Monday, November 7, 2011

Call Centers: A Deep and Still Largely Untapped Vein of Operational Profits

I know what you are thinking. “How can he say we haven’t tapped profits from our call centers? We have improved our Website, improved our IVR, and tapped our user community which, taken together, has drastically reduced the number of customer calls we get. So now we don’t need as many agents and our offshore outsourcing has cut our labor rate on the agents we do need. Really, how much value is left?”

The sentiment above summarizes the strategy and results for many companies. It is a good strategy and it has extracted significant savings. However, though it sounds clichéd, it really is the tip of the iceberg…the visible, easy to get at, sources of cost. The lion’s share of the value…revenue and cost…is there for the taking for anyone willing to take a systematic approach. To wet your whistle, I estimate further cost reductions of 40% and call center revenue increases of 3X are easily achievable.

And don’t think that because you outsource and don’t have call centers that this doesn’t apply to you. The outsourcer is charging a price per minute or a price per call. Into that price is built the same inefficiencies you had before you outsourced, in some cases, even more.

Sources of Value

Part of taking a systematic approach is to analyze all the sources of revenue and cost that pass through contact centers and/or that are affected by contact center activity. Once we thoroughly understand the sources of revenue and cost and the drivers of those sources, we can make sure we have plans in place to systematically attack them.

There are four value buckets: three are cost buckets and an “other” bucket that has both revenue and specific kinds of cost not covered by the other three. The three cost buckets are Direct Labor (DL), Indirect Labor (IDL) and Capital Expense (CapEx).

Of these, Direct Labor is obviously and by far the biggest driver of cost. Moreover, the amount of Direct Labor to some extent drives the amount of Indirect Labor (Trainers, Monitors, Coaches, SMEs) and directly drives Capital Expense (Phones, computers, and software licenses for the agents). Drive down Direct Labor and you can proportionately reduce IDL and CapEx.

Before looking at the specific drivers of the amount of Direct Labor, let’s overview what is in the “Other” value bucket. First, poor process capability (agents not handling a call the way they are supposed to every time) results in a significant amount of cross-sell, up-sell revenue being left on the table. Despite the monitoring, the coaching, the incentive plans, etc, agents routinely fail to cross-sell. We have seen situations, where the agents only cross-sell 25% of the time they were supposed to. In situations like these, without improving the sales process one iota, just getting the offer rate near 100% will deliver significant top line benefits.

Another form of revenue left on the table comes from debt collection processes…not reaching the right party, not getting to settlement talks, not agreeing on a settlement number. Improving the collections process instantly puts more money in the hands of the company. (And again, please don’t think “I don’t have to worry about that because we outsource collections,” because your debt collector is leaving your money on the table too.)

In addition to revenue, there are costs in this “Other” bucket. One form is costs accruing in another part of the business due to mistakes being made in the call center. There are numerous small examples, like items returned because an address was entered incorrectly, but a huge source of unnecessary cost comes in the form of warranty processing errors. The agents don’t check the warranty and authorize returns when the warranty has expired, or they have returns sent to the wrong location, or they fail to properly instruct the customer on what to do which leads to other departments having to respond or backstop the call center. These areas are often referred to as “hidden factories” and it is obviously better to stop burning the toast than it is to improve your ability to scrape it.

Third, many companies, especially outsourcers, pay fines either to the government or to their clients for failure to follow the proper process on the call. For example, disclosures not being read to customers can lead to fines (e.g., mini-Miranda rights that collections agents are required to read to debtors.) Further, failure to properly recap calls can lead to outsources paying fines and having to cover customer change fees.

Finally, a source of both cost and revenue, is chargeback processing. Chargebacks come from customers who ordered and received something but then call their credit card company and claim they never ordered the item. Make no mistake, this is not an error made by the agent; this is fraud. But anything the center can do to reduce the possibility of fraud, will increase revenue and decrease the costs associated with responding the fraudulent chargebacks.

So, to conclude this section, now we know where the money is. It is tied up in Direct Labor and in other one-off items where poor call center process capability is resulting in revenue leaks, fines, and costs accruing in other parts of the business.

Drivers of Direct Labor

We know that Direct Labor is the biggest driver of contact center costs. By looking at what drives the amount of direct labor a center carries, we can systematically reduce direct labor expense and, as mentioned, IDL and CapEx, which are correlated with the amount of DL.

