I just finished reading Blake Landau’s blog (at customermanagmentIQ.com). The blog is titled An “AHA!” Moment About AHT Average Handle Time and talks about how agents may “hang up” on customers to keep their handle time down to meet their target or quota. I have seen such foolishness in almost every call center I have visited.
The Problem with AHT:
The problem here is deep, rooted in command and control thinking born from scientific management theory (F.W. Taylor). Not only is this thinking old, but is displayed both in the podcast interview and comments on Ms. Landau’s blog. This productivity mindset is over 100 years old and has run it’s course. Better methods are at hand, but require a change of thinking from command and control to a systems thinking one.
The AHT target is the problem. The AHT becomes the de facto purpose of the call center agent meaning their focus is on the target and not the customer. The agent is left with a choice to either serve the customer or risk being paid attention to or not receiving some incentive for not achieving some arbitrary numerical goal (target).
Additionally, the target does not account for the variety of demand that an agent receives. I have seen on many occasions where the customer demand is a hard call (time consuming) and no agent wants those calls when they are under the gun of an arbitrary target. Sometimes they hang up and some times they don’t give complete answers to customers leading to more failure demand (call backs, errors, follow-ups, escalations, etc.), this just increases call volume at great expense.
The Command and Control Solutions:
One comment to the blog suggests that having someone with greater than 15% AHT need to have the agent paid attention to. The arbitrary 15% bothers me where does that number come from? Why isn’t it 20% or 7% or some other number. This person clearly does not understand variation (see Service Metrics: What You Need to Understand).
Almost all the responses were from command and control thinking. Items like more quality monitoring, scorecards, coaching, training, etc. that only add waste to a poorly designed system. Most of these solutions focus on the individual (except scorecards) and the problem here is that 95% of performance comes from the system (work design, technology, management thinking, constraints, regulations, policies, procedures, scripts, etc.) and only %5 is attributable to the individual. Scorecards are just doing the wrong thing, righter (see: Balanced Scorecard . . . MBO in Sheep’s Clothing). These solutions have the displeasing odor of command and control thinking.
A Better Way: Systems ThinkingOne thing I have found is that command and control thinking doesn’t work very well. Systems thinking (by nature) focuses on the customer. Decisions are made outside-in and not top-down starting with understanding purpose from a customer perspective, deriving measures from this purpose and liberating method. The focus becomes serving the customer rather than some arbitrary target. With an understanding of customer demand, we can design systems against this demand. In a management paradox, this improves service and cut costs by eliminating failure demand. This is something that command and control (production) thinkers don’t understand . . . to them there is always a trade-of between costs and good service.
The better way eliminates the need for quality monitoring, scripts, mandates, procedures, targets and the like saving organizations from wasteful costs. Other benefits are improved culture from putting the decision-making back with the work and allowing agents to think again instead of “dumbing them down” with costly technology and monitoring. The real question is . . . are you ready to change thinking to get the benefits?
Tripp Babbitt is a speaker, blogger and consultant to service industry (private and public). His organization helps executives find a better way to make the work work. Download free from www.newsystemsthinking.com “Understanding Your Organization as a System” and gain knowledge of systems thinking or contact us about our intervention services at [email protected]. Reach him on Twitter at www.twitter.com/TriBabbitt.Share This:
thank you for the discussion Tripp. I think it’s interesting the way you write about the added benefits of an improved culture specifically the antithesis of “dumbing them [the agents] down’ with costly technology and monitoring.” I think this is a good debate that we should focus on. Obviously there are A LOT of opinions on this topic…
I’m the person that does not seem to understand variation. Yes I’m not an expert in statistics, yet I do know, from experience and analysis that: – There is a clear correlation between AHT variance (or breadth) and Customer Satisfaction / First Contact Resolution, proving that: – If somebody is way out of scale on the average this is an indicator for poor knowledge, quality and abuse (latter in cases out of scale below of average). This indication also proved to be right when quality monitoring these agents. – Use your (limited) coaching & training resources to focus improvement of knowledge & quality on these “outliers” and you will see fastest and highest increasing results on the metrics that do matter for the Customer such as FCR – As a bonus you get overall AHT to drop, because the agent is now better shaped and equiped to deal with customer requests. 