Measuring Results and Behaviors

0:08 recall that in course one 0:10 we talked about setting smart goals and 0:12 objectives 0:14 these goals and objectives have to be 0:16 linked to the strategy and vision of the 0:17 organization that's an 0:18 absolute first key step it's very 0:20 important 0:22 then we need to measure people 0:24 performance and behaviors against those 0:26 objectives 0:27 now some of these objectives will lend 0:28 themselves to quantitative measurement 0:31 others will not for example it might be 0:33 easier for me to quantitatively evaluate 0:35 your sales numbers or your cost cutting 0:37 numbers 0:38 but in addition to hitting those numbers 0:39 companies also care about 0:41 you exhibiting certain behaviors such as 0:45 you hit those numbers in a way that 0:48 supports 0:49 and reinforces corporate culture that 0:51 you grow yourself as a leader and you 0:52 develop others along the way 0:55 and ultimately these dimensions of 0:58 performance 0:59 competencies and behaviors feed into 1:01 your performance appraisal form 1:03 this is a segment of a typical 1:05 performance appraisal form 1:06 in most companies managers would fill 1:08 them out once a year at the end of the 1:10 year 1:10 many companies are now transitioning 1:12 toward feedback systems 1:14 where such information would be 1:15 communicated to employees more 1:16 frequently than once a year 1:18 accenture microsoft amazon netflix are 1:21 good examples 1:23 but let's go back to this performance 1:24 appraisal form for a minute 1:26 you can see that this employee is being 1:28 evaluated in five different dimensions 1:30 project management customer service 1:32 production support teamwork and 1:34 communication skills 1:36 so the first step is to assign a 1:38 quantitative score to the employee 1:40 on each of these dimensions by comparing 1:42 his or her performance 1:44 and behaviors relative to standards and 1:46 objectives 1:47 so for example for assigning a score for 1:49 project management your manager might 1:51 consider 1:52 goals of completing projects on time and 1:55 on budget 1:56 for customer service your manager might 1:58 consider scores from customer 1:59 satisfaction surveys 2:02 lexus an automotive manufacturer ways 2:05 customer satisfaction scores very 2:07 heavily in its appraisals of dealers 2:10 for teamwork many companies would go out 2:13 to 2:13 your peers to get that evaluative 2:15 feedback 2:16 some companies would rely on the input 2:19 from your primary manager 2:22 in addition to assigning scores on each 2:24 of these dimensions 2:25 your manager would also need to provide 2:27 some vignettes or examples of behaviors 2:29 and outcomes to explain those ratings 2:32 and even if your company does 2:33 performance appraisals once a year 2:36 don't wait until the end of the year to 2:38 come up with a list of those behaviors 2:39 and outcomes 2:40 in the beginning of the year start a 2:41 text file on all your direct reports 2:44 your teammates you need to evaluate 2:45 and write down their key behaviors and 2:47 outcomes following the completion of 2:49 major tasks and projects 2:52 in addition to getting scores on 2:55 individual dimensions here 2:57 an employee would also receive an 2:59 overall performance score 3:02 now in some companies it would be either 3:03 a simple average or weighted average 3:05 of the scores on different dimensions if 3:08 it's a weighted average it would 3:09 typically be weighted by 3:10 the importance of different objectives 3:12 or different dimensions 3:15 in other companies you would rely a bit 3:18 more on managerial discretion 3:19 meaning that the manager would come up 3:21 with a holistic score that they believe 3:24 best describes the overall performance 3:26 of course in light of the 3:28 ratings on the individual dimensions 3:31 this overall score is really important 3:33 it determines your likelihood of 3:35 promotions 3:36 your salary raises your compensation 3:38 your bonuses and so on 3:40 a quick side note on the scales most 3:43 companies use scales that vary from 3:45 three point to seven point scales to 3:48 measure behaviors and outcomes 3:50 it's fair to say that on most behaviors 3:53 and outcomes a five point scale 3:55 would be sufficient to capture 3:56 variations in behaviors and outcomes 3:59 it is less complex and more easily 4:01 interpretable than a seven point scale 4:03 there's also research showing that five 4:05 point scales can be more motivating than 4:07 three point scales 4:09 because employees are more likely to 4:10 believe that they can move up on a five 4:13 scale than on a three-point scale 4:17 so at this point i want us to pause and 4:19 recognize that there are two 4:21 fundamentally different approaches to 4:23 assigning ratings to your employees 4:26 one is what we call an absolute rating 4:28 system 4:29 what i mean by this is that when i 4:31 evaluate a given employee 4:33 i don't compare that employee to other 4:36 employees in the organization 4:37 i evaluate them independently 4:40 and the second approach is a comparative 4:42 rating system 4:44 where i evaluate every employee by 4:47 comparing him or her 4:48 to their peers 4:52 the most typical form of a comparative 4:56 rating system 4:57 is a forced distribution of forced 4:59 performance curve 5:01 recall that in school or college he had 5:04 some classes that were graded on a curve 5:06 where say no more than 15 percent of 5:09 