MULTI-SKILLING
 

How many things can they do?

Is multi-skilling the key to call centre success?  Nick Lane, of AOL Europe, looks at the benefits and drawbacks

MULTI-SKILLING is increasingly being adopted by call centre managers, under pressure to get every last bit of performance from their staff and equipment.  They aim to economise wherever possible while maintaining quality of service.
The aim of multi-skilling is to create agents who, in a single shift, can:

• handle emails and web-based contacts alongside voice calls, and

• handle different types of call, from technical assistance and service calls to sales, debt collection and customer registration.

Successful multi-skilling can deliver both greater workforce flexibility and low staffing costs.  But will it work for you?
There are three areas in which multi-skilling is typically used:

1  Call overflow in which call types are routed to one or more destinations.
For example, calls may be routed to single-skilled agents in main business hours, but sent to multi-skilled agents in the evenings -- saving on single-skilled agents outside peak times.
Call overflow can also be used to cut the number of abandoned calls and stay within your service goal (SG).
When it comes to planning call overflow, using workforce management software, your main concern must be to evaluate the impact on agent requirements.  Clearly, in my night shift example, overflow will need to be calculated as the basis for staffing requirements for the multi-skilled team.  However, simply because the ACD is configured for overflow calls does not mean necessarily that overflow has to be included in the planning.
Take another example.  Say a group of Silver membership agents receive overflow from the Gold membership queue if Gold agents are busy.  If in one hour, it takes six agents to handle all the 100 Silver calls, then anything less than 20 calls overflowing from Gold membership would not alter the agent requirement (and thus the SG).  Therefore, in such an instance, it may not be necessary or profitable to include call overflow in the planning process.

2 Contact blending is commonly used by workforce management vendors when introducing basic multi-skilling.  The idea is to split an agent's schedule amongst different activities -- in this example, inbound hotline, fax and email work:

Planners will typically schedule for inbound first and then manually schedule other activities based on staffing requirements and the SG.  A workforce management product will automate this, based on the SG and other defined parameters.  This type of "activity blending" was the first step for many call centres on the path to skill-based routing and multi-skilling.
While this multi-skilling concept can be performed using any sound workforce management system, it has a number of disadvantages. For example:

• Where managers plan to meet their SGs for inbound calls first and then plan other activities "on top" (according to overstaffing), contact blending will do little to account for non-phone activities that have a different set of SGs.  In other cases, it will only be possible to create a total staffing requirement (i.e. with no intra-day distribution) for non-phone activities.  This again will cause problems with SG targets.

• In other instances, managers will find it difficult to react quickly to changes in volumes.  Say an agent is planned to perform an activity for a set period.  If that activity has lower than expected volumes, it is not always possible to seamlessly provide the agent with another activity as he/she may be at another desk, logged into another programme, etc.

"Caller, ring back, please -- I
was just enjoying a headache!"

3  Simultaneous activities  The aim here is to use multi-skilling to integrate the different activities agents can perform.  Contacts are evaluated on their relative importance (SG) and then distributed to agents with the appropriate skill sets required to handle them.  In this way, agents can receive and process incoming contacts in potentially any medium for which they are qualified (email, fax, inbound, web chat).
Advances in technology, such as the Aspect Enterprise Contact Server, enable multimedia contacts to be simultaneously distributed to facilitate this form of multi-skilling.
Progress in customer relationship management software also supports this method of skills-based working -- resulting in true multimedia customer service centres.
This "integrated" method of multi-skill has several key advantages:

• requirements for different contact media can be scheduled simultaneously according to their relative importance (SG)

• any changes in volume can be seamlessly reacted upon (e.g. with fewer emails, agents will simply spend more time on other contacts (such as inbound calls); and

• call centres can gain the full benefit from economies of scale.

Benefits and drawbacks  Clearly, there are good reasons why you may wish to implement multi-skill.  In short, it can:
1 Provide greater agent accessibility (with calls being routed to a greater number of agents)

2 Utilise all available agent skills

3 Provide a facility for further learning and utilisation of new skills

4 Provide flexibility in the handling of calls (e.g. enabling calls to be routed according to business factors such as opening hours, contracts, etc.).

5 Enable centres to handle an identical workload as with single-skilled agents, but with fewer agents -- and still reach desired SGs

Unfortunately, it's rare that all of these benefits can be realised.  Take the technology for multimedia routing of customer contacts.  While powerful contact distribution technologies are now available, they will not always come at the right price and deliver a sufficiently rapid return on investment (RoI) for organisations of all shapes, sizes and requirements.
Then there are planning issues.  Put simply, attempting to utilise ALL agent skills can create a workforce planning nightmare.  For example, agents may have unique skill sets, but varying competency within each skill.  Multiply this by foreign languages (common overseas) and the number of planning possibilities for each agent could become alarming.
There is a very fine balance between increasing agents' skills and overloading them with information.  For example, in many call centres absenteeism has risen when multi-skill routing was introduced.
If the tools and technology to assist agents are not well advanced, the consequence can simply be more work -- with less capability for the agents.
 