2nd argument: – Call center scheduling depends on 2 factors: volume and AHT – WFM tools work with averages too (as well as some variations on the average) – having a small bandwith of individual AHT around the overall AHT helps to get staffing requirements right and also helps managing ASA better (how many times have you heard a traffic manager swear because agents take too long in wrapp-up with a high waiting queue). Both AHT and ASA are not the most important metrics when it comes to Customer Satisfaction. AHT analysis provides you with insights in your agent-population quality delivery (if you do not use AHT as a agent target, which you should not). I say: use that power. Last but not least: I fully agree: the AHT target is the problem. From where I stand: the size does not matter (or should not matter), too high variation does. Whether this is 15 % above avg, I don’t really care. From my experience I know that most contact centers where quality is an issue (low FCR, poor CSAT, high employee attrition etc) there are significant numbers of agents (up to 30 % of the agents) that have AHT > 15 % of average and below (also up to 30 % of the agents). Limited resources alow for so much focus you can handle. Please read my full article http://tinyurl.com/ohnsw8 By the way:I am fully in favour of system thinking, customer expection BPM, lean etc.. Anything that drives companies to improve the customer experience, takes away customer frustrations etc. I do not think that one needs to go without the other though. It is not “or” it is “and”. No system / metric / measurement / approach is superior. It is about crafting a combination of tools & methodologies that works for your company in the specific situation that you are in at the time that you are at.. If times change (and we know they do) so should your tools / metrics and methodologies. I’ll be looking around your blog sometime soon. The topics you write about are very interesting. With highest regards, Wim Rampen
Wim: Thank you for your comment. I did read your blog before I responded. There is no doubt outliers exist in any call center. The problem is how we derive outliers and who we blame. For Outliers: First of all, we need to account for the variety of demand call centers get. Did the CSR (Call Center representative) get all the “hard” calls that take longer to handle? If so, is it in my best interest as an agent to get rid of the “hard” calls (hang up, cut short the answer, etc.) to achieve an arbitrary AHT number? I have and am sure you have seen this behavior this is not good service. Additionally, this drives a call back from a angry customer (failure demand) that having been dealt with correctly the first time would not have additional calls and time. Something production (command and control) thinkers don’t understand. Secondly, if we are dealing with the same demand and someone is an outlier, we better have the statistics to back it up. It was Walter Shewhart and W. Edwards Deming that gave us the control chart in how to determine “outliers” statistically (called special causes by Deming). There is a statistical way to determine if there is a special cause that I link to in my blog. Arbitrary percentages over or under will not suffice. Because I typically find statistically that agents are not special causes and they are within statistical limits of performance, I know most agents don’t need paid attention to with re-training, coaching, quality monitoring, performance appraisals and other waste. The answer is to improve the system in which they work. This means redesign of the system in which agents work in to improve performance. Deming taught me to improve performance that we had to change the system, it was John Seddon that taught me that to change the system we have to change the thinking. Ultimately, improving performance means improving thinking.95% of the time the system needs to be improved, 5% of the time it is the individual. Thirdly, we have to be careful of correlations. The story of the rooster crowing each morning and the belief that his crowing caused the sun to come up . . . until the day he overslept and the sun still came up. Sure customers would like to be on the phone for a shorter time, but their real purpose is to get the demand purposefully attended to (whatever it may be and however long it takes). Who to blame: If we understand variation and the theories behind it we would be spending more time fixing the system (technology, work design, managment thinking, regulations, scripts, procedures, measures, theory, etc.) all play a major role (95%) in the performance and I haven’t even talked about the rest of the system beyond the call center that drives the customer demand (both value and failure). Managment owns the system, they are to blame. I am sorry, but I can’t subscribe to the statement no system, metric, measurement, approach is superior. I was fortunate to have met Dr. Deming early in my career to understand variation. I am cursed by the fact that thinking has not improved in the 20 years since. There is a better way and hopefully is one beyond this. The productivity (command and control) mindset is old and has run it’s course, long proven is a better way. People need to know that there is. Thanks, Tripp
