From Brian's Desk

Contact Center Excellence – Stand Out From The Crowd

Archive for the month “October, 2012”

Contact Center Big Data

Big Data seems to be running in a pretty tight race on the media-scope with virtualization, right behind Cloud (my previous post).   What is Big Data?   Put rather simply, it’s a collection of data that is too large and too complex to be housed in or processed by a data base management system.   Typically, Big Data is a combination of existing data sources that are somehow related.  The task of analytically processing the Big Data is to discover latent or unobvious relationships  between sets of data, to discover ‘business truths’ if you will.

Big Data is not only structured, record-like data, but unstructured data as well.  Consider all of a company’s emails in it’s email management system.  There may well be important relationships and suppositions one can make if all of the emails could be read and understood.  Fortunately, tools are now available that allow an enterprise to consider such a task. Without software tools, the task could take untold resources and time in a large enterprise, and the return on the intelligence gained is likely to pale in comparison to the cost.

We clearly have a large amount of unstructured data generated within the contact center.  Chat transcripts, emails, ticket documentation, and voice recordings.  Data and voice analytics software is now available to allow the contact center organization to discover much more about the company’s customers than nearly any other organization within the company, probably the only exception being POS (point-of-sale) data.  But even POS data does not capture product likes and dislikes, and brand sentiment like contact center data.  And, powerful as these sources are by themselves, bringing together the POS and contact center data can lead to further discoveries, ‘business truths’ and ‘customer truths’.

I made the case in an earlier post ( that instead of call or contact shedding to cut costs, the contact center should be encouraging customers and potential customers to call or make contact via chat, email, SMS text or otherwise.  Where are we shedding to?  An IVR?  Self-service on the web or a self-service mobile app?  How much valuable data is being lost when we do so?  Let’s consider that the contact center agents are frequently one of the least paid groups of staff you have;  how much more expensive is it for marketing to have these customer conversations?  I contend the more data you have from your customer, the more intelligence you have about the customer, the more ‘truths’ you can discover.  An IVR is mostly a one-way conversation, and how many call centers are recording the IVR interactions?  Web self-service can give you clicks and potentially point out shortcomings in the self-service process and information.  However, you are left guessing about sentiment, likes and dislikes.

Big Data right now is a key competitive differentiator as few know what it is and how to use it to best discover ‘truths’.  Are you going to be a leader or a follower?


Is Cloud-Based For You?

The weather report for the contact center industry is – mostly Cloudy.   The promises from the vendors are all there; let us manage your call center hardware and software so you can focus on managing your business, it’s lower cost and faster to implement, you get on-demand capability, you get automatic failover.   In general the promises are valid.  IF your cloud vendor has a full contact center solution or sharply delineated functional boundaries and IF you don’t need to integrate with your back-office applications, then Cloud might just be appropriate.

A full contact center solution entails everything; the phone switch or PBX, the IVR, the ACD, the agent desktop, the eMail system, the chat application, call and screen recording, knowledge base, real-time monitoring, and workforce management.  Anything less is not a Cloud contact center solution, it is a point solution.  A Cloud solution that delivers a call into the center is a point solution.  To the extent it is functionally complete and has few integration points it could be useful.  If your center is an integrated multi-channel contact center, the point solution needs to not only deliver calls, but chats, eMails, and any other inbound channel you have.  That point solution also needs to handle your outbound channels.  I call this a contact management solution.   Integration is the enemy of Cloud solutions.  If you are looking at a Cloud-based contact management solution and it doesn’t have all of the functionality mentioned above, you are left with integration that is potentially expensive to develop and implement, and potential run-time latency problems.

Desktop Cloud puts the agent desktop or CRM application in the Cloud.  As previously stated, integration being the enemy of a Cloud solution, the desktop Cloud has little practical advantage for the contact center.  It needs to integrate with the contact management solution, it needs to integrate with a screen/voice recording solution, with the WFM solution, potentially the company’s eMail application and chat application in addition to back-office applications and databases.

For the same reason the Desktop Cloud has little practical value, other solutions sometimes disguised as Software-As-A-Service, or SAAS, that provide point WFM, KB or QA solutions also provide little or no value.  There is simply too much integration work required.

Even for the full Contact Center solution, integration requirements with back-office applications and databases can sway a decision.   Dig deeply into vendor claims regarding integration.  Press them for examples and current customers.  One problem is finding a full Contact Center Cloud solution.  No software vendor does it all.  They need to partner and integrate, but this will be done in the Cloud so that a full solution can be offered.  This is typically al-a-carte and done on a customer-by-customer basis.