There is a very fine balance
between increasing agents'
skills and overloading them
with information

Increased agent occupancy  Another issue is whether multi-skilling is going to have a detrimental effect on the service goal.  Multi-skilling may well create significant personnel (and hence cost) savings but there is generally a trade-off for requiring fewer agents in a complete multi-skilled environment.  There are two major detrimental impacts to consider:

• Agent occupancy  In many call centres it is necessary to put a ceiling on maximum agent occupancy.  This may be to reduce agent attrition or because of a rule dictated by either a government or union.
  In some regions, such as in Central Europe, union rules are commonplace.  It is often stated, for example, that an agent should have a break of five-15 minutes each hour.  Setting a ceiling for occupancy will impact on the number of agents required.  It may potentially cut multi-skilled personnel savings to be reduced or even result in no savings at all.

• Agent retention  The higher the agent occupancy the greater the chance of agent turnover.  Agent attrition will typically lead to higher costs (in terms of both additional training agents and the need to provide incentives for existing agents to remain) and can also hit service levels.

"I hate it when I lose a multi-skilled
agent, but I hate it even more when
I lose all their knowledge"

Workforce management for multi-skill planning  One of the most decisive factors dictating the success or failure of multi-skilling is proper planning.  It often requires more care than the planning of single-skilled agents.  Whichever methods you use (see above), the issues remain the same.   The main factors to consider are:

• ACD routing  Some workforce management vendors have developed solutions that attempt to "mimic" and simulate skill-based routing performed by the ACD.
  While this may have been viable in the past, newly-developed technologies have made this far more difficult.  Imagine a situation where each agent has their own unique skill combination, composed of different telephony and multimedia skills, priorities and skill levels.  Combine this with business rules, agent availability and ACD simulation.  Now you've created a big problem.
  Simulating the ACD is, in reality, not always necessary.  With intelligent routing, planning can dictate where calls will be answered.
  Take a simple multi-skilled routing set-up for handling incoming sales and service calls which is configured so that sales calls are routed to sales agents first and then as a second priority to multi-skilled agents.  Service calls are routed in the same fashion.
  By staffing correctly, managers can dictate where calls are answered.  For example, by reducing the number of sales agents, sales calls can be flowed to cross-trained agents, as the availability of sales agents is reduced.
  One of the main benefits of planning in this fashion is the possibility of providing multiple solutions to a problem.  For example, if sales service levels are poor, there are three potential solutions:

1 Increase the number of sales agents (thereby increasing service levels)
2 Increase the number of cross-trained agents (thus increasing your ability to answer overflow calls within service levels)
3 Increase the number of service agents resulting in more service calls being answered by service agents.  (Consequently cross-trained agents can take more sales calls, thus increasing the service level)

Even a combination of all three is possible.  The flexibility of multi-skill together with intelligent routing enables planners to make the best use of available resources.

Wages  This is commonly recognised as an issue in countries such as America while it's frequently overlooked here.  Multi-skill agents are often more expensive than those with a single skill and are often those with longer service.  Points to consider:

• how much do multi-skill agents cost vs. their efficiency?  Is it cheaper to schedule more single-skilled agents than multi-skill and what would the impact be on efficiency?; and

• if multi-skill agents earn more money, then many take the view that they should be fully utilised.  In other words, calls should be offered to multi-skill agents before being offered to single-skilled agents.

One possible problem with this approach is turnover.  If a multi-skill agent is being over-utilised and leaves, then their knowledge goes with them.  Training costs are often high enough to make this approach prohibitive.
In conclusion, implementing multi-skill must be very carefully thought through from both a business and resource planning perspective.  The availability of the technology does not necessarily provide a suitable reason for employing multi-skill.
Furthermore the different types of multi-skill utilisation, together with their benefits and drawbacks, should be considered to provide the maximum opportunity for reaping the full benefits.

PROFILE
Nick Lane is senior business planning manager, AOL Europe (UK), and is responsible for planning more than 1,500 agents across seven sites.
In 1999, Mr Lane was one-and-a-half years into studying for a PhD when he took a temporary job with the workforce management software company, TCS -- and never looked back.
He joined Aspect Communications in 2000 in a support role, and quickly moved into consultancy and a specialisation in multi-site and multi-skill planning.  In April, 2002, Mr Lane joined Citibank as resource planning manager looking after three sites in Germany.
He then joined AOL Europe in September 2002 as resource planning Manager, responsible for resource planning across three countries and 16 sites.  When AOL decentralised in August 2003, he was promoted to his present post.

Fast facts: If you wish to contact Mr Lane, please click on Reader Enquiry and enter  No. 65/31