Submitted by ggggeo on
BLUF: My question is how can I state performance objectives in a way that satisfies my boss while still helping my team to focus on the right behaviors? I'm really stuck here, so thanks in advance for any advice.
Background: I am in the midst of helping my team develop/update annual performance objectives. They are pretty much finalized, but I am struggling with the "right" (or even "good") metric for one of the roles that reports to me.
The role has both programming and some market research design/recommendation responsibilities. My organization has set a number of billable hours for three levels of performance (like Hi/Med/Low) and my boss would like to see some additional productivity metrics and has suggested hours per project.
I struggle with this because we have no control over the number of projects that come in, we just program every survey the company throws at us. We also have no control over the length of the survey, how much is "recycled" from a previous survey, how many changes occur after the "final" survey is handed off to us (and we never say "Sorry, you already gave me the FINAL, no changes!") I have tried thinking of and scouring the net for reasonable proxies, but everything I found incentivize the wrong behavior, e.g., number of lines of code, which is easy to quantify but doesn't lead to good or efficient code. In addition, they also have a responsibility to add value to surveys they program, such as make recommendations to improve data collection and or survey respondent experience, and I am concerned that that will be the first thing they cut out if they have to reduce the time spent on surveys.
All that said, I am not concerned with their level of productivity. Two people are currently in this role, and they are able to handle all the work that comes to them, and can easily bear the expected increase in workload as our company adds 10% or more surveys this year. I have tried to focus the objectives around on-time delivery of programmed surveys, but my boss isn't 100% sold.
Do you have any opportunity to help guide clients in the creation of the survey? Can your team help clients figure out what they want to measure so you asked for fewer changes which will reduce the total time per survey? How are you guys at accurately estimating the time required and delivering to deadline? Can you guide clients towards proven questions that also happen to be in your library? Hope that helps you get unstuck. - Sam
Thanks for the reply! That's a great idea about the question library. One of my directs had a similar idea and I've had him working on that.
We don't have much input in the way of survey design, when it is delivered to us it is in a specific Word template that theoretically has everything, but as the person programs it we often find that certain scenarios are missing from question skipping logic, or that the way the survey writer has it in Word isn't optimal and we will make that suggestion before we program it.
The bulk of changes come after the survey has been programmed and the survey designer is testing it and seeing it on the web. They find things they missed the first time (wording, missing options, etc.) or often that the survey is programmed the way they wrote it, but they realize they didn't write it correctly. We try to pre-identify those as much as we can.
As far as deadlines, the company has a standard 4-5 day turnaround time for survey programming and testing, and we pretty much never need or take more than that time. Issues sometimes arise when a survey is handed to us several days late while still expecting the same original turnaround date, but that is a separate issue (I think!)
So we really don't set the timeline, we have little control over required post programming changes, and we consistently meet deadlines. I'm still a little stuck as to how best to demonstrate and build metrics around their productivity.