iSixSigma

How much Available Time is too much?

Six Sigma – iSixSigma Forums Old Forums General How much Available Time is too much?

Viewing 22 posts - 1 through 22 (of 22 total)
  • Author
    Posts
  • #28675

    Edwards
    Participant

    I am working on a call center project in a customer service area in the finacial industry. The goal of the project is to maintain a minium sigma level of  3  with a defect being any calls answered or abandoned over 90 seconds and the opportunities being all calls received.
    Currently we have a 1.5 sigma with an average speed of answer of 121 seconds +. The interesting part of the problem is when we look at the representative time spent in different states, I am seeing that the representative are available to take a call 20% of the total time they are working. This is very confusing. How could there be so much time waiting on a call, and still our ASA is so high? 
    Has anyone ran across simular problems?
    Does anyone have any benchmarking information on the precentage of time that a call center represenative should be available, talking on the phone, involved in after call work?
    Any help would be welcomed.
    David  

    0
    #71710

    Erik L
    Participant

    Just postulating here, but perhaps there is a mismatch between the spare time available to the influx of calls.  Have you looked at your resource loading at peak call times?  To your metric of ASA, are you sure that the mean is the best metric to look at here.  Do you have a normally distributed population of data?  Have you looked at the median response?  If the data is skewed, the mean is drawn to the heavier tail and may not be the best choice.  Have you looked at stratifications of the data?  Are there shifts of operators receiving calls? Multiple locations receiving calls? Years of experience?  etc. etc.
    Regards,
    Erik

    0
    #71712

    Kelly Maidman
    Participant

    What is the goal of lowering the ASA?
    Perhaps ASA is not the best measurement?
    How about measuring: 1st call resolution, 1st call Cycle time, or First Pass Yield on calls without escalations?
    (just my $0.02)
    -Kelly

    0
    #71716

    EPS
    Participant

    Check your measurement system as well! Maybe there is something wrong with it.
    Good luck!
     

    0
    #71718

    walden
    Participant

    Sorry David, long winded reply here.
    I’ve completed the same analysis on call centers also in the financial service industry and to get where we needed/wanted to be it required data compiled in half hour intervals.  This helped ensure we maintained the proper staffing and readiness throughout the day as call volume changed.  We found call volume peaks at between 3-4 (just before market close) and again from 10:30-11:30 (when all parts of the country were awake and at work).  The key is to match staffing with volume by period so at the end of the day the result you, the firm, and the customer expect are met.  The trick is to make your queue time/service levels imune and insensitive to call volume changes.
    With respect to expected talk time and after call work there is no standard across call centers.  Tech help desks may tend to see calls of 45 minutes while an account value queue may see calls of only 1-2 minutes.  Both talk time and after call work is related to the knowledge of the rep and the acessability of information the rep has.  Bottom line, bench mark your current process and Kaizen the process through process improvement projects. 
     

    0
    #71742

    Fernie
    Participant

    David,
    The key is in your call volume peaks, as well as your staffing levels. Actually, these two should be considered as some of the top X’s when you get to the analyze phase.
    It is very common to see Agents waiting for calls for specific intervals of time (low call volume periods), which builds up the AVAIL %, and see high call volume periods (which builds up the ASA and ABAN % ), during the course of any given day.
    If you can identify your volume peaks, you should be able to forecast your staffing needs, as well. Look at historical data, and let the numbers talk to you.
    Good luck,

    0
    #71743

    Fernie
    Participant

    I forgot to mention another potential source of information:
    If your ASA is so high, you may have a problem with your call allocator, or call routing system. Talk to your PBX /IVR administrator and see if they can run a trace for you.
    Good luck, again
     
     

    0
    #71782

    Bill Nelley
    Participant

    I used to work in a call center years ago…I’ll try to dig through my files to find some benchmarking information for you.
    Have you tried to segment or stratify your data? It might be interesting to see how it shapes up by state, country, representative, hour of the day, etc. and any other factor that you can think of.

    0
    #71788

    Cone
    Participant

    The problem is simple, answering times are not normally distributed — they follw a distribribution that is skewed right. A simple simulation will show you that your efficiencies (% of time operators are busy) will be down around 60 – 70% before you start getting the defect level you are seeking. The body of knowledge is not generally covered in Black Belt training — it is call queueing theory.

    0
    #71791

    Bobby
    Participant

    David,
    Worked 6 years in a call center.  I agree with the direction you are bing directed.  The one thing I would add is something that we came across.  A few call reps were manipulating the phone systems, even during peak periods, where their % avail would be high even though there were calls waiting in queue.
    The 30 min breakdown of call statistics is the best place to start.  Once you have done this, especially if you can break it down to the representative level, you can see if there is any system manipulation.  Once this has been eliminated, the half hour breakdown will allow you to see any issues with staffing levels, which is most likely the cause of your problems.
    Good luck.