Primarily what I see on the market are either Call Management or Contact Management Cloud solutions.  Again, focus on clean integration and clean handoffs.  You will typically need integration with the agent desktop and with quality recording applications.  Ensure your requirements relative to response and latency are well understood by your vendor.

Cloudy is okay, just make sure the Cloud doesn’t rain on your parade.


I was at Contact Center Association’s fall conference last week and saw a presentation on gamification in the call center.  This is basically gaming for the call center agents, not the customers visiting company web sites.  As I sat and listened to the talk about tokens or tickets being handed out for every minor bit of good behavior (like showing up for work on time), I couldn’t help being a bit concerned that contact centers might turn to gaming and entertainment to mask engagement or morale issues.

In my book “Contact Center Excellence” I cite work from Gartner on engagement; in the average organization there are fewer than 2 actively engaged employees for every actively disengaged employee.  That means if you have a 1,000 seat contact center and you believe you have 400 actively engaged staff, you will also have over 200 actively disengaged staff.  The actively disengaged are doing their best to disrupt established processes and programs, and are the ones with the highest unplanned shrinkage, lowest quality scores, and lowest overall operational scores.  They typically do not like award programs, and will do what they can to subvert the programs.  Thus, gamification is not going to win over the actively disengaged and is therefore not going to correct an engagement problem within the center.

There are many approaches to improving engagement, and Gartner lists eleven such approaches, and I tailor that list in the engagement chapter in my book. Entertaining awards like gaming may help improve engagement, but the contact center would be wise to understand the current level of engagement across their center, and put a comprehensive plan in place to improve engagement.  One component is certainly awards and rewards, but putting a fancy and fun set of games in place by itself is not going to boost morale or engagement.

Following one of my posts in which I am unable to cite a quote, one of my readers assisted by referencing H.L. Menken.  I bring up the quote again, because this is another case where it applies – the topic of engagement; “for every complex problem there is a solution that is simple, elegant, and wrong.”

So, who is up for arguing with me?

Why We Don’t Want to Hear From You

For those companies that do not have a brick-and-mortar presence, why don’t you want me to call you?  With the availability of voice and data analytics, I contend a 10 minute phone call is very valuable.  We get to understand brand sentiment, we get input on customer product likes and dislikes, we get input on customer service likes and dislikes, we get to ask questions and capture responses.  How much usable and actionable information do you get when a prospect visits your web site?  Sure, you might get an understanding of how many visitors are clicking on what links, what pages are visited most often, how long the visitor stays on a page and the path visitors take through your pages.  Sure, the cost of collecting this data is low, but so is the robustness and value of the content.

So, why are customer service contact centers so maniacally focused on talk time or handle time?  Why are they so focused on cost reduction and ‘shedding’ calls?  And, where are these calls going?  To the company website where the quality and robustness of the data we can get is minuscule in comparison to what we can get in a phone call?  If a 5 minute call is valuable isn’t a 10 minute call more valuable?  Sure, maybe not twice as valuable, but you really need to ask; do we want to stop at 5 minutes?  Time for a reality check here.  If you are paying $12 per hour for your agents, 5 minutes cost you $1.  How much is your company spending on your marketing budget to get that 5 minutes of intelligence?  Oh, so you are shedding your calls to your IVR are you?  Great if you are using a voice-recognition IVR and you are recording and running voice analytics on your IVR interactions.  If not, what intelligence are you gathering?

Raise your hand if you are in marketing and would like to have more intelligence about your customer.  Keep your hands raised if you are spending a good deal of money getting the intelligence you need, and keep them raised if you are in fact getting the intelligence you need.

So, instead of focusing on cost per call, now more than ever as companies are primarily virtual, why are we not focusing on value per call?

Why Don’t They Understand Me?

Let’s face it, the world is not getting simpler.  Technology is more robust and complex.  Each generation has its own intimate relationship with technology, and rarely keep up with the changes easily embraced by the next generation.  I am a baby-boomer and frustrated by those of the post-WW II generation that typically refuse to learn or even use what I am comfortable with; email and PCs.  And, there is no way I am as comfortable using the latest technology that is second-nature to my Gen-Y children.  I have a different opinion than some as I believe younger people are simply more tolerant.   This is in contrast to the thinking that the younger mindset is; if I can’t figure out how to use something, there are so many other options and I’ll just go and get one of those.  There is some truth to that argument, however the way in which generations approach the problem is very different.  If I can’t figure out how to use something, I’ll pick up the phone and actually speak with a human.  My introduction to technology was the Commadore-64.  Subsequent technology was progressively easier to use and of higher quality.  Software development was expensive (and it costed big $) and therefore rigorous processes were developed focused on design and exhaustive testing.   Six sigma methods were applied to reduce errors, and by and large, the software was of progressively higher quality.  This has all changed very dramatically in the past 10 years.  Software (apps) are now developed by freelance programmers, following no standards in terms of design and quality, and many are either free or available at a very low price.  For those that have grown up in this wild-west of apps, they are used to apps not quite working as designed and having no support available if they find errors.  Apps are a commodity; if the one I just purchased for $0.99 isn’t working they way I want it to work, there are so many $0.99 alternatives.  Therein lies my contention that post baby-boomers are more tolerant of the current technology.

