Operations is not just about process!

April 30 2011 by ontrac_admin in Metrics and Measurements, Operations Management |1 Comment
OnTrac Operations Assurance Framework

OnTrac Operations Assurance Framework

Imagine the perfect operation! No people! Everything is automated and the outcome is highly predictable. Wouldn’t that be nice? Unfortunately, that is not the scenario in most operations today. Yes, bits and pieces have been automated, and several clever systems and procedures have been put in place. But still, people are the key to running the operation, and there lies the problem.

Management creates a strategy for the company and this is finally broken down into a series of tasks for individuals at various roles and levels. These tasks are parts of smaller processes, that in turn are parts of larger processes. Every individual in the process is counted upon to perform his/her task as per the required schedules. The task itself, when performed, should be of the appropriate quality. In order to gain efficiencies, some of these tasks could be automated. The process could be made more efficient through business process reengineering or through lean/six sigma projects. At times, roles and responsibilities can be changed in conjunction with the new processes (popularly called “transformations”).

Even when all of these are done, we are still left with a set of people who need to do a series of tasks on a sustained basis, failing which the desired results will never be achieved.

Example from Banks: A great example in recent times comes from the banking domain. It wasn’t too long ago, that visiting a bank’s branch was the only way to do most transactions. Among all the transactions, a vast majority were related to withdrawals and deposits. The entire banking system had systems and processes in place to handle customers (who stood in queues) to conduct such transactions. Those of us who remember this era, can relate to several unpleasant experiences when there were delays in such simple transactions. Either the clerk was not available, or took too long to service the customer ahead of you, or was waiting for some input from someone else, before proceeding to service the customer’s request. Lets fast forward to current times where we now use ATMs for withdrawals and deposits. Introduction of ATMs has significantly changed the landscape in terms of utilization of common banking services. It has improved convenience for customers by an order of magnitude. Now that the benefit has cascaded to all customers, we still experience issues. Customers experience several problems with ATMs such as non functional ATMs (due to technical problems, breakdowns, no cash availability etc.). Long queues can be seen in several ATM kiosks. At times, everything is fine, but the ATM is not available because the bank staff is loading cash into the machine (a process that can take 1 to 2 hours).

The glamour of big bang efforts: The act of reengineering processes and/or automating them is definitely glamourous. Those who are involved in such projects (especially those with high impact) end up achieving a high level of visibility in their respective organizations. This means more promotions or better compensations. However, a larger part of the job is to ensure that the new system or process is sustained over a period of time without the service quality suffering. If the bank’s staff is regular in its preventive maintenance schedule, chances of ATM breakdowns will be much fewer. If the time at which the machines are loaded with cash are adhered to (it should typically be done during non peak hours), inconvenience to customers can be minimized. These tasks are completely dependent upon people. Management needs to monitor that these tasks are done as per the routines they were designed for.

Key to sustenance in BPO industry: Taking this scenario to the BPO industry, there are several projects that companies undertake, to improve process efficiency. But once that is done, people need to run the resulting operation efficiently. For this, managers at all levels need to keep an eye on routine tasks that are supposed to be done. More often than not, managers keep an eye on the outcome through metric dashboards (e.g. they keep an eye on productivity, quality, customer satisfaction, sales, turnaround times, attrition, schedule adherence etc.). While these are important, managers also need to keep track of operational routines, the performance of which will result in better outcomes. These are routines that need to be performed by the staff, team leaders/supervisors etc. For example, managers need to keep track of team leader’s routines such as shift briefings, performance monitoring, leave planning, succession planning etc. Monitoring these routines is itself a process and requires the use of tools (such as a simple excel based tracker) or better software that provides a higher level of business intelligence. We at OnTrac are in the process of developing this software that supports our Operations Assurance Framework to help managers track these routines. Write to us at info@iamontrac.com if you want more information on it.

To conclude, the moral of the story is simple. Our premise is that whatever is the desired outcome, it is important to track not just the process outcome itself, but also the tasks that people do, that go as an input to drive the output. What do you think? Do share your thoughts in the comments below.