The first driver of direct labor is call volume. Simple math, if call volume goes up you need more agents to handle it if you want to maintain your service levels (% Answered in X secs). That is why the self-service options that companies have developed have reduced costs. But another step is possible here. A significant percentage of call center volume is call backs from customers because an issue wasn’t resolved correctly the first time or calls because something wasn’t explained correctly and the customer calls when they get their first bill. We have had clients where an additional 10-20% of their call volume was preventable with some changes to how the agents were handling calls.

A second driver of DL is handle time. The longer the average handle time, the more agents you need to handle the volume. As important as this driver is, with some minor exceptions, call centers have made very little progress in systematically reducing handle time in their centers. The only lever they have is experience, training, and coaching and this has proved remarkably ineffective at decreasing handle time.

A third driver of DL is shrinkage…agents off the phone for any reason. The more the agents are off the phone the more agents you need to make up for the time they aren’t on the phone. Agents get off the phone for breaks, lunches, meetings, training, 1-1 coaching, and after-call work (ACW). Now we are not recommending reducing breaks and lunches and setting up sweatshops. But what if you could reduce the need for training, meetings, and 1-1 coaching and not affect performance? What if you could reduce ACW to near zero and still accomplish the work the agents were doing in ACW? We have seen ACW amount to 15% of average handle time. Reducing ACW reduces the need for DL.

The final big driver of Direct Labor is turnover. Call center turnover is very high…in many centers it was 100% or more prior to the recession. When turnover is high, a higher percentage of inexperienced agents are on the phone. Inexperienced agents have longer handle times, they make more mistakes (contribute to “hidden factories,” and they are off the phones more for 1-1 coaching (more shrinkage). All this drives DL, not to mention IDL in the form of the coaches and trainers and HR people needed to terminate old agents and onboard new ones.

The Solution: Increase Use of Automation

We have discussed how the poor process capability in call centers…agents not doing what they are supposed to do, the way they are supposed to do it…increases cost and results in revenue left on the table. We have also discussed how turnover and shrinkage end up requiring more DL.

There are a host of reasons why agents don’t do what the process requires them to do…they weren’t trained, they forgot, it is too hard to do or remember, they were tired or distracted, they choose not to do it, etc. We could spend a lot of time trying to seal off each of those passages that leads the process to being incorrect or too slow or we could leverage automation which will virtually guarantee the process is performed correctly.

Agent-assisted automation is pre-programmed system actions and pre-recorded audio, directed by the agent to make sure the process is correct. The agent drives the process. If the customer has unique questions the agent uses his/her live voice to address the customer’s needs. If not, the agent uses the pre-programmed paths to resolve the customer’s request.

You build the optimal path and you make it easy for the agent to follow it. This alone will increase the likelihood that the process is done correctly.

But with some simple CRM integration work you can error-proof the process. Error-proofing makes it impossible for the agent to process an order or a customer request unless the key steps…cross-sells, disclosures, credit card checks, mini-Miranda…are carried out. Once the software controlling the automation indicates to the CRM that the steps are accomplished, the order can be submitted.

So what is the theory about how the automation extracts this untapped operational profit. The optimal path you build is obviously error-free and with the CRM integration and the error-proofing, this greatly reduces hidden factories due to call center mistakes.

The optimal path is faster. You have taken out unnecessary steps, streamlined what is said to the customer. Shorter calls reduce the need for DL.

Your pre-programmed system actions allow you to process in parallel which reduces ACW and allows you to lower DL.

The software does the heavy lifting so there is less need form monitors, training and coaching.

The job is easier and more enjoyable for the agents, which lowers turnover, increases call center metrics and reduces the need for IDL.

Results

I listed the theory of the savings above. What are the actual results? Well, the proof of the pudding is in the eating and the eating is good. Going back through untapped profit vein drivers we discussed at the beginning:

  1. Do you have a big cost center doing warranty processing? Are customers returning merchandise authorized by the call center that is out of warranty? Using automation to check the warranty and communicate to the customer that the unit is out of warranty, we helped a high tech company go from a 15% error rate on warranty processing to zero.
  2. Handle Time too long? After Call work too long? Use automation to build the ideal process, to process things in parallel, and to do the work for the agent that the agent was doing at the end of the call. On cell phone activation calls, we reduced AHT by 40% and eliminated ACW which had been 15% of AHT.
  3. Paying fines for failure to leave proper messages and give mini-Miranda rights on Collections calls or failing to recap travel arrangements? We have driven the former to zero defects for every collection agency we work with.
  4. Agents giving lots of reasons for not cross-selling? With automation, the cross-sell happens every time. We increased cross-sell revenue 5X, just by making the offer every time.
  5. Agents off the phone a lot for training and coaching? If the automation is doing the heavy lifting, there is less need for monitoring and training and the agents spend more time on the call. One client fully recovered their investment in our software just through the savings in agent training time. All the other improvements our software generated were pure upside.

