Help with KPI's
Discussion
Guys and Gals,
I'm trying to sort out some KPI's for our helpdesk..
We run a slightly unconventional two tier second line desk, with "unskilled" call loggers passing the calls to the tech team for technical vetting (and hopefully fixing!)
So, our technical team deal with purely outbound calls...
What we are attempting to do is come up with a numeric value that displays the "effectiveness" of an agent...
Does anyone have any suggestions for statistics to use, formulae to apply, or websites to visit for inspiration?
TIA,
slinky
I'm trying to sort out some KPI's for our helpdesk..
We run a slightly unconventional two tier second line desk, with "unskilled" call loggers passing the calls to the tech team for technical vetting (and hopefully fixing!)
So, our technical team deal with purely outbound calls...
What we are attempting to do is come up with a numeric value that displays the "effectiveness" of an agent...
Does anyone have any suggestions for statistics to use, formulae to apply, or websites to visit for inspiration?
TIA,
slinky
Interesting. Clearly KPI's for the call takers will be difficult as their sole purpose is to recieve calls.
However, for the tech's a KPI can be established but what you do need to have an idea of is their target of calls closed first time without requiring an engineer visit; say >80% Green, >=70% amber and <70% red.
KPI's can only be used when the management measuring them has a clear idea of their own expectation, I'm not sure there is a standard?
However, for the tech's a KPI can be established but what you do need to have an idea of is their target of calls closed first time without requiring an engineer visit; say >80% Green, >=70% amber and <70% red.
KPI's can only be used when the management measuring them has a clear idea of their own expectation, I'm not sure there is a standard?
We're not really concerned about stats on the call takers.. they do their job... (kinda..)
What were looking at is attempting to work out how efficiently the tech's time is being used, and the attempt to put some form of value on this.. whether it be a percentage or just purely a number...
At present, we monitor clearance percentage, so could look at an average over the month (which they are targetted on) but we need to look at the bigger picture and assess the correct usage of the resource..
We can audit the amount of time spent on outgoing calls to customers, number of calls made, calls cleared, calls passed to higher levels of support/field service teams etc etc etc...
In all honesty, I don't think KPI is the best TLA for this, what we really need is just a PI, or PQ (Performance Indicator or Quotient) to give an at-a-glance numeric value of how well an agent is performing...
Whether this is do-able or not, however, is another question!
slinky
What were looking at is attempting to work out how efficiently the tech's time is being used, and the attempt to put some form of value on this.. whether it be a percentage or just purely a number...
At present, we monitor clearance percentage, so could look at an average over the month (which they are targetted on) but we need to look at the bigger picture and assess the correct usage of the resource..
We can audit the amount of time spent on outgoing calls to customers, number of calls made, calls cleared, calls passed to higher levels of support/field service teams etc etc etc...
In all honesty, I don't think KPI is the best TLA for this, what we really need is just a PI, or PQ (Performance Indicator or Quotient) to give an at-a-glance numeric value of how well an agent is performing...
Whether this is do-able or not, however, is another question!
slinky
You need something to measure against, and that something needs to be something which is important to the business, in this case perhaps the speed of fix and quality of fix.
The last helpdesk I was involved with placed a severity and priority indicator on each call (1-10 1 being highest) and was given 2 times - an initial response time (important to keep customer informed) and a fix time (this could be extended with the customers agreement) based on the priority and severity rating.
The relevant team (networking, coms, development etc) were passed the call and were measured on these two times and measured against those times deemed acceptable by the business (90% called back by initial reposne time 90% fixed in agreed time etc)
The customer was then called back once the job was completed and was asked to confirm that a. the job was done and b. asked to give a rating on the overall service 1-5 5 being very satisfied. The call was then closed and total open time recorded.
This was quite a big business and the system was managed by a fairly complex system but I think the basics work quite well.
Cheers
Paul
The last helpdesk I was involved with placed a severity and priority indicator on each call (1-10 1 being highest) and was given 2 times - an initial response time (important to keep customer informed) and a fix time (this could be extended with the customers agreement) based on the priority and severity rating.