Tripp, I would really like to understand how to determine outliers better through statistics. If you can point me in the direction that would be great. Now with regard to your statement: The problem is [how we derive outliers and] who we blame. Ultimately, improving performance means improving thinking.95% of the time the system needs to be improved, 5% of the time it is the individual. I do fully agree that it is hardly ever the CSR to blame. That’s also why I state that AHT is a metric for “management evaluation”. The system is created by Management and it is the management who needs to change it or make it work like it is supposed to. Juts for clarification: with The System you are not only talking about processes and IT-systems to guide that procesesses? The system is the entire company and the way it deals with its challenges (including culture, HR-management, procedures, decision making methodology, knowledge management etc). I agree that it’s 95 % the system (created by management) that drives performance, not the individual. When the CSR cannot perform (on quality / service experience etc) it is management who needs to change the system or have the system work better for that individual. It’s not the CSR that “did it wrong”, it was management who did not train or coach well enough and did not provide the right knowledge-tools, processes and procedures to facilitate a great customer service experience. One closing “reality check” remark: if we can all blame The System, it releaves us from the duty to blame ourselves. If you want to change The System, your daily work in it, your role in it, yourself is a good place to start. Thx Wim http://contactcenterintelligence.wordpress.com/
Wim: Thank you for your comment. Sources for the understanding of variation: Within two weeks I will be releasing for sale “Creating and Using Capability Charts” targeted for service organizations. In the meantime, you can read the blog http://blog.newsystemsthinking.com/blog/bryce-harrison/0/0/service-metrics-what-you-need-to-understand or read Understanding Variation by Dr. Don Wheeler (more for manufacturing, but a good read). I am glad that we agree on the 95/5 rule as this is a big area that is missed. Too much focus on the individual and not enough on the system. My original list of system components was not meant to be comprehensive and I am appreciative of your additions. I will depart from you on the actions where you say “it was management who did not train or coach well enough and did not provide the right knowledge-tools, processes and procedures to facilitate a great customer service experience.” The knowledge tools, processes and procedures do not make the customer experience, it is the change in thinking to one of understanding purpose, customer demand, flow, and value. The items you mention are where command and control managers go to improve. With regards to your closing remark as we can just blame the system. I would point back to the understanding of variation to know when it is the system and when it is individual. I would also point out that most command and control systems are designed to make CSRs accountable, but the decision-making has been separated from the work. If you can’t make decisions about your own work, how can you be accountable to a bad system. The systems thinking I aspire to puts the decision-making back with the worker (CSR) and not management making decisions off of some report or other anecdotal information. Thank you again for the dialogue.
hi . thx for your interesting discussions/comments. Recently my manager has asked me to decrease my AHT because i had like 111 being the highest. we are a call centre within a big company and provide specific service. We do both inbound and outbound calls.. i would like to know how AHT is calculated and whether outbound calls are also counted for calculating AHT. I am very much interested in decreasing my AHT although i firmly believe that this will affect the way in providing excellence customer care which is the motto of the company. I am among one of the workers who always strive hard to resolve issues and make it a point to achieve the goal .. However if i need to choose between decreasing AHT or getting reprimanded. I will rather decrease AHT … although it sadden me a lot that performance is based on AHT as i dont see my colleagues resolving issues and escalating calls or even properly documenting the call as i do.. Besides there are variance as i work over the weekends and the type of calls i receive are different. There are lots of systems failure.. pc responding slowly, at times we are lose our system…, information not always updated, unable to issue codes through the sytems… unable to close the call as the system does not recognise the code… and we also act as helpdesk during weekends. I also believe that some of my colleagues just brush off to close the calls by delivering the minimum of what they should be doing. Although i am frustrated i am willing to Decreased my AHT, please feel free to email me if u need further information or i will check for ur valuable comments here… thank you
