improving answer rate without increasing FTE

Assuming 150 calls per FTE and 7.5 working hours, the average handle time should be close to 3 minutes if the FTE does nothing but answer the phone. Here are some reality checks: What is their idle time as a percentage of total hours worked? What is their aux time (other than breaks)? What is their after call work? If idle time, aux time and or after call times are low, you know they’re in full production mode. If idle time is high, then you should be looking at your staffing model because it doesn’t predict incoming call volume very well (what are you staffing towards: ASA?). If aux time is high, then you have to look at other work that they perform? Why and what? If after call work is high, you’ll need to take a look at what they are documenting and why they spend much time there. The kaizen is definitely a good idea, but the diagnostics could give an immediate diagnostics of the situation.

Step 1: Create histograms of time to call abandons vs. pick up of call to see if there is a threhold at which clients start to hang up.
Step 2: Compare total time availalbe (for FTEs) and break it out into “on call time” (i.e. handle time), “aux time’, “idle time”, “after call work” and create percentage of each of the times relative to “time available to make call”, or “handle time”
Step 3: Review pattern of incoming call volume and compare with # of FTEs available: Example 8 – 9 AM 400 calls recied 30 calls abandoned 150 FTES. Do this for a 24 hour time period and five days and create a p-chart 30/400 = 0.075 abandonded (This can be done in a simple excel spreadsheet or using Minitab.
Other questions to ask: What is your attrition rate of employees? How many systems do they need to toggle to help the customers? Are there scripts or is there a call process flow that the reps can follow? Do you have excessive downtimes? How long does it take your reps to bring up a field? What is the complexitiy of the questions? Do they make both outbound and inbound calls? Do you have an opportunity to create a two-layered experience system with one internal transfer and reorganization of inbound verus outbound calls … I don’t know your call center. But in the past this type of analyis yielded quite some interesting results. Also, sit with a few reps and double-jack to see for yourself, what they are dealing with. This in itself can be an eyeopening exercise. Overall, this needs to result in a project. But I doubt that you have identified the problem and are rather scratching the surface of the sysmptons. It would be good to do this high level analysis to guide in a problem direction. Others will say go to the shop floor and look for yourself, that’s an alternative too. Good luck!

Great detailing from Answer rate
Parallely figure out these quick ones
– most of the calls belong to which category – query, complaint or requests – this shall be industry specific, am sure such categorization is available for your industry as well. Can few of them be routed to self care options, which can be IVR, SMS based, Web Based solutions to cut down the calls. At the end of it pareto principle holds valid, pick the top ones and see any one can be routed giving ease, faster resolution to customer
– convergence as highighted by answer rate is critical to look into. How many screens an agent refers too, which one most often, in case multiple can we converge them into one.
– drilled down agent analysis as well comes handy. Higher tenure agents take less time to resolve (hypothesis), rostering plans etc can accordingly be changed for high volume hours.
– BASIC data should be right, the volume that you are currently seeing , is it actual, or it includes repeat attempts by customer as well. MSA is very critical, if possible get data of calls at each point before it reached the agent. Answer rate very rightly pointed to systems, hardware, software, medium, etc which effects call transfer etc.
Hope this helps!
 

Well all the things have been answered by ppl more experience than me. But O), here is my two bit from my recent experience at my new job at a call center:
1. Agents have to discuss about a policy with customers in their overall call if the customer is out of warranty period: each of them is doing this in his/her own manner – leading to inconsistency in delivery, all information not provided correctly etc…you can guess the rest.
2. If the customer is out of warranty and accepts the charge for the out of warranty service then he/she has to provide debit/credit call details before we can proceed.
Solution that I have:
1. Automate the script for the out of warranty policy and let customer know that he/she is going auto for 2 mins
2. Automate the acceptance of debit/credit call details in IVR if the cust. accepts the pt 1.
Gain: Standard script, defined time, debit/credit card details security
Suggestion for you: Listen to the calls, identify what you can automate – implement ASAP

Hi all,
I would also suggest you review the pooling theory. The more agents are in a pool to answer a large number of calls the better the answer rate will be. This comes with an assumption that agents are cross-skilled or can be cross-skilled quickly.
The alternative option is understanding the peak call times and using agents which dont necessarily have the call knowledge to take messages. This means that the client will have their call answered (thus lower abandons) and they can be called back when the peak period is over.
Regards,
Ross.