So, for us baby-boomers, and there are a good many of us, we have a higher expectation of quality in technology products and less curiosity to figure out how it works.  In short, we simply have a different mindset regarding technology than younger generations.  I have placed a good number of calls into support organizations in the past couple of years because (1) either I don’t understand technology, or (2) technology doesn’t understand me.  I recall a time when my wife and I were driving from Raleigh, North Carolina to Baltimore, Maryland.  We each had a cell phone and each had the same Driving Directions app on our phone.  You would imagine that the same Driving App from the same cellular carrier would guide us on the exactly the same route to our destination.  Well, in short, no.  As we neared our destination, the app on one phone was guiding us on a completely different route than the app on the other phone.  I wasn’t quite sure if the case was (1) I didn’t understand the technology, (2) the technology didn’t understand me, or (3) perhaps the technology simply didn’t work.

I am specifically focusing on technology as I see that as the largest gap of cross-generational contextualizing.  My daughter can’t understand why I can’t understand.  The median age of workers in a technical support call center are of my daughters age.  Just like with her, they don’t understand why I don’t understand.  No amount of training on how to troubleshoot the technology is going to fix that problem.  Being able to take apart the widget and put it back together is not going to bridge the life-experience gap.

This is very much like the cultural gap experienced during the last decade when companies figured they could easily outsource calls to countries overseas.  This was bumpy to say the least, and many companies lost brand affinity because of it.  The match of skills was not the issue.  The issue was culture and the lack of cultural understanding and context as calls moved to countries with cultures very different from the countries where the call originated.  The second half of the last decade was devoted to overcoming the cultural divide.  When I called, I was now speaking to “Bob”, clearly with an accent foreign to mine.  The out-of-country call center had newspapers (remember those?) and TV piped in from the originating country.  There was specific training focused on accent neutralization and cultural awareness.  Was it successful?  Surely more successful than if nothing had been done.

Should the generational divide be treated as a cultural divide?  I think the answer is an emphatic “yes”.  It is more than just teaching technology and teaching generic soft skills such as engagement and empathy, although I cannot understand how one can truly engage a caller without understanding the cultural context.  Much more important in the engagement process is understanding the generational context of the caller.  If the caller has self-identified in the IVR and our CRM app knows the age of the caller, perhaps we have at least a starting point for the context setting.  But, how do we do this at the call center if we know nothing about the caller until the call is connected to the agent?


Rework and Errors

Rework in the contact center can take many forms.  Rework means spending additional resources doing work that should have been completed correctly the first time.  There are various reasons for the rework, it is measured in different ways, and remedied in different ways.  Rework means having to retrain staff because the training, for whatever reason or reasons, was not effective the first time it was given.  Rework means answering a second or third call from the same customer that should have been handled the first time.  Rework means data correction in the back office because data was either captured or entered erroneously.

Within the industry, we have FCR to measure the first type of rework mentioned above.  There are various approaches to addressing FCR problems; additional training, more robust knowledge, and asking the caller if in fact the call was resolved before ending the call are several.  Many have told me that the largest gains they witnessed in FCR occurred when Handle Time, or at least Talk Time, was removed as an agent-level measurement.  Without the time pressure, agents are able to spend the needed time on a call or contact and are therefore more likely to resolve the query or issue.  I recall a quote, I really cannot attribute unfortunately, but it states “for every complex problem, there is a solution that is simple, elegant and wrong.”  I strongly advise leaders to really understand their Talk Time, and their After Call Work and not make a simplistic and hasty move to change what and how efficiency is measured and at what level.  In my book, “Contact Center Excellence”, I make the case for managing variability and analyzing outliers rather than managing to averages.  Leaders need to have a good understand of how their Talk Time varies across the agents in the center, and then understand why.

