.
If your friends and colleagues are poorly treated – especially for reasons you have to guess because no one told you – It will affect your performance on the job. Best case it decreases your performance on the job temporarily , and worst case you will walk out of the door and find another job . How you will react is also probably different from how any of your colleagues will react to the same decisions .
The new VP for sales might cost more than existing Sales leaders . That is market reality and you can’t fight it if you want to hire top talent . Analytics can even prove that the salary is par for course . What it might not tell you is the cost of replacing the existing top gun sales leader who is pissed off when she finds out the new person is making more . Unlike assets – people talk !
Analytics can’t be big brother
Aggregated data is the problem – but analytics can still figure out a lot about individuals by gathering fine grained data about each person, and in many cases without violating laws . This still doesn’t mean it is ethical to “snoop”. If you leave a social media exhaust , you are not really permitting your employer or anyone else to use it against you . World will be a terrible place when employers (and government ) keeps tabs on you all the time .
Finance and HR don’t look at “people” the same way
If you ask the CFO and Chief HR officer of a big company on how many people work there – the chance is high that you will get two different answers . And each will only trust their own number . There are good “technical” reasons for this discrepancy ( like how each function defines FTE ) – but this also causes a lot of bad decision making when HR and finance data is combined to make people decisions
I could go on with what else causes bad HR decisions , but let me wind up here with a parting thought . In near future , Robots will become part of what HR supports today . All this analytics that don’t work as intended when it comes to people – they will work much better when it comes to machines . So perhaps it was not all wasted effort after all 🙂
3 thoughts on “Analytics about people – nothing but land mines all around ”
My son is 18 years old and doing casual construction work. As you’d expect, it’s obvious to him that different work gangs have different productivity. What’s not so obvious is that the same people, when placed in different groups, have wildly different productivity – i.e. the combined productivity of 2 work gangs, made up of of the same 10 people, can be very different depending on the mix of people..
It is simple enough to say that this is the responsibility of the site foreman (or your industry’s equivalent), but there is a disconnect between what head office sees when allocating workers to a site, and the amount of control my son has over his productivity and the productivity of his gang,
As I’m sure you know, the growth of the casual / freelance work force comes from the idea that the transaction cost of hiring and firing has massively reduced over the years. Relying on the numbers to tell us what a group of people is worth assumes we are measuring ALL the differences between the people in that group, AND their relationships. For example, one of the missing “transaction costs” that has not been allowed for is the loss of knowledge about the capabilities of an individual “resource”; i.e. Analytics may tell us that these 10 people have the appropriate certifications to do the required work, and give us an idea of their work experience, but it doesn’t tell us about their productivity as individuals or as a group.
In short, analytics reduces people to a fungible resource, because analytics isn’t smart enough to know the difference.
Your son’s case is a powerful example – thanks for sharing .
If you get a chance – check out this book https://www.amazon.com/Extreme-Ownership-U-S-Navy-SEALs-ebook/dp/B00VE4Y0Z2 it shows the case of good managers and bad managers impacting their teams