If you think this is just a shareholder play and the other stakeholders…the agents and the customers are being hung out to dry…you would be mistaken. The agents love this solution. They get to rest their voices; they don’t have to be as worried that they will make a mistake; they get to concentrate on really listening to the customer as opposed to focusing on trying to remember the dozens of things they are supposed to do.

Use of automation has been shown to decrease agent stress and increase satisfaction. When a senior leader from our client visited the outsourcer using our pre-recorded automation, it was as quiet as a church and the agents seemed less stressed and genuinely happy to be working there. We have not had a chance to study turnover reduction, but we are confident we are affecting this huge cost driver.

On the customer side, every study of customer reaction and satisfaction shows that the customers are completely accepting of automation. First, the pre-programmed system actions, they never see or hear, so no problem there. Few customers even comment let alone complain about the agents’ use of pre-recorded audio. And the customer satisfaction scores are the same for agents using the software and agents handling the call entirely with their live voice.

And why should we be surprised that customers don’t care about the agents use of automation? We hear prerecorded audio in the IVR. We hear it when we pay automated machines in parking lots. We hear it on buses and trains. We hear prerecorded preflight announcements. Does anyone care that the flight attendants don’t do this? No.

To be sure, organizations have taken costs out of their call center operations. In this difficult economic environment, where companies are looking for and turning over new stones to try to find revenue and profit growth, this article should serve as a reminder that deep seams can sometimes run through areas abandoned as no longer productive.

Call Centers: The Still Largely Untapped Operational Profit Vein

I know what you are thinking. “How can he say we haven’t tapped profits from our call centers? We have improved our Website, improved our IVR, and tapped our user community which, taken together, has drastically reduced the number of customer calls we get. So now we don’t need as many agents and our offshore outsourcing has cut our labor rate on the agents we do need. Really, how much value is left?”

The sentiment above summarizes the strategy and results for many companies. It is a good strategy and it has extracted significant savings. However, though it sounds clichéd, it really is the tip of the iceberg…the visible, easy to get at, sources of cost. The lion’s share of the value…revenue and cost…is there for the taking for anyone willing to take a systematic approach. To wet your whistle, I estimate further cost reductions of 40% and call center revenue increases of 3X are easily achievable.

And don’t think that because you outsource and don’t have call centers that this doesn’t apply to you. The outsourcer is charging a price per minute or a price per call. Into that price is built the same inefficiencies you had before you outsourced, in some cases, even more.

Sources of Value

Part of taking a systematic approach is to analyze all the sources of revenue and cost that pass through contact centers and/or that are affected by contact center activity. Once we thoroughly understand the sources of revenue and cost and the drivers of those sources, we can make sure we have plans in place to systematically attack them.

There are four value buckets: three are cost buckets and an “other” bucket that has both revenue and specific kinds of cost not covered by the other three. The three cost buckets are Direct Labor (DL), Indirect Labor (IDL) and Capital Expense (CapEx).

Of these, Direct Labor is obviously and by far the biggest driver of cost. Moreover, the amount of Direct Labor to some extent drives the amount of Indirect Labor (Trainers, Monitors, Coaches, SMEs) and directly drives Capital Expense (Phones, computers, and software licenses for the agents). Drive down Direct Labor and you can proportionately reduce IDL and CapEx.

Before looking at the specific drivers of the amount of Direct Labor, let’s overview what is in the “Other” value bucket. First, poor process capability (agents not handling a call the way they are supposed to every time) results in a significant amount of cross-sell, up-sell revenue being left on the table. Despite the monitoring, the coaching, the incentive plans, etc, agents routinely fail to cross-sell. We have seen situations, where the agents only cross-sell 25% of the time they were supposed to. In situations like these, without improving the sales process one iota, just getting the offer rate near 100% will deliver significant top line benefits.

Another form of revenue left on the table comes from debt collection processes…not reaching the right party, not getting to settlement talks, not agreeing on a settlement number. Improving the collections process instantly puts more money in the hands of the company. (And again, please don’t think “I don’t have to worry about that because we outsource collections,” because your debt collector is leaving your money on the table too.)