Rework in the form of retraining can be very costly for a contact center, not only in terms of the time of the phone costs and training team costs, but cost to customer satisfaction, and perhaps back-office costs.  Sometimes training effectiveness can be measured by FCR; a low FCR across the center for a given call topic is a good indicator that training was inadequate.  Call (or email or chat) transfers can also be an indicator.  And, if the transfers are to back-office staff, or chats to back-office staff with questions, you will hear about it from the back-office leadership.  At the last center where I was a leader, we had a productivity tool that captured all chats; the reason code for the chat, the transcript of the chat, and an indicator from the ‘chatee’ if they felt the chat was needed as the result of a knowledge or training gap.  Management could quickly determine if there was a training issue.  Online training has become a very good alternative to dragging staff into a training room and can be taken during downtime or during breaks scheduled for that purpose.

Rework due to errors can have a major effect on downstream business processes.  Consider a sale made with an error in the shipping address.  There may be additional expense to the company to reship a product, and potentially additional remediation expenses for a customer dissatisfied that their package did not arrive on time.  Errors in gathering the correct insurance information at patient registration can have significant impacts on a healthcare provider’s revenue cycle.  It is always easier and less costly to fix an error before it leaves the contact center.  Much better to have your contact center application perform the correct edits and audits automatically.  The technology exists to ‘watch’ the keystrokes at the agent desktop and either automatically correct errors, or display an alert with an explanation.  (Pardon the commercial plug, but Cincom’s Synchrony product does have this capability.)

In conclusion, measure the amount of rework you have then take steps to minimize it.  What are you doing to eliminate rework?

A Look At Utilization

Utilization, quite literally, means “put to use”.  When agents are idle, they are not “put to use” and the idle time is therefore just cost.  The task of a contact center manager is to increase utilization thereby reducing idle time and cost.  Let’s face it, operating a contact center is a numbers game.  The higher the contact volume and the more uniform its distribution throughout the day, week, and month, the greater the chance of having little idle time.  The opposite is also true.  The lower the contact volume, and the more variation in the arrival of the contacts, the more idle time you will have.  Therefore, a single ‘pool’ of agents will always have less idle time, and a higher utilization, than if the single pool is broken up into smaller pools.

For a variety of reasons, sometimes valid, contact center managers break up the single pool by segmenting agents according to skills or channel and hence create multiple smaller pools of agents.  This segmenting is valid when the skills needed to handle the various types of contacts are very different.  This difference in skills may be of the more vs. less form, i.e. account balance vs. researching an erroneous entry on a bank statement.  In this case, presumably the account balance inquiry can be handled by an agent with lower skills, and at a lower cost.  The difference in skills costs may well outweigh the cost of a lower utilization.

The skills difference may be reflective of the diversity of the business, for example the skills to handle a mortgage refinance question are very different than the skills to handle a question relating to life insurance options.

Or the skills difference may be related to communication, some agents are just better at speaking and others are better at typing (to handle chats and eMail).

Whatever the skills differences, be sure they are significant enough to warrant the decrease in utilization.  Run your staffing models using a combined pool and separate pools of agents so that you are well aware of the cost of operating with the separate pools.  If the cost due to the utilization decrease is more than you want to see, try cross-training and/or improving your knowledge base to get closer to a single pool model.  The larger the pool, the more utilization you will have, so first exhaust other options before coming to the conclusion to segment your agents.

Quotes For My Book

“In this book, Brian Flagg makes the case for a powerful transformation of the support center. He shows why metrics-based thinking isn’t sound strategy, why the combination of talented people and robust knowledge makes a difference, and how employee engagement saves money—big money. He digs deeply into each area covered in the book. If you want to step into a new era of support center, this book is an essential reference. Keep it close at hand; I know I will.” – Roy Atkinson, well-known writer and analyst on support center and customer service topics.


“Brian offers a wealth of solid information and thought leadership on how to create and sustain a successful contact center operation. Brian explains in language we can all understand what works and what doesn’t. This book is a must read for any contact center manager interested in taking their operation to the next level.”  – Barbara Burke, Consultant and Author of The Napkin, Melon and the Monkey.


“I would recommend reading and referencing this book, not only the front line management (tactical perspectives)  but the senior level management (strategic perspectives) for better focus on the value of servicing the Customer  (service, sales, experience, loyalty, etc.).”  – David E. Hadobas President, CEO and Founder CCNG International, Inc.