Strategic Monitoring-The HOW Part!

March 31 2011 by ontrac_admin in Quality |No Comments

Transaction Monitoring

Monitoring is a key activity that determines consistency of output

They say change is the only thing that is constant. Everything changes! And this phenomenon also affects operations. It is relatively easy to set up a green field operation, but to run it reliably in tip top shape for extended periods of time is a very big challenge. Several aspects affect an ongoing operation and one of the most important ones, is human behavior. Even after several rounds of automation, re-engineering and process improvements, most operations still have some aspects of human activity involved in the delivery process. Therein lies the problem. How do we get people to consistently delivery what they are supposed to, for extended periods of time? People lose focus, they get bored, skills can erode, confidence can wane, changes in personal lives can impact professional performance etc.

But then, again, customers really do not care about all these problems. They expect predictability and reliability from the service provider. They have paid for this service and would like to see their money’s worth.

The case of a restaurant:

Lets take the case of a restaurant. Many restaurants start operations with a big fan fare with a focus on excellent food, service, menu items, hygiene, atmosphere etc. They attract customers and generate a good word of mouth. Over time thought, something happens to many of them (not all, but many). A slow and steady degradation of service takes place. Perhaps some of the waiters have become a tad lax, or the kitchen is a bit slower in service. Perhaps there is a broken piece of equipment that has not yet been replaced or, there is some garbage lying behind the restaurant that has not been cleared for a few days. Bottom line is that there will be a degradation of overall service, the service provider (in this case, the restaurant) may not even realize. However, customers start to see it immediately because they have multiple points of reference across two dimensions.

a. There is a significant time gap between their visits to the restaurant. Therefore, they can easily perceive the difference between the current visit and the last.

b. They are able to compare with many other restaurants, and therefore can judge the relative quality of service

On the other hand, the local management of the restaurant may view things differently. For example, it is possible that one of members of the kitchen staff leaves a bag of trash behind the restaurant one day. The management does not view this as a serious incident. But this incident provides other workers the license to do the same and very soon, there is a hugh pile of trash behind the restaurant. Another example is that of cleanliness. It is possible that the automatic dish washer breaks down. Therefore, the management hires extra hands to manually do the dishes. However, the quality of the wash is not the same. The dish washer is never repaired or replaced and the temporary workers become permanent. Each of these actions by itself adds a small delta to the degradation, and the management is not able to see it. But added together, it can have a tremendous impact.

The WHAT of Monitoring:

Coming back to the BPO industry, the same phenomenon occurs with respect to the quality of transactions. Left unchecked or unmonitored, a similar drift will occur eventually resulting in poor quality. Therefore, most BPO operations have quality monitoring processes designed to observe key elements of transactions (or calls) so that there is consistent output from the customer’s perspective. Typically, these processes involve a quality monitoring form with various parameters that define the transaction. A person from the quality monitoring team observes transactions or calls and rates them. This data (from a sample of transactions) is then observed and analyzed to check for deviations. All of these typically address the “WHAT” of the quality monitoring process. But in the battle for better service quality, this is not enough. Further improvements can be done by overlaying HOW monitoring is done.

The HOW of Monitoring:

Strategic monitoring deals with the HOW part of the activity. Monitoring of transactions must take into account two aspects i.e. observations when the work is being done, and observations after the work has been done. Both of these will provide very different sets of insights. The first type of monitoring is called Synchronous Monitoring because the act of observing and the work are done simultaneously. The second type is called ASynchronous Monitoring because the two do not occur together. Further, in voice processes (unlike data processes), there could be two types of synchronous monitoring i.e. one where the associate knows he/she is being observed (side by side) and one where he does not know he/she is being observed (remote live).