In addition to revenue, there are costs in this “Other” bucket. One form is costs accruing in another part of the business due to mistakes being made in the call center. There are numerous small examples, like items returned because an address was entered incorrectly, but a huge source of unnecessary cost comes in the form of warranty processing errors. The agents don’t check the warranty and authorize returns when the warranty has expired, or they have returns sent to the wrong location, or they fail to properly instruct the customer on what to do which leads to other departments having to respond or backstop the call center. These areas are often referred to as “hidden factories” and it is obviously better to stop burning the toast than it is to improve your ability to scrape it.

Third, many companies, especially outsourcers, pay fines either to the government or to their clients for failure to follow the proper process on the call. For example, disclosures not being read to customers can lead to fines (e.g., mini-Miranda rights that collections agents are required to read to debtors.) Further, failure to properly recap calls can lead to outsources paying fines and having to cover customer change fees.

Finally, a source of both cost and revenue, is chargeback processing. Chargebacks come from customers who ordered and received something but then call their credit card company and claim they never ordered the item. Make no mistake, this is not an error made by the agent; this is fraud. But anything the center can do to reduce the possibility of fraud, will increase revenue and decrease the costs associated with responding the fraudulent chargebacks.

So, to conclude this section, now we know where the money is. It is tied up in Direct Labor and in other one-off items where poor call center process capability is resulting in revenue leaks, fines, and costs accruing in other parts of the business.

Drivers of Direct Labor

We know that Direct Labor is the biggest driver of contact center costs. By looking at what drives the amount of direct labor a center carries, we can systematically reduce direct labor expense and, as mentioned, IDL and CapEx, which are correlated with the amount of DL.

The first driver of direct labor is call volume. Simple math, if call volume goes up you need more agents to handle it if you want to maintain your service levels (% Answered in X secs). That is why the self-service options that companies have developed have reduced costs. But another step is possible here. A significant percentage of call center volume is call backs from customers because an issue wasn’t resolved correctly the first time or calls because something wasn’t explained correctly and the customer calls when they get their first bill. We have had clients where an additional 10-20% of their call volume was preventable with some changes to how the agents were handling calls.

A second driver of DL is handle time. The longer the average handle time, the more agents you need to handle the volume. As important as this driver is, with some minor exceptions, call centers have made very little progress in systematically reducing handle time in their centers. The only lever they have is experience, training, and coaching and this has proved remarkably ineffective at decreasing handle time.

A third driver of DL is shrinkage…agents off the phone for any reason. The more the agents are off the phone the more agents you need to make up for the time they aren’t on the phone. Agents get off the phone for breaks, lunches, meetings, training, 1-1 coaching, and after-call work (ACW). Now we are not recommending reducing breaks and lunches and setting up sweatshops. But what if you could reduce the need for training, meetings, and 1-1 coaching and not affect performance? What if you could reduce ACW to near zero and still accomplish the work the agents were doing in ACW? We have seen ACW amount to 15% of average handle time. Reducing ACW reduces the need for DL.

The final big driver of Direct Labor is turnover. Call center turnover is very high…in many centers it was 100% or more prior to the recession. When turnover is high, a higher percentage of inexperienced agents are on the phone. Inexperienced agents have longer handle times, they make more mistakes (contribute to “hidden factories,” and they are off the phones more for 1-1 coaching (more shrinkage). All this drives DL, not to mention IDL in the form of the coaches and trainers and HR people needed to terminate old agents and onboard new ones.

The Solution: Increase Use of Automation

We have discussed how the poor process capability in call centers…agents not doing what they are supposed to do, the way they are supposed to do it…increases cost and results in revenue left on the table. We have also discussed how turnover and shrinkage end up requiring more DL.

There are a host of reasons why agents don’t do what the process requires them to do…they weren’t trained, they forgot, it is too hard to do or remember, they were tired or distracted, they choose not to do it, etc. We could spend a lot of time trying to seal off each of those passages that leads the process to being incorrect or too slow or we could leverage automation which will virtually guarantee the process is performed correctly.

Agent-assisted automation is pre-programmed system actions and pre-recorded audio, directed by the agent to make sure the process is correct. The agent drives the process. If the customer has unique questions the agent uses his/her live voice to address the customer’s needs. If not, the agent uses the pre-programmed paths to resolve the customer’s request.

You build the optimal path and you make it easy for the agent to follow it. This alone will increase the likelihood that the process is done correctly.