students could get an excellent score 5:12 a forced distribution approach 5:15 is very much similar where managers 5:17 assign ratings to employees 5:20 in accordance with pre-specified 5:22 performance distribution percentages 5:25 so the classic example here is g's force 5:28 distribution 5:29 which was used in the 80s and the 90s 5:31 when jacquel trent the company 5:34 this curve required that no more than 20 5:36 percent of managers are assigned 5:38 to exceeding expectations the top 5:40 bracket 5:41 70 would be assigned to the middle 5:43 bracket meeting expectations 5:45 and 10 had to be assigned to the lowest 5:47 bracket not meeting expectations 5:50 those would be employees that would be 5:53 the first to leave the company if their 5:54 performance does not improve 5:56 the top 20 would be in line for the 5:58 biggest salary raises the biggest 6:00 bonuses 6:01 jack welch referred to it as the 6:03 vitality curve 6:05 and he called it the kindest form of 6:08 management 6:09 which is a very interesting statement 6:10 when you think about this 6:12 now jack welch's logic was that very 6:15 often 6:16 we're fired from companies for 6:17 inexplicable reasons and this approach 6:21 gives you a lot of useful information on 6:23 your performance 6:24 it gives you an opportunity to change 6:26 your behavior and you get a head start 6:28 on your job search 6:29 if you don't improve so you called it 6:32 the kindest form of management because 6:33 you get a lot of useful information in 6:35 your performance you know exactly where 6:37 you stand in your organization 6:39 you know your prospects with the company 6:41 and you get a lot of that information 6:43 well in advance of major decisions with 6:45 respect to your career 6:47 know this approach to evaluating 6:49 performance force distributions was 6:51 really popular in the 80s 90s and early 6:53 2000s when lots of companies rushed to 6:56 mlag's approach 6:58 the enthusiasm for its distributions has 7:00 subsided significantly in recent years 7:02 and for good reasons as we'll talk about 7:03 shortly 7:05 but there's still a lot of companies 7:07 that are avid advocates of forced 7:09 distributions 7:10 lendingtree is a good example it's the 7:12 leading online 7:13 mortgage provider their curve is 7:16 actually a little harsher at the top 7:18 where no more than 15 percent of 7:20 employees can be assigned the top rating 7:22 which is a one 7:24 75 percent should be assigned a rating 7:26 of two 7:27 meeting expectations and bottom ten 7:29 percent rating of three 7:30 not meeting expectations ones receive 7:34 the largest bonuses 7:35 threes are the first to be fired if 7:37 their performance does not improve 7:39 and so in performance appraisals at 7:41 landing tree the key question you're 7:42 wondering about going into your 7:44 performance appraisal 7:45 is either one are you two or three 7:49 glaxosmithkline a leading pharmaceutical 7:51 company goldman sachs a premier 7:53 investment bank 7:54 all of them use variants of force 7:56 distributions this is another example of 7:59 a force distribution used by aig 8:01 a leading insurance company top ten 8:03 percent at aig 8:05 get a rating of one the top rating the 8:07 next 20 percent are twos 8:09 next 50 percent of threes and the bottom 8:12 twenty percent of fours and fives 8:14 as you might have guessed the ones get 8:16 the largest bonuses 8:18 the opportunities for promotion largest 8:20 salary increases 8:22 bottom twenty percent might not gain any 8:24 bonuses at all and the fives are the 8:25 first to be let go if their performance 8:27 does not improve 8:30 another variant of a comparative rating 8:32 system 8:33 is a forced peer ranking where a manager 8:37 rank orders these are teammates 8:40 subordinates direct reports 8:42 from the best to the worst 8:47 some consulting companies still use that 8:50 system where at the end of each project 8:52 manager would rank quarter all the 8:53 consultants from best to worst 8:56 and your end of the year score would be 8:58 essentially a weighted average of these 8:59 scores where the weights are the times 9:01 you spend on different projects 9:05 a variant of a four-speed ranking or 9:07 even a way to get to this ranking 9:10 is a system of pairwise comparisons 9:13 where instead of assigning 9:14 one overall ranking right away i conduct 9:17 a system of pairwise comparisons where i 9:19 compare 9:20 each teammate's performance to the 9:22 performance of every other teammate 9:24 so here i would take france's 9:26 performance and compare his performance 9:28 to janice 9:29 isis michaels lucianis and schindlins 9:32 then i take jenna's performance i had 9:34 already compared her performance 9:36 to francis so i would compare jenna's 9:37 performance to isis 9:39 michaels lucianis and chinlins in this 9:42 particular case 9:43 you can see that as a manager i would 9:45 have to make 15 pairwise comparisons 9:48 the formula is n multiplied by n minus 1 9:51 divided by 2 9:52 where n is the number of teammates so in 9:54 this particular case it's 6 multiplied 9:56 by 9:56 5 divided by 2. and then i simply count 10:00 the number of wins 10:01 for each employee the person with the 10:03 largest number of wins 10:04 is propelled to the top of the ranking 10:07 the person with the fewest number of 10:08 wins is at the bottom 10:09 and you have a nice tie breaker which is 10:11 head-to-head comparison 10:13 in case you have a tie