    0
    #71800

    Ron Jarrett
    Member

    David, 20% may not be enough available time if you are looking at it over a long time period. What are your peaks and valleys? You need to look at the availability on an hourly interval, or maybe 1/2 hour intervals. Based on your situation, you probably have reasonably consistent call volumes, for a given hour on a given day of the week or month. You probably have 40% availability during some intervals and 0% during others. In my experience, you need to provide excess “overflow” positions (people who only receive calls during certain periods of the day or after a certain wait time). By doing this, we reduced the average wait time from 3 minutes + to less than 15 seconds. I hope this helps. Ron

    0
    #154881

    Laura Gibbons
    Participant

    Hi David,
    Are you saying that 80% is your occupancy rate (the inverse of a 20% availability)? If it is, that is fantastic…If you are not at 80% occupancy, what other off-phone activities are they doing that is bringing down their occupancy? What is your shrinkage factor: 1.05…1.20…?
    I just finished a call center project that increased our occupancy (thus decreasing availability to take calls/sitting around waiting), which was financially successful in deliverying results.
    Good luck – Call center projects are some of the hardest, in my opinion, but some of the most fulfilling for the black belt and the organization.
    ~Cheers,
     
    Laura

    0
    #154898

    Lebowski
    Participant

    Laura Gibbons,
    Congratulations on your project success. Rather thann make the guy feel badly because you are so successful and actually told him nothing about what you actually did, just maybe you have some suggestions?
    It must be difficult to have such a worldly view of which projects are difficult and which are so satisfying considering you only began training 14 to 15 months ago. How about

    0
    #154899

    Lebowski
    Participant

    Laura Gibbons,
    The end of that last post was a request for the way you measured the difficulty and satisfaction of those projects. Maybe it is just very satisfying to be in a call center and decide that your projects are the most difficult and satisfying. I am sure the Six Sigma world will be much better off with new breath of fresh air of objectivity blowing through it.
    Lebowski

    0
    #154901

    Mikel
    Member

    Call center projects are trivial. Just queueing and turnover.

    0
    #154906

    karthikeyan
    Participant

    Hi, I genarally take the support of data in regards to IVR usage, more the IVR is user friendly less the hit that will take on abdn calls. You should also try and eliminate the hold time by digitizing the part of the process that requires ACW. Add the the ACW to the performance rating that will automatically kill ACW. Hope this layman language would help. All the best.

    0
    #154911

    Lebowski
    Participant

    Karthikeyan,
    I don’t get the indigenous language but it looks like a better attempt at some helpful information than the cheerleader stuff from an earlier post.
    Love those TLA’s
    Lebowski

    0
    #154912

    Lebowski
    Participant

    Stan,
    Are we going to get the second installment in the trilogy of trilogies? Are we about to the part where the evil wizard disguises himself as a cowboy?
    Lebowski

    0
    #154940

    Laura Gibbons
    Participant

    Hi David,
    Are you saying that 80% is your occupancy rate (the inverse of a 20% availability)? If it is, that is fantastic…If you are not at 80% occupancy, what other off-phone activities are they doing that is bringing down their occupancy? What is your shrinkage factor: 1.05…1.20…?
    I just finished a call center project that increased our occupancy (thus decreasing availability to take calls/sitting around waiting), which was financially successful in deliverying results.
    Good luck – Call center projects are some of the hardest, in my opinion, but some of the most fulfilling for the black belt and the organization.
    ~Cheers,
     
    Laura

    0
    #154957

    Lebowski
    Participant

    Laura Gibbons,
    This was not good enough to rationalize posting it twice.
    Lebowski

    0
    #154967

    Mikel
    Member

    It wasn’t even good enough for once.

    0
    #154969

    Vidyadhar
    Member

    First of all, I would suggest that you look at historical data on arrival patterns, look at what the average time that a person is willing to hold before it is abandonned. from that you could determine your ASA target, next put in a traget for percentage abandoned calls within 3/4 ASA. Also the average call duration, arrival patterns, staffing & scheduling are some of the factors that could affect occupancy. We had a similar project, where the major impact for occupancy was because of the staffing plan for a shift. However, it would be advisable to validate all the factors, even attrition & absenteesim could affect your occupancies.
    The idea should be to optimise with minimum variation, Try to keep your occupancy between 80-85. if it is less than 80% indicates underutilisation of resources & >85% would probably result in a risk to your service levels.

    0
Viewing 22 posts - 1 through 22 (of 22 total)

The forum ‘General’ is closed to new topics and replies.