But with some simple CRM integration work you can error-proof the process. Error-proofing makes it impossible for the agent to process an order or a customer request unless the key steps…cross-sells, disclosures, credit card checks, mini-Miranda…are carried out. Once the software controlling the automation indicates to the CRM that the steps are accomplished, the order can be submitted.

So what is the theory about how the automation extracts this untapped operational profit. The optimal path you build is obviously error-free and with the CRM integration and the error-proofing, this greatly reduces hidden factories due to call center mistakes.

The optimal path is faster. You have taken out unnecessary steps, streamlined what is said to the customer. Shorter calls reduce the need for DL.

Your pre-programmed system actions allow you to process in parallel which reduces ACW and allows you to lower DL.

The software does the heavy lifting so there is less need form monitors, training and coaching.

The job is easier and more enjoyable for the agents, which lowers turnover, increases call center metrics and reduces the need for IDL.

Results

I listed the theory of the savings above. What are the actual results? Well, the proof of the pudding is in the eating and the eating is good. Going back through untapped profit vein drivers we discussed at the beginning:

1) Do you have a big cost center doing warranty processing? Are customers returning merchandise authorized by the call center that is out of warranty? Using automation to check the warranty and communicate to the customer that the unit is out of warranty, we helped a high tech company go from a 15% error rate on warranty processing to zero.

2) Handle Time too long? After Call work too long? Use automation to build the ideal process, to process things in parallel, and to do the work for the agent that the agent was doing at the end of the call. On cell phone activation calls, we reduced AHT by 40% and eliminated ACW which had been 15% of AHT.

3) Paying fines for failure to leave proper messages and give mini-Miranda rights on Collections calls or failing to recap travel arrangements? We have driven the former to zero defects for every collection agency we work with.

4) Agents giving lots of reasons for not cross-selling? With automation, the cross-sell happens every time. We increased cross-sell revenue 5X, just by making the offer every time.

5) Agents off the phone a lot for training and coaching? If the automation is doing the heavy lifting, there is less need for monitoring and training and the agents spend more time on the call. One client fully recovered their investment in our software just through the savings in agent training time. All the other improvements our software generated were pure upside.

If you think this is just a shareholder play and the other stakeholders…the agents and the customers are being hung out to dry…you would be mistaken. The agents love this solution. They get to rest their voices; they don’t have to be as worried that they will make a mistake; they get to concentrate on really listening to the customer as opposed to focusing on trying to remember the dozens of things they are supposed to do.

Use of automation has been shown to decrease agent stress and increase satisfaction. When a senior leader from our client visited the outsourcer using our pre-recorded automation, it was as quiet as a church and the agents seemed less stressed and genuinely happy to be working there. We have not had a chance to study turnover reduction, but we are confident we are affecting this huge cost driver.

On the customer side, every study of customer reaction and satisfaction shows that the customers are completely accepting of automation. First, the pre-programmed system actions, they never see or hear, so no problem there. Few customers even comment let alone complain about the agents’ use of pre-recorded audio. And the customer satisfaction scores are the same for agents using the software and agents handling the call entirely with their live voice.

And why should we be surprised that customers don’t care about the agents use of automation? We hear prerecorded audio in the IVR. We hear it when we pay automated machines in parking lots. We hear it on buses and trains. We hear prerecorded preflight announcements. Does anyone care that the flight attendants don’t do this? No.

To be sure, organizations have taken costs out of their call center operations. In this difficult economic environment, where companies are looking for and turning over new stones to try to find revenue and profit growth, this article should serve as a reminder that deep seams can sometimes run through areas abandoned as no longer productive.

Thursday, September 29, 2011

The Unturned Stone: The Untapped Profits Hiding in the Poor Quality of Modern Call Centers

Attacking Call Volume

In the last post we did a show me the money analysis to seewhere all the untapped cost and in some cases revenue was lying:

(Direct Labor x DLRate) + (Indirect Labor x IDL Rate) + (Capital Expense X CE Rate) + (Other) =Total Opportunity Cost of Call Center
The keys from that discussion were:

The industry's efforts to date have been focused on rate reduction…move to lower cost markets and your labor and real estate costs go down.

2) Direct Labor is the biggest bucket of cost.

3) Though Indirect Labor and Capital Expense havetheir own drivers, the amount of Direct Labor is the biggest driver of thosetwo expenses.

4) The catchall “Other” bucket of cost and revenue in some cases drives the other costs and in some cases is a cost or revenue lost category of its own.


