Case study: how a high technology manufacturer used process contribution analysis

This company, a market leader in high-tech components, had an enviable reputation for its order entry, warehouse and despatch functions. A customer was answered instantly, a highly disciplined order entry process captured customer requirements quickly, picking lists were despatched electronically to a sophisticated automated warehouse and goods were despatched within one to two hours, typically for next day delivery.

Stock control systems ensured high levels of order fill and line fill. By contrast, its customer help desk, designed to answer queries ranging from simple product and stock enquiries through technical specification questions to advice on component application, suffered long answer hold times, even longer call transfer times and an unacceptably high call abandon rate.

Call agents worked with a relatively simple call centre system which enabled them to query a basic abstract of product data and allowed “interesting” calls to be logged onto a data file so that repeat queries about the same product issue or problem did not have to be researched again.

All other documentation — catalogues, technical manuals, data sheets — were held on paper, either on the agent’s desk or in a central library. Agents had no access to the main business systems and data. There was no access to customer data.

The business issues arising from this were:

1. Customers experienced disappointing levels of service from the help desk compared with the order line.
2. Customers might use the service to gain product advice and then, forced to place another call to the order line, placed a call to an alternative and, possibly, cheaper supplier.
3. If the customer wanted copies of technical documentation, they were despatched manually by a photocopying or reprint service from the central library.
4. If call queues were long — as indicated on a wall board — agents felt pressured into a “don’t know” or “no we don’t” response because searching through available material was too time consuming.
5. Details of customer requests were not logged so a “no we don’t” response which might relate to a genuine need for a new product could not be tracked.
6. When calls were transferred to a second tier agent, the query had to be repeated.
7. Front desk (first-tier agents) lacked support systems and processes. As a result, they spent too much time answering a question without realising that they were getting out of their depth. A call transfer to a specialist was avoided because the transfer time was unacceptably long.

Customer value analysis revealed the following:

1. Customer expectation of service levels were frequently not met, causing potential defections.
2. Telephone response times and wait times were long.
3. Accuracy of requested information was suspect — “no we don’t” might be a pressurised guess.
4. Literature fulfilment did not meet the high standards set by the order line.
5. Customer needs and wants — in terms of product specification, availability and demand — could not be tracked, addressed and analyzed, to the detriment of product marketing.
6. A customer was not identified and there was no means of tracking repeat issues.
7. Return calls to customers for responses to detailed research could not be managed centrally.
8. Competitors were offering the same product at a cheaper price, encouraging defection.

So the basic needs included:

1. Achieve decent standards of telephone response.
2. Give the right answer to basic questions quickly and easily.
3. Prompt and complete despatch of literature.
4. Once the help desk had identified the product, it had to take the order or at least transfer the customer to the order desk with product detail transfer.

The attractors from the customers’ perspective included:

1. Capture my requirements — and if they can’t be handled immediately, make it obvious that they are being dealt with.
2. If I don’t use the precise terms for a product as listed in the catalogue, then I need to speak to someone who can resolve this difference and relate my needs to your product portfolio.
3. If you need to transfer me to someone else, transfer the substance of the conversation so far without me having to repeat it.
4. If my product issue is common to others, give me the impression that you are on top of it as a potential problem and have worked out a fix.

This led to a number of capabilities that must be provided:

1. Better and more comprehensive availability of information to agents.
2. Improved search techniques.
3. Automated literature despatch.
4. Order taking.

There were also a number of capabilities that would provide significant competitive differentiation:

1. A work control system to ensure fast and accurate follow-up to unresolved issues.
2. Automated capture of customer identity.
3. Customer call history.
4. Fuzzy search, “sounds like” capability.
5. Call with data transfer.
6. Feedback of customer requirements into business and product planning.

These capabilities were prioritised according to analysis of customer comments and feedback and led to the development of a system with the following enablers:

1. A new call handling system (software) that linked to a more detailed product database, and connected directly to the main business systems — in this case, through extract of product and customer data.
2. Conversion of paper documentation into an image system for concurrent display on multiple agent screens.
3. Extension to the databases to enable automated cross- reference of product numbers to technical data.
4. Accumulation of products discussed, customer requirements, technical data discussed to permit seamless transfer to other agents.
5. Capture of product data and direct linkages to order entry systems, stock availability without rekeying.
6. Creation of customer history and contact history with call results, call outcomes, notes and follow-up action system.
7. New structure for product naming and categorisation together with advanced searching algorithms.
8. CTI for capture of calling line identities.
9. Automatic post or fax despatch of technical data.

At the same time, processes were analyzed. At the first stage of the project, results showed that:

1. Product searching speed needed to be improved.
2. Agent skill levels needed to be more precisely matched to the complexity of the call, optimising staff utilisation.
3. Skill levels had to be matched to the agents’ ability and authority to give complex technical advice, optimising quality and accuracy of the advice.
4. Call wrap-up times needed to be improved through automatic literature fulfilment and capture of call details during the call.
5. Linked, direct, order taking needed to more efficient than simple call transfer to the specialist order line.
6. Time on a call could usefully be extended in order to capture customer feedback and specific information — for re-use in downstream processes. For example, product development, product quality issues for supplier management, safety issues.

The first wave of findings led to a number of new capabilities or previously identified capabilities that needed higher priority in the implementation plan:

1. Searching algorithms needed to be driven by different kinds of keyword depending on the level of agent experience and technical knowledge.
2. Agent skills, experience and authority needed to be defined more accurately and processes defined to match these criteria.
3. Complex documents were to be withheld from lower skilled agents with a prompt to refer to higher levels.
4. The tiers of staff deployment needed to be redefined to match experience and capability.
5. Application enhancements to the basic call agent software must be built to allow faultless fulfilment and automatic capture of call details.
6. A second set of application enhancements were need to provide seamless access to business systems.
7. Call scripting guidance needed to ensure opportunity for extended capture of customer feedback and automatic capture of basic call details.
8. Better control of call backs and scheduling of research work by senior staff.

Enablers arising from this new and revised set of capabilities included:

1. Departmental reorganisation and re-profiling of skills.
2. Extensions to the image system library database to define document classification to correspond to agent skill and software to control it.
3. Search algorithms further extended.
4. Automatic print and fax subsystem introduced.
5. Screen scraping of host computer sessions, automatic menu navigation and application-to-application data transfer through dynamic data exchange (DDE).
6. Call scripting.
7. Workflow.

The solution design process, using a technique to integrate and customise a number of software package solutions, was driven by this prioritised set of capabilities and enablers. The system was prototyped — to test user acceptance — and delivered in three phases to accommodate basic needs and attractors.

The system was now capable of delivering service levels comparable with the order line. The order line achieved high service levels in part through actively discouraging agents to offer product advice. The customer needed to have consulted the catalogue and have stock numbers and quantities prepared in advance.

The new capabilities of the help desk — sophisticated product search, comprehensive availability of product information and automated links to order-taking — encouraged a degree of laziness in customers who could call the help desk, use the services of the agents to perform catalogue search and product selection and place an order, all through the same call. This distorted agent productivity figures and incurred costs through the wrong application of resources.

Process analysis identified the need to adjust the help desk processes to minimise this excess cost without removing a service which clearly provided a significant competitive differentiator for customers.

The solution was to introduce a variable delay in the hold time on inbound help desk calls to provide sufficient discouragement to those customers who had discovered a way of “defeating” the system. The delay ensured a sufficient incentive to encourage desired customer behaviour.

For further information and to purchase contact Colin Coulson-Thomas