Monitoring processes must involve a ratio of different types of observations based on the requirements. Lets take two different examples:

1. Data Processes: Lets say that the productivity level of an associate is not up to the mark. However, the quality of work is quite good. One of the best way’s to check for low productivity is to physically observe (synchronous monitoring) the associate. It is possibly that he/she may not be using the knowledge base correctly, or may be having issues with certain types of transaction etc. So, the proportion of synchronous monitoring should be higher in the mix (of synchronous to asynchronous monitoring).

2. Voice Process: Lets say that the analysis of an associate’s quality scores shows that his quality scores typically go down during the latter part of the shift. This, at first appears to be strange. However, a higher proportion of side by side (synchronous) monitoring could reveal the roadblock that occur due to physical impediments such as a change in lighting during the latter half of the shift, or the arrival of another team that tends to create distractions in the second shift etc. It could also be associate fatigue or some other factor that can be caught only by physical observation.

The strategy therefore involves creating the right mix of the two types of monitoring. Therefore, Team Leaders and Quality Analysts should have a monitoring plan that takes into account the proportion of synchronous or asynchronous monitoring that need to be done for each associate.

The link below provides one such template for monitoring both data as well as voice processes.

Monitoring Template

Service Efficiency

February 28 2011 by ontrac_admin in Uncategorized |1 Comment
Service Efficiency

Money down the drain?

Prior to the advent of the Japanese automobile industry, the focus of the automobile industry in the US was on the production of large cars or gas guzzlers as they were called. Then came the oil shocks in the 70s and fuel efficient cars suddenly became the rage. As economies recovered, larger vehicles came back in vogue, although they were also a lot better engineered. In recent times, with the price of oil crossing $100 several times, fuel efficiency has once again became a focus are for automobile manufacturers.

The learning from this is that whenever a resource becomes scarce, its usage needs to be moderated carefully. Whenever customers feel the pinch of higher fuel prices, they demand fuel-efficient automobiles.

What does this have to do with the BPO industry? Well, we may not be producing cars, but whatever we produce, clients are now demanding that we do it more efficiently. Post recession, clients have restructured themselves to operate more efficiently. Therefore, there is a trickle down effect of this demand for efficiency to the BPO industry. Clients are demanding better pricing or better quality for the same price. As the industry matures and its rate of growth slows down, the companies that will survive are those that can do “more with less”. So what does this mean for the BPO industry?

Focus on Service Efficiency:
Service efficiency eventually refers to the cost of providing service to the client (this cost include all aspects of service delivery, but for the purpose of this article, we will focus on floor operations). In the early days of the industry, service providers focused primarily on business acquisition. The idea was to become as big as possible, in as short time as possible. Therefore, the types of work that they acquired from clients, and the way they managed that work was not the most efficient. There are several examples of third party companies acquiring captives, with the primary objective of “bulking” themselves up. But as times got tougher, they realized that the focus now needs to shift towards operating whatever they have, more efficiently.

Components of Service Efficiency:
From a floor operations perspective, there are three areas that provide an insight into the company’s operations i.e. productivity, quality and inventory carrying costs. Most client-facing metrics will eventually be part of these larger areas. The challenge for a BPO company, is to be able to meet these metrics with lower number of resources, and/or, cheaper resources.

Lets take the case of quality. Many BPO operations (especially in the non-voice segment) are able to meet stringent quality norms set by clients. During the first phase of their evolution, they used expensive resources to get this done. During the second phase, they used cheaper resources, but developed much better training and coaching mechanisms. In the third phase, they used even cheaper resources, but changed the way they monitored quality. Their sample sizes increased and they specifically started to distinguish between internal quality and external quality. External quality, is what the client really sees and measures. Internal quality refers to rate at which errors are identified through internal quality checks, before the external client can discover them. However, the process of improving internal quality is resource intensive, given that it requires experienced people who could have been deployed into core production. The next phase therefore, is to eliminate internal quality checks and measures. This will require an extremely high degree of accuracy right at the point production is completed, eliminating the need for internal quality. So what does one need to do, in order to use trained low cost resources with zero internal quality checks? The answer lies in a grey area combining technology with human behavior.