So the biggest opportunity area is the reduction of absolute amount of direct labor. To unlock thevalue, we will need to look at the drivers again. As stated, the amount of direct labor youneed is a function of Call Volume, Handle Time, Service Levels, and Shrinkage (Time Not Spent on the Phone). Let’stake each in turn.

Sine Qua Non: A CallVolume Pareto

That call volume is a big driver of cost is not news to callcenter leaders. They have been trying toreduce the volume of calls by implementing more self-service channels…websites,Interactive Voice Response, and Community Support Boards.

But the way to tell if a company is trying to systemicallyand systematically go after call volume is if they can produce a pareto chartof call volume…a graph that shows the call volume and handle time associatedwith exactly why customers are calling.

Tracking Handle Time and Volume together allows you to orderthe opportunities in terms of Volume X Handle Time. That way low volume calls that may take along time to solve have as much weight as high volume calls that are easy tosolve.

There are a lot of ways to proceed from here. One way simple way is to break the list intothree buckets: 1) Calls due to NotResolved the First Time or Call Center Errors, 2) Calls that are preventablewith a process/procedure change in call handling, 3) Calls that are preventablewith a change in another part of the business and 4) first time requests forservice that are not preventable.

Let’s take the last category first. You set up a call center to service yourcustomers. That is what you are therefor! If a customer has a change ofaddress or a death of a spouse or moves, it is perfectly legitimate for them tocall and request service. You should ofcourse offer web or IVR self-service, but some customers just want to talk tosomeone and that is fine. This is a lowpriority bucket to worry about.

The bucket you must focus on first is Calls due to NotResolved the First time or Call Center Errors. If customers are calling back because of an error you made, you areadding to your own cost and you just have to fix it immediately. An example is a customer calling to fixsomething about their bill only to have the same problem appear the next monthbecause the agent forgot to do something in their system. Another example is a tech support call thatdoes not resolve the customers issue due to the agent not following the properdiagnostic path. Both are pure waste.

The second bucket is Preventable Calls with a ProcessChange. A good example here is the highpercentage of customers that call after they get their first month’s bill. If the proper disclosures are done and stepsare taken to ensure understanding, many of these are likely preventable.

The last category to be discussed is calls that arepreventable with a procedure change in another part of the business. There are lots of examples here as callcenters are often the place where all the poor quality problems show up. Product defects generate calls. Problems with field work generate calls. Problems with billing generate calls.

Armed with a Volume X Handle Time Pareto Chart, you can goto other parts of the business or to the GM over the whole business andhighlight the cost associated with errors in other departments. Having specific cost detail helps increasethe focus on finding a fix.

Of course another approach is to just develop a strategy forthe top Volume X Handle Time drivers. Fromthere, assign resources to come up with coordinated strategies forsystematically attacking each driver, no matter what bucket it is in.

But at the end of that analysis, you will have a list of processchanges you want to make and you will a list of things you want to make surethe agents do/say every time. Whatnow? Pull the agents off the phone fortraining? More monitoring and coaching? Put a sheet of paper on the agent’s chairstelling them to do it this way instead of that way?

This is what we have been doing and it is not working. Call center leaders have to think aboutautomating and error-proofing work with agent-assisted automation (http://en.wikipedia.org/wiki/Agent-assisted_Automation). Build the proper “first bill” set ofdisclosures, prerecord them and let the agent play them. Automate the updates in a second system whenthey are made in the first system so the agent doesn’t have to remember to doit. Develop the proper diagnostic pathfor a particular problem and let the agent play the questions to the customerto ensure the proper steps are taken.

It is the overreliance on monitoring and coaching thatcreated many of these problems. Lookingto monitoring and coaching to fix them, is, well, you decide how smart that is.

But however you do it, the big picture idea to not lose sight of is that because callvolume is the big driver of Direct Labor and because Direct Labor is the bigdriver of Indirect Labor and Capital Expense, systematically reducing callvolume is going to have a multiplier effect on cost reduction.

Monday, September 26, 2011

The Unturned Stone: The Untapped Profits Hiding in the Poor Quality of Modern Call Centers



The call center industry has been around for about four decades. When it began, call centers and mail were the only ways for customers to get their issues resolved. Over the last couple decades, websites, self-service kiosks (e.g., ATM’s), chat and many other channels have been developed for our customers to get help.