The relevant team (networking, coms, development etc) were passed the call and were measured on these two times and measured against those times deemed acceptable by the business (90% called back by initial reposne time 90% fixed in agreed time etc)
The customer was then called back once the job was completed and was asked to confirm that a. the job was done and b. asked to give a rating on the overall service 1-5 5 being very satisfied. The call was then closed and total open time recorded.
This was quite a big business and the system was managed by a fairly complex system but I think the basics work quite well.
Cheers
Paul
Don't forget that if you set a target then the support techs will meet it - although the quality of the support may go down with it.
All we see (at a large IT company called 'Hat Pisces') is that when we set KPIs and other performance targets, the targets are met and the number of calls go up as the tech just 'splits' calls (where possible) into sub-calls, getting the customer to call again to do another part of the problem.
All we see (at a large IT company called 'Hat Pisces') is that when we set KPIs and other performance targets, the targets are met and the number of calls go up as the tech just 'splits' calls (where possible) into sub-calls, getting the customer to call again to do another part of the problem.
m-five said:
Don't forget that if you set a target then the support techs will meet it - although the quality of the support may go down with it.
All we see (at a large IT company called 'Hat Pisces') is that when we set KPIs and other performance targets, the targets are met and the number of calls go up as the tech just 'splits' calls (where possible) into sub-calls, getting the customer to call again to do another part of the problem.
Can they really be blamed for splitting calls if they don't think they can meet the customers and managements requirements at the same time.
This is the perfect time to get your staff involved with specifying what they want, as well as what your customers and yourselves want.
I know I'd be against being measured so closely especially if I didn't have any input into what was to be measured. If I couldn't provide the service I felt the customer needed it would have a demotivating effect on me as well.
Whats the perceived problems your having?
thepeoplespal said:
Don't forget that if you set a target then the support techs will meet it - although the quality of the support may go down with it.
All we see (at a large IT company called 'Hat Pisces') is that when we set KPIs and other performance targets, the targets are met and the number of calls go up as the tech just 'splits' calls (where possible) into sub-calls, getting the customer to call again to do another part of the problem.
Can they really be blamed for splitting calls if they don't think they can meet the customers and managements requirements at the same time.
This is the perfect time to get your staff involved with specifying what they want, as well as what your customers and yourselves want.
I know I'd be against being measured so closely especially if I didn't have any input into what was to be measured. If I couldn't provide the service I felt the customer needed it would have a demotivating effect on me as well.
Whats the perceived problems your having?
Of course they can't be blamed for it - it's just that you have to consider what will happen if you set bonuses based on targets.
Imagine telling your builder that he will get a bonus if he finishes building your house early. Of course he will finish early and get his bonus, but you probably wouldn't feel very smug when it starts falling down!
If there are currently no issues with the amount/quality of support and you just want a measure then fine.
However if you are using the measure as a tool to then go out and specify maximum remedy times, and minimum support levels so that you can 'optimise' your workforce, then you'll have to take care!
We are aware that there are deficiencies within our area and wish to backup the analysis of these issues with statistics... Not to be used in a big stick fashion, but to assist in the recognition of problems...
Targets are already in place, so this won't be used to change targets..
slinky
Targets are already in place, so this won't be used to change targets..
slinky
Simple performance measures of individuals only work when the tasks being performed are uniform. Helpdesks don't ususlaly work that way. They normally tackle problems with substantially varying degrees of complexity, and there is also tremendous variability in the quality of the solution they produce.
How do you measure the complexity of the problems being tackled, and how do you define and measure the quality of the solution being produced? I would suggest that both are so subjective and so open to non-random bias that the team couldn't make these assessments for themselves. And I imagine it would make not be economical to have a third party make that assessment for them.
Surely this type of assesment is the job of the team leaders on the help desk?
How do you measure the complexity of the problems being tackled, and how do you define and measure the quality of the solution being produced? I would suggest that both are so subjective and so open to non-random bias that the team couldn't make these assessments for themselves. And I imagine it would make not be economical to have a third party make that assessment for them.
Surely this type of assesment is the job of the team leaders on the help desk?
Gassing Station | Business | Top of Page | What's New | My Stuff