  1. Deskilling: Deskilling a process to a level where individual components can be automated has become a mantra in the recent past. By deskilling, a complex task can be broken down into simpler activities that can be done by resources that do not require a high degree of skill (this is quite similar to what Frederick Taylor did over 100 years ago at the Bethlehem steel company). The beauty of this process is that it gives companies the ability to further outsource specific pieces of work, to other service provider who focus on providing services either from geographies that have cheaper resources, or who focus on specializing in that type of service, or both. For example, a company that worked in the area of mortgages for a large client, figured out a way to deskill the work being done at their office in Mumbai. As a next step, they figured out a way to outsource a specific component to another company that specializes in documents tagging and that was located in a smaller town. This way, they reduced their cost by about 15%.


  2. Technology: So what does technology have to do here? Without technology, it would be quite difficult to outsource parts of work to be processed. But a bigger benefit is the reduction of cycle time and improvements in quality that could be achieved. In the above example, the company automated the process of looking through the state laws handbooks for restrictions on property sale at various zip codes in the US. This process shaved off a few minutes from the process, thereby giving the company another about 10% of savings through improved productivity. One need not develop expensive software to automate. There are several other examples where simple macros have helped companies reduce cycle times or improve quality.


  3. Behaviour: While there are many people who do not like to hear it, but most work done by people (in any industry) is routine in nature (though there are certain types that are not). Whether it is a taxi driver ferrying a passenger, a software programmer writing code, a BPO associate processing a transaction, work tends to be routine. A well-known business personality once exclaimed that it a CEOs job was routine, because he had been the CEO of several companies and had seen it all. The problem is that human beings are not wired to do routine stuff. We like new food, new clothes, new movies and also new work. Therefore, getting people to stay focused on doing routine jobs, so that physical or emotional roadblocks don’t get in the way is a challenging affair. This is where management comes into the picture. It involves doing a wide range of activities (which are again routine) to ensure that people are happy and productive at their work. More the deskilling, more routine the job. Therefore, it becomes all that much more important for managers to constantly maintain “people connect”.

This trio of Deskilling, Technology and Behaviour is the key to improving service efficiency. Specific implementation of these across various areas will vary, but that tough job is for the managers of the business.

Egypt: BPO Operations in times of Crisis

January 31 2011 by ontrac_admin in Crisis, Operations Management |1 Comment

Note: We originally intended to write about a different topic. However, we felt compelled to write about operations in times of crisis, given the current situation in Egypt.

Every business dreads the day when its operations comes to a standstill due to reasons completely beyond its control. The damage to the business in such circumstances could be enormous. It can even lead to the company ceasing to exist. We are obviously referring to the impact of the current situation in Egypt on its fledgeling BPO industry. Till about a week ago, Egypt was a poster child and an example of the enormous potential that North Africa holds with respect to tapping the BPO industry. Several companies have set up shop at various locations in the country with cities such as Cairo and Alexandria being attractive destinations. Egypt is a surprisingly multilingual country. Many companies provide support in English, French, German, Spanish etc. Not to be left behind, some Indian BPOs have also set up operations in the country. Everything was great, till last week arrived.

The current situation is different from, say, a natural calamity such as an earth quake or a tsunami. Natural calamities do not have politics associated with them. Also, they are immensely “forgivable” by investors due to their very nature. However, the case in Egypt is different. This is a case of civil unrest, that can potentially have a longer impact on the industry at large. It is interesting to observe how some BPO companies tried to keep their operations going inspite of such challenges. Each company seemed to have its tolerance limits.