So by one measure, the ratio of customer touches per call center agent has gone through the roof and the service sides of these businesses are dramatically more productive. The number of call center agents may or may not have dropped depending on the growth of the business, but businesses are supporting more revenue with less call center agents.

Moreover, the cost of those agents has also dropped dramatically relative to other costs. This is because businesses keep moving their call centers to lower cost labor and facility markets, including, for better or worse depending on your perspective, offshore.

So more touches, relatively less agents, and at a lower cost equals success right? If close enough is good enough for you, then declare victory. But if stopping in the middle of a mountain climb is not your idea of success, then this article about the dramatic amounts of cost and revenue being lost in modern contact centers will be of interest.

Show me the Money

A quick overview of drivers of cost and revenue in contact centers will help reveal where and how this money is being tied up.

At a high level, the total cost of contact centers is:

(Direct Labor x DL Rate) + (Indirect Labor x IDL Rate) + (Capital Expense X CE Rate) + (Other) = Total Cost of Call Center

Now, of course, there is no one rate for Direct Labor, Indirect Labor and certainly not for Capital Expenses, but this rate idea is just there to highlight that costs are affected by those rates.

And this is in fact what businesses have done to reduce their call center costs. They have moved to lower cost labor markets. There used to be call centers in Silicon Valley and New York and Boston, but not anymore as the real estate and labor costs are too high. So now you find call centers in more rural areas of the US and even offshore to further reduce the costs. Can we continue to reduce those labor and real estate rates, yes, but as other parts of the world become more developed the rate of rate reduction is rapidly diminishing.

So that leaves the amount of direct labor, the amount of indirect labor, the amount of Capital Expense, and Other. What drives them?

As you will soon see, direct Labor is the biggest piece, so let’s start there.

Direct Labor

As an absolute number, the amount of direct labor is the biggest bucket of cost. At a conceptual level, the amount of direct labor you need is a function of Call Volume, Handle Time, Service Levels, and Shrinkage (Time Not Spent on the Phone). Let’s take each in turn.

Call Volume. This is one of the two big drivers. In general, if the call volume increases you need more agents to handle it. Your call volume could double because your revenue is doubling. This might not be so bad. However, your call volume could double because there are problems with your product or service. This will be expensive.

Handle Time. The second big driver is Average Handle Time. If the AHT per call is 10 mins, you will need a lot more direct labor than if the AHT is 1 minute.

Service Levels. If you want to answer 95% of your phone calls in under 30 seconds, you will need more agents to be waiting around in case the call volume spikes. If you decide answering 95% of your calls in less than five minutes is good enough, you will need less.

Shrinkage. If all your agents never leave their desks and you got 8 full hours of productivity out of each of them, you would not need as many agents. But they have to have breaks and lunches so you need more. In addition, there is the time they are off the phone for training and meetings and other call center functions. The more off phone time, then the more agents you will need.

Indirect Labor is all the people in the call center that are not on the phones: Coaches, Subject Matter Experts, trainers, monitors, IT personnel, etc. Capital Expense is all the technology used in the call center, and sometimes real estate. Generally anything that is depreciated.

Now lest this begin to get too complex, the good news is that the big driver of Indirect Labor and Capital Expense for Call Centers is the amount of Direct Labor. While not the whole story, the more people on the phones the more coaches that are needed, the more monitors, the more desks and computers, the more software licenses, etc.

We are left with one rather non-descript bucket: Other. The other bucket has important cost drivers that can be more challenging to quantify, but are important to look at. Here are some of the biggest.

The Effects of Poor Call Center Performance. We could get really theoretical here and try to quantify how poor service agent performance affects Customer Loyalty, share of wallet, decision to renew, etc. This is important and should absolutely not be overlooked, but we want to be more concrete in this article.

One effect of call center performance that adds up quickly is sales rate. Let’s say the upsell rate in a call center averages $5 a call. Not bad. What if you find out that the agents are trying to upsell on only 20% of the calls? That is a lot of money being left on the table.

Another quantifiable aspect of poor call center performance are any costs incurred anywhere in the company when the contact center makes a mistake. A common example is warranty processing. What if the agent does not check the warranty and the customer sends the unit in to be fixed and it is not under warranty? What if the agent does not check the customer’s address and the goods are sent to the wrong location and have to be resent internally? What if the agent does not remind the customer to remove their software before returning the console and the customer writes to complain and employees have to spend time working that issue?