More Cost Savings Through Efficiency

What better way is there to reduce effort than through automation?  By definition, automation requires little or no effort.  I led a large service desk for a retailer, and one of the tools my team developed was a dashboard.  When a call came in, through phone digit information we were able to get the store # of the calling store, and then use CTI to ‘ping’ every network device in the store, reach into the change management system to pull any changes in that store in the last 48 hours, and reach into the service ticket database to gather open high severity tickets for that store location.  A multi-tabbed dashboard was ‘popped’ to the agent 2 seconds before the call reached the headset.  By the time the agent gave his or her greeting, the agent knew who was calling and had a pretty good idea why they were calling.

Similarily, while in the interaction, data can be automatically moved from one screen to another and used to prepopulate screens, or ‘push’ search terms to another part of the screen or another screen.

So, when brainstorming approaches to efficiency, don’t forget the most efficient method; automation!

Saving $ Through Efficiency defines efficiency as “accomplishment of or ability to accomplish a job with a minimum expenditure of time and effort”.   We can therefore maximize efficiency by minimizing the time spent on a task and/or the effort spent on the task.  By maximizing effciency, we lower cost.  The task, in contact center terms, is the interaction. Time spent on the task is time handling the interaction, and effort is the agent’s effort handling the interaction.  The distinction is made here between the two as the agent time spent is directly impactful on the customer’s perception of the interaction.  A call that is longer than it needs to be keeps the customer on the phone for more time than they believe is necessary, and eMail that goes unanswered for three days keeps the customer from getting an answer to a query or a problem addressed for a longer time than they view as necessary.

Agent effort is more indirectly impactful for the customer.  They may or may not understand when an agent is struggling with their technology or searching through an ineffective collection of knowledge, especially in an offline channel (eMail).  However, I contend that for phone, chat and SMS channels, effort spent is pretty well correlated with time spent.  The reverse is not so well correlated.  An agent can spend too much time trying to reach an engagement score on the QA scoresheet by simply, pardon the technical term, chit-chatting, and spend no more real effort than an agent sitting alongside that is much better at engaging a caller.

Minimizing the effort spent during an interaction means, quite simply, making the resources an agent needs to use during an interaction more robust and easier to use, and more supportive of the expected flow of the interaction.  Unfortunately, contact centers don’t have a good measure, or even any measure, for effort other than correlating with time, and hence efficiency is measured as Handle Time (Talk Time + After-Call Work Time).

There are a variety of resources an agent must utilize when handling an interaction.  There may be multiple back-office applications or databases, one or more knowledge bases, and the CRM or Trouble-ticket application. There may be hard or softphone controls to start a recording, place a call on hold, or transfer a call.   There may be a chat or SMS interface, an eMail system or fax viewer the agent needs to utilize.

It is easy to visualize how all of these resources can contribute to chaos on the agent desktop (or multiple desktops; I have seen to many instances of this), with multiple signons, toggling back and forth between resources, and copy/cut and paste of data.  A unified desktop that brings order to the chaos is very effective at reducing effort.  Effort is reduced with a single signon, mash-up screens within the unified desktop that bring together data from multiple back-office applications and databases and presents in a way that supports the interaction.  All resources, available in an easy-to-understand single desktop can dramatically lower the effort an agent needs to spend to handle an interaction.  As an aside, a unified desktop can also cut your training bill in half as agents have a single interface and a single application to learn.

Another key component to effort reduction is having a well-structured and robust knowledge base, a single knowledge base that can return search results within the context of the interaction.  For example, if the subject of the interaction is billing, a knowledge search for ‘bill’ should not be returning records from your corporate phone directory of people named Bill.  Likewise, a search for “bill” for a call for a company contact should not be returning records for an overdue bill.

I said that the principal measure of efficiency used today is time-based.  This is not entirely correct.  Most QA checklists or scorecards I have seen incorporate some measure of the agent’s familiarity with and efficient use of their resources.  Unfortunately, the measure is less objective than Handle Time and therefore requires frequent QA calibration to portray a level of accuracy.  It is also resource-intensive to measure as it requires a QA person to listen to the recorded interaction and watch the screen-shots.  Agent-behavior can have a significant impact on time-based measurements.  How adept the agent is at listening and controlling the conversation, how adept the agent is at portraying empathy and how long it takes the agent to establish engagement all drive the talk time for a given call.

One approach I have used with some success is to have all agents handle eMail (offline interactions) in addition to phone calls or chats (online interactions).   Comparing the handle time of the same type of phone and eMail interactions will give you a better indication of whether the agent is struggling with the resources or with the conversation.

By implementing a unified desktop, a robust and context-sensitive knowledge application, and by understanding and adjusting agent behavior, you can easily get a 10-15% reduction in operating costs.

So, let’s here it!  What other ways have you found to increase efficiency?

Post Navigation