Many companies are new to the industry. They either have no business continuity plans, or, have very rudimentary ones. At the first sign of trouble, operations were disrupted. We would hazard a guess, that perhaps 80% of companies would fall in this range. We would also guess that most of these companies would be less than 2 years old and therefore have not had enough time to develop mature operations. Having said that, there are others who have leveraged their global expertise. They have quickly invoked strong business continuity procedures and transferred work to other geographical regions including India and Eastern Europe. They also kept sufficient number of people at the office given that there were arrangements for people to stay for a few days within the premises. However, with the failing technical infrastructure that resulted from disruptions in telecom lines and internet connections, companies had to resort to alternate modes of service. An example is that of a process in one company which quickly shifted to fax as a primary mode of support for its customers. Agents were “retooled” within a day to support such queries. Slowly, this too died its death as fax was no longer possible. Interestingly, during this period, the companies that transferred support to other geographies, also made arrangements for technical knowhow to be readily available in a “cloud” and not at servers located in Egypt. This was a great decision because data stored in the cloud was easily accessible to the other geographies. There is also an instance of a company flying out its subject matter experts to an East European destination, where they could provide necessary support to the local agents. Of course, at some point, even companies with such excellent BCP plans will throw up their hands, as is the case now. Sadly, almost no BPO company is able to continue providing services to its clients. We know of companies actively involved in getting their foreign nationals out of the country and most of these employees will never return again. They have had to leave their belongings, barely managing to get back along with their passports.

At this point, it is very difficult to say what will happen both in the short and long terms with respect to the industry. If stability is restored in the country in the next few days, and the technology infrastructure assumes normalcy, this might just be a blip in the long term. However, if this continues, the BPO industry will probably be dead for a long time to come in Egypt.

Better Employee Utilization – The Geese Strategy

December 29 2010 by ontrac_admin in Metrics and Measurements, Operations Management, Utilization |No Comments
A Busy BPO Center

What does it take to keep employees busy?

A major determinant of the viability of any BPO operation is employee utilization. Obviously, when this number is low, it would mean that the operation is saddled with excess capacity costs. Therefore, an important objective of every operation, is to ensure that employee utilization is optimal at all times.

Simple Process with one type of work: The drivers for employee utilization in processes where the staff performs only one type of work, are limited. It primarily depends upon the volume of work, the number of people and their productivity. Its fairly obvious that if the volume of work increases, the utilization rates will increase. If the number of people are more than required, the utilization will decrease. Ironically, if productivity is high, it will tend to decrease utilization. Now lets assume that productivity is at the required level. Lets also assume that the work arrival rates are erratic i.e. it is not really possible to accurately predict how much work will arrive at any point in time. In such a scenario, a vast majority of operations tend to overstaff, to compensate for spikes. This automatically leads to lower utilization. What can be done in this case?

The answer is quite simple. Here is an example:

Pre-Scenario:

No. of inbound transactions/calls in a day (of 8 hours): 7500

Average Handle Time (or Average Processing Time): 10 minutes

Total Workload per day = 7500 x 10/60 = 1250 hours

Assume staff works for 8 hours. In that case, No. of people required = 1250/8 = 156 (approx)

Total number of people in the team on floor = 212

Staff Hours available per day = 212 x 8 = 1696 hours

Utilization = 1250/1696 = 74%

**We are ignoring shrinkage so that we can keep the calculations simple for illustration purposes

Post Scenario:

The Geese Strategy: When geese fly long distances, they do two things. One: They fly in a delta or V formation. As a group, this provides them with a higher lift. Two: They exchange places i.e. the leading geese falls back after some time, and another takes its place. The trailing geese require lesser energy because they get the benefit of the lift provided by the leading geese. This strategy can be used as follows:

Divide the team into three groups. The utilization target for group A will be 90%. For group B, it will be 85% and for group C, it will be 80%. Rewire the ACD (or the transaction routing system) such that all transactions/calls are first routed to group A. Once the target occupancy for group A is achieved, route the overflow volume to group B. When group B achieves its target occupancy, route the remaining volume to group C.

In this strategy, it is also important to rotate staff every couple of hours between the three groups, so that each individual is not exposed to high occupancies all of the time. The numbers then work out as follows:

Group A

Target Utilization: 90%

No. of people: 71

Staff Time available: =71 x 8 hours/day = 568 hours

Total Workload handled: 568/90% = 511.2 hours

Group B

Target Utilization: 85%

no. of people: 71

Staff Time available = 71 x 8 hours/day = 568 hours

Total Workload handled: 568/85% = 482.8 hours

Group C:

* This group only needs to handle 256 hours of work (1250 – 511.2 – 482.8).