Turnover. Turnover is a huge driver of seen and unseen cost. The part that is easy to see is that if I have a lot of turnover, I need more indirect labor to hire and train them and they spend a lot of time off the phone getting coached. The part that is hard to see is that, in general, the more inexperienced my agents are, the slower they are and the more mistakes they make. High turnover centers are much more expensive to run than low turnover centers.

Fines. Some call centers have to pay fines due to non-compliance. This can be true of any call center, but it is especially true of outsourcers. Outsourcers often pay fines to their clients for a failure to perform all the call steps correctly, especially recapping the call.


Over the next few posts, I will describe specific ways to attack each bucket.

Friday, September 23, 2011

On Twitter

I am not sure why it took me so long, nor why I decided to sign up, but you can follow me on Twitter at: DennisAtKombea.

Also, I am about to begin a long series of posts...all under the title of The Unturned Stone...about the slacked-jawed amounts of money corporations are leaving on the table because the call centers they run and outsource to are so staggeringly inefficient.

Stay tuned...

Wednesday, April 13, 2011

Large, Occasional Data Breaches are really, really bad. The Small Daily Ones are even worse.

Another compromised data center is making the headlines. Epsilon, the online marketing company was hacked. This time, it appears only email addresses were obtained. With luck, the worst those affected will experience is an increase in email marketing spam.

Could more be done to protect company data centers? Yes, always. But it is worth asking if we should be spending more money to further lock down the back door to the data centers, especially when the call center "front door" is banging in the breeze.

Email lists were compromised at companies like HSN Inc, Scottrade, Marks and Spencer and many others. In addition to storing sensitive information, these companies have tens of thousands of call center agents around the world at their own centers and at outsourced locations that are taking credit cards and other sensitive information over the phone from their customers every day. In addition, there are thousands more coaches, team leads and call monitoring personnel that have access to recordings of calls where customers' personal information can be easily accessed. None of these people have access to the secure data center, but they have access to a blinding amount of identity and financial information.

In the thousands and thousands of call centers around the world, there are few controls consistently in place that prevent employees from recording sensitive information with a recording device or on a computerized or paper note pad. Even if safeguards existed, nothing can stop agents from simply memorizing the important details.

Not only can this information be stolen, it is stolen. Thousands of times a day, everyday sensitive customer information is recorded, copied, memorized and stolen. It can be used by the person who stole it or the information can be sold to individuals who in turn sell the information to others, with devastating consequences:
Overseas Credit Card Scam Exposed.

Despite the insidiousness and pervasiveness of the problem, companies do little to prevent it because the problems rarely come to light on a large scale and because the breaches are extremely difficult to trace back to the offending agents and the companies where they are employed.

Even more maddening is that simple solutions are on the market which allow agents to "collect" the private information without ever seeing or hearing it. For example, the customers can be transferred to an Interactive Voice Response (IVR) system, enter their information, and then get transferred back.

This IVR option has been around forever, but is rarely used. Part of the reason is lost sales. You are right on the edge of booking a customer’s order and you have to transfer them to the IVR to enter the credit card information. During that transfer time and with the agent not on the phone, it is easy for customers to rethink their purchases.

It is also not always the smoothest transition. The customer can be left waiting for the agent to pick up or the agent can be left waiting for a customer who has already decided not to go through with the purchase and has hung up.

Newer applications provide for a better customer experience and are easier to implement. SecureCall, our CRM plug-in, allows customers to enter their card information over the phone directly into the Customer Relationship Management system while the agents remain on the phone. The DTMF tones are converted to monotones so the agent cannot record or decipher the numbers.

As a side note, though the primary purpose for considering and implementing one of these solutions is to prevent agent fraud and credit card theft, these solutions also help to fight fraudulent chargebacks. In a fraudulent chargeback, the customer is claiming they never ordered the goods they were sent. When deploying solutions like these, part of a merchant’s argument is that their agents don’t have access to credit card numbers and don’t enter credit card numbers, so the customer must have provided this information. As a result of this argument and other features, these solutions have helped merchants overturn a higher percentage of fraudulent claims.

If the reduction of agent fraud and consumer fraud are areas you want to shore up, you should look for solutions that:

1) Provide a great customer experience

2) Don’t affect close rates and average order size

3) Are easy to implement, including the training of the agents

4) Are completely PCI Compliant

5) Are cost effective


The lack of front door security around the call center wreaks financial havoc on millions of unsuspecting consumers. With the simple solutions that exist out there, this is inexcusable. It is also just a matter of time before this glaring leak blows up some company's front porch.