Slashdot Mirror
How Adobe Got Rid of Traditional Stack-Ranking Performance Reviews
Hugh Pickens DOT Com writes "Bob Sutton reports that in 2012 Adobe moved from yearly performance rankings to frequent "check-ins" where managers provide employees targeted coaching and advice. There is no prescribed format or frequency for these conversations, and managers don't complete any forms or use any technologies to guide or document what happens during such conversations. They are simply expected to have regular check-ins to convey what is expected of employees, give and get feedback, and help employees with their growth and development plans. 'The aim is to give people information when they need it rather than months after teachable moments have passed,' writes Sutton. Donna Morris, Adobe's senior vice president for People and Places, says her team calculated that annual reviews required 80,000 hours of time from the 2000 managers at Adobe each year, the equivalent of 40 full-time employees. After all that effort, internal surveys revealed that employees felt less inspired and motivated afterwards—and turnover increased. According to Sutton, Adobe's bold move seems to be working. Surveys indicates that most Adobe managers and employees find the new system to be less cumbersome and more effective than the old stack-ranking system where managers must divide employees into groups — for example, maybe 15 percent of people can be assigned the highest rating. 'That goes against our core value of being genuine,' says Ellie Gates, director of management effectiveness at Adobe. 'Our goal should be to inspire people to do their best work.' Since the new system was implemented, voluntary attrition has decreased substantially, while involuntary departures have increased by 50% because the new system requires executives and managers to have regular 'tough discussions' with employees who are struggling with performance issues—rather than putting them off until the next performance review cycle comes around. 'It is reducing unnecessary cognitive load, while at the same time, nudging managers to engage more often and more candidly with direct reports to help them develop their skills and plan their careers,' concludes Sutton. 'It also bolsters accountability because managers have far more responsibility for setting employee compensation than under the old system.'"
Stack Ranking only works on a short term basis where you want to trim the fat.
If you do it for too long, two things happen (a) you start cutting into good performers (b) people will not collaborate to make others look good
Stack ranking is for managers who are not grown-ups.
Train your people. Teach them to improve. Defend them while they learn. Make them better and they will make the company better.
If you are in charge and something goes wrong, it's your responsibility. It's your fault. And if it isn't your fault, it's still your fault because you're the manager.
You are responsible. 100% of the time. No exceptions.
Take responsibility for your job and do your job. Train your people. Take care of your people. Grow up.
In Japan, the Manager is responsible if an employee screws up. If an individual does something merit worthy, the entire team takes credit. As well, the freshman hires have no say in their job assignment - higher ups work all that out and take any heat...they sort it later as skills and relationships mature.
Stack ranking, meet regression to the mean.
Yes, p accomplishes 1-p (for some p, always) but it's not necessarily the same people in the p pie slice year over year. The Pareto does not state that 20% of the people with account for 80% of the output and will continue to do so, because as we all know 100% of what drives performance is whether you have it, or you don't, end of story.
Sapolsky on Heights And Lengths And Areas Of Rectangles:
What does stack-ranking achieve as a long-term evolutionary pressure? It helps the company accumulate the people who are best at concealing their dips, no matter how the chill winds blow.
Just what you want cultivate, a whole cadre of engineers specializing in meteorology.
There was a different passage about genetics I was trying to find. A population will only retain multiple genetic phenotypes if each of those phenotypes is advantageous in some circumstance or environment. Any phenotype that dominates across the board, in nearly every circumstance, soon extinguishes the competition.
That we have so many phenotypes indicates that human circumstance is extremely fluid.
" 'The aim is to give people information when they need it rather than months after teachable moments have passed,' "
This is a very important thing, and it's kind of sad that it IS something that people in business sometimes realize, rather than being something we could assume everyone capable of forming complete sentences understands.
And that's why it is really important that Timothy, who thinks we form a passive 'audience' here, who somehow imagines that occasionally pushing the button to publish a user submission under his name, without even fixing the obvious typos first, qualifies him as the creator of the site, really needs to feel some backlash today. Not in 6 months when the whole site goes, today.
It's also why each and every member of the staff that encouraged the delusion that this 'beta' was a reasonable, workable idea needs to be gone now. Not in 6 months when the whole site goes, today.
Because those teachable moments are short, and these are not minor little mistakes anyone could make. These are possibly the biggest mistakes anyone in their position could make, in regards to their work. They are mistakes that you would expect from someone who was just recruited from a business school last week and had spent no more than 20 minutes lurking before deciding to change everything.
=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Friends don't let friends enable ecmascript.
Nobody buys Playboy for the articles. They do it for the hot, nude women (sadly, sans grits). It just so happens that /. is exactly the same. No one reads /. for the articles. The articles were news two days ago. And no one reads /. for the summaries. The summaries are almost always wrong.
Everyone reads /. for the comments.
Well no, I for one read /. for the hot, nude women.
Wir sind geboren, um frei zu sein - Rio Reiser
Since one of the things that drives me the MOST crazy about my current job, and MOST makes me think about quitting, is in fact the near-constant requests for writing various self-assessments, goal documents, and other such things that are not actually related to my job, and which don't actually seem to be used for anything other than making the people responsible for requiring all of those documents look like they're doing useful things. Drives me crazy. After all that, our performance reviews tend to basically say "yep, you're doing fine. Have a raise that is exactly in line with inflation, just like everyone else is getting."
Calling us the Audience is like the Bee Keeper calling the Bees the audience.
Bees make honey. You can set up bee boxes and have bees live in the boxes and make honey that you can harvest. But the bees are free to leave at any time. The only reason the bees stay is because the boxes are less trouble than building a beehive. Try making the bee box unusable and the bees will just go build a beehive elsewhere. Don't believe it? They've been building beehives for a lot longer (*cough* Usenet *cough*) than bee boxes (*cough* Slashdot *cough*) have been around.
I'll see your senator, and I'll raise you two judges.
That's absolutely where I am with it also. I've had no fewer than 5 calls from Adobe before and after the CC business and I've told them each time I'm never buying another one of their products until they sell perpetual licenses. They insist they won't buckle on the issue, but I believe they will have to eventually.
It would be harder to resist CC if they put in a single compelling feature in a product since CS6, but that simply hasn't been the case.