Target Utilization: 80%

Workload = 256 hours

Therefore, staff hours = 256 / 80% = 320 hours

Therefore, no. of people required = 320/8 = 40

The total number of people now required  = 71 + 71 + 40 = 182. That is a saving of 30 people, who could be used for other purposes.

With this strategy, people will be busy enough, while at the same time, the erratic flow can easily be handled (there is enough buffer in groups B and C). The utilization target rates here are only indicative and should be adjusted downwards to suit the needs to the business.

Processes with Multiple Tasks: In some processes, the staff handles multiple types of tasks e.g. inbound and outbound calls, or inbound calls and transactions etc. The general thought process is that multiskilling employees allows us to get better utilization. While this is true, we can further improve utilization by using the Geese strategy as follows:

Lets take the an example of an operation where the staff has to perform three types of work i.e. handle inbound calls, make outbound calls and handle inbound transactions. Any employee, at any point in time, could be doing any of these three tasks. The problem is that employee utilization in this process was at a low 73%. Again, the issue here is that it was difficult to predict the arrival of transactions or the inbound calls. However, outbound calling was controllable by the process. Here is the data:

Pre-Scenario:

Inbound:

No. of calls arriving per day: 500

AHT: 3 mins

Total workload =500 x 3/60 = 25 hours

No. of people required = 25/8 = 3 (approx)

Transaction Processing:

No. of transactions in a day: 1000

Average Processing Time: 20 mins

Total workload = 1000 x 20/60 = 333.33 hours

No. of people required = 333.33/8 = 42 (approx)

Outbound Calling:

No. of calls to be made per day: 600

AHT: 5 mins

Total Workload = 600 x 5 /60 = 50 hours

No. of people required = 50/8 = 6 (approx)

Total people required = 3 + 42 + 6 = 51

Actual no. of people on the floor = 70

Total available staff hours = 70 x 8 hours per day = 560

Total workload = 25 + 333.33 + 50 = 408.33 hours

Utilization, therefore = 408.33/560 = 73%

**We are ignoring shrinkage so that we can keep the calculations simple for illustration purposes

Post Scenario:

In this case, we can see that the largest volumes are in the transaction processing work. Therefore, create a group A, that will only handle transactions. Target utilization should be 90%. Group B will consist of people making only outbound calls. Their target occupancy will be 70% and the third group (C) will be only take inbound calls. Group C’s target utilization will be about 65%.  The point of keeping sufficient buffer in group B and C is to take care of excess work that could come to group A. If the utilization goes above 90% in group A, the excess work should be moved to group B and then to group C.

Group A (Transactions only):

No. of people: 47

Staff time available = 47 x 8 hours per day = 376 hours

Total workload = 333.33 hours

Utilization = 333.33/376 = 89% (almost 90%)

Group B (Outbound only):

No. of people: 9

Staff time available = 9 x 8 hours per day = 72 hours

Total workload = 50 hours

Utilization = 50/72 = 695 (almost 70%)

Group C (Inbound only):

No. of people: 5

Staff time available = 5 x 8 hours per day = 40 hours

Total workload = 25 hours

Utilization = 25/40 = 63% (almost 65%)

The total number of people now required  = 47 + 9 + 5 = 61. That is a saving of 9 people, who could be used for other purposes.

With this strategy, people will be busy enough, while at the same time, the erratic flow can easily be handled (there is enough buffer in groups B and C). The utilization target rates here are only indicative and should be adjusted downwards to suit the needs to the business. Also, it is important to rotate people across these three groups every 2 hours so that they are not all exposed to high occupancy rates all the time. In this case, multiskilling is a big benefit because it enables such rotations.

Smart routing strategies (such as the Geese strategy) can contribute significantly to help achieve better employee utilization, while helping to use excess resources for other value added activities

Next Page »