Pallavie: Thanks for the comment. You don’t mention your industry, so it is hard for me to understand the ttypes of demand you receive. The first thing a manager should look at is the variety of demand by type and frequency. You mention there is variety in demand, so some calls will naturally be longer or shorter based on this demand. Variation in phone calls is to be expected. I do believe that your manager is working on the wrong problem (AHT) and instead first working on the type and frequencies I stated earlier. You might try identifying with the manager the amount of failure demand (unwanted demand from customers like chase calls, missed appointments, etc.) that the call center receives, you will probably run between 25 – 75% failure demand and reducing this number will address getting less calls and therefore happier customers and more importantly (to you) more room to provide good service rather than time sensitive service. All the types of demands you are mentioning make me believe you will have no problem identifying with the manager failure demand. Also, all this documenting is waste what is the company doing with the information? There is another broader problem you face. The whole productivity mindset of the manager and trying to reduce costs through AHT. I suspect you are in the UK from your writing and would suggest that you try to attend one of my Vanguard partner’s events that explain why command and control thinking is sub-optimal and what systems thinking can bring to the table. Or you can read the book, Freedom from Command and Control where the first 50 pages or so are devoted to explaining why your managers thinking is flawed and actually increasing costs and failure demand. Please feel free to email me or post other questions. Regards, Tripp
Please consider reading my novel, HANDLE TiME. It is a satire about life as an American call center employee. HANDLE TiME covers many aspects of the agent experience; from getting hired, getting trained, working under the performance metrics (which include AHT, of course); the customer interchanges, the socialization between employees, management-employee relations, etc.. HANDLE TiME is at amazon.com http://tinyurl.com/handletime is the abbreviated address. there is a video book review as well as are several other professional book reviews on the page. If you Google the term ‘handle time, you will see the book cited. If you Google HANDLE TiME by LiNCOLN PARK, you will see scores of references to reviews, the official youtube book trailer, etc. I had absolutely no idea what the worklife of a front-line agent entailed, originally. After taking a job in a center to research the position, I was so astonished by what goes on, I HAD to write a book! My conclusion is that AHT is counterproductive. It terrifies the agents and forces them into revenue-choking actions; and it fosters a the growth of potential legions of pissed-off, dissatisfied customers who are not receiving the First Call Resolution they deserve for their patronage of the larger company. Thanks for your time. Best Regards, LiNCOLN PARK http://lpwrites.com
Thanks for your blog, ..
We are running few processes here in India and serving few Domestic Satellite TV Service Providers here in India.. we stared our AHT Target last year with 8 Mins / Troubleshooting Call but its now at 5 Mins, our client wants us to deliver below 4.5 mins, we are struggling alot here , our main problem is agent attrition, we have try many things but its not moving a second …..what should we do now??
Robin: Hi Robin. The point you make is a good one. How do you reduce AHT? If we think about what a
Ht truely is – it is the AVERAGE handle time across the contract or group. I would firstly encourage you tpo think about AHT as “it is what it is”. Assuming you have a good group of agents, then the AHt is the output of your processes and prcedures and their effectivness. What do your people say are the reasons for the AHT to be where it is? Have you actually asked the front line where they can improve in process? In my contact centre, my team (around 100 FTE) I have a KPI and Process group. This is made up of my BEST performers. They design our processes by getting feedback from those who know best – our front line agents. They also design our KPI’s based on ap rogressive framwork. AHT is the output of your whole bsuiness and how ikt does things. Start by asking people what takes time? Double jack for a week with many different people and understand where the differences are…. Never stop asking your front line where you can improve and enhance process. It is the secret to a hightly efficient operating model.
I know this thread is old but it is one of the few I have found discussing metrics in a more encompassing manner. I am a frontline call center agent, and started thinking about the futility of a single-point goal for AHT and other metrics we use. Obviously such metrics are subject to distortion from system problems and outages as well as external events. Beyond that they do not really tell one anything about how the result is being achieved.
It seems to me that, contrary to common assumptions, outliers both ends of a metric may be engaging in either positive or negative actions, though obviously as one reached the far end the negative is more likely. Short calls may reflect poor attention to cusomer needs or very efficient behavior, long ones may reflect poor knowledge/habits but also could indicate conscientious attention to customer or systemic problems. I know from my own experience that my CAHT is higher than it could be because I often address problems I find within our system, and do not always have the time or ability to switch to another call state to do an “extra” such as sending a correction to knowledge base documentation – yet doing so ultimately improves the center’s performance.
Nothing to me is more absurd than calling in a rep to advise them that CAHT or other metric is beyond the magic number unless one knows why that happpened. If one does not address the cause then the cure is left to chance as well. If one focuses at least on a sample of outliers the potential exists to get a wealth of information from the behavior of those canaries in the coal mine.