Good customer experience management is not a one-time event, but requires reviewing customer journey maps whenever something in a process has changed. The implementation of the lovely new open-road tolling in Massachusetts missed a key part of the customer experience: how to tell the customer when something is amiss with his transponder. And this could be — and should be — proactively pushed to the customer.
The new TSA service design requires removing large electronic items. The goals are to improve the efficacy of the scans and to speed over passenger processing. However, this may not speed up passenger processing due to increased workload at passenger work areas. This is a great example of shifting a bottleneck, and, unfortunately, the lack of attention to all the “workers” and job tasks in the overall process flow.
TSA lines aren’t as bad as feared, due to the implementation of new process designs and technology. It’s really just back to basics of operational management and Lean philosophy. No earth shattering innovations, just stuff they could have — and should have — done years ago.
We’ve seen that the head of the VA doesn’t understand service delivery models and now we’re hearing the details that show the TSA doesn’t understand basic process analysis. This should be a surprise to anyone with any understanding of capacity analysis – it certainly wasn’t to me – but it is astounding that we have people running these critical agencies that appear to lack such basic knowledge of their agencies’ core tasks. TSA capacity management is just poor.
We all know that the lines at TSA for security checks have gotten long. TSA wants to tell us that’s because of an increase in passenger traffic. And various pandering senators – think Ed Markey – think it’s the checked baggage charges. But now we’ve learned that checked bags have missed their flights — and in large numbers — due to screening holdups.
The Outcome of TSA Capacity Management: Long Lines
The problem is obvious and was well articulated by an American Airlines senior manager testifying before Congress. The TSA didn’t adjust its staffing model when the protocols for screening baggage changed.
Let’s explain. We all remember that TSA failed to detect contraband in smuggling tests in over 90% of attempts. (With such poor efficacy, why even bother to do the screening?) So, TSA responded by increasing the screening time to make sure bad stuff got flagged. Any of us who travel could see that screening time was longer for each given bag. (If I wasn’t afraid of being arrested, I would take some hard measurements to quantify the times.)
Capacity is defined as items screened per hour or day, NOT the number of screening lines. When you lengthen the time to screen a bag, you reduce capacity.
For example, if you double the time to screen, you would need to double the number of screening lines to keep things in balance. TSA didn’t do that. In fact, it appears the number of screening lines decreased.
What did they think would happen? And their solution? Tell people to show up earlier. It’s your fault, passengers, for missing your flights!
TSA Recommendations Make Long Lines Even Longer
But as I have argued, having demand (people) arrive earlier does NOT address a demand/capacity imbalance! Here’s what it does do: It makes lines get even longer. Why?
The recommendation creates a mob mentality. Since other passengers are going to show up earlier, then I need to show even earlier just to be sure I get through.
- Wicked long lines waiting for security screening, which would be a very tempting soft target for terrorists.
- More crowded gate areas with all these people who have arrived wicked early.
As for that new screening system being tried out in Atlanta, is it really attacking the bottleneck, which is the screening time? I’m dubious it’s all that’s being touted. It may be an improvement, but my guess is that its impact will be marginal.
As for Senator Markey, please stop pandering. If you shift more bags to being checked, you’re NOT attacking the bottleneck; we’ll just have more checked bags miss flights. I know you’re a lawyer, but how about taking an operations management course at one of the nearby Boston business schools.
The head of the US Veterans Affairs, Robert McDonald, demonstrated a phenomenal lack of understanding of the nature of a service product and measurements relevant to those service products when he said that VA wait times was not an important measure.
VA Wait Times in McDonald’s View of Service Measurement
“The days to an appointment [wait time] is really not what we should be measuring. What we should be measuring is the veteran’s satisfaction… When you go to Disney, do they measure the number of hours you wait in line? Or what’s important? What’s important is, what’s your satisfaction with the experience?… And what I would like to move to, eventually, is that kind of measure,” as reported by the Christian Science Monitor at one of its events on May 23, 2016.
Really? He thinks Disney doesn’t measure wait times? Of course it does! Why? Because wait time is part of the customer experience, it’s part of the product-service bundle they sell to their customers.
McDonald’s bona fides for the VA job was his military background and that he was CEO of Proctor & Gamble. How can Mr. McDonald run a service operation when he lacks such basic understanding of services? There are no wait times in the production of Tide.
Let’s put the issue into a context that Mr. McDonald would understand. Let’s say customers love the cleaning power of Tide, but the box they hate — no carrying handle or it was flimsy and would break into smithereens if it were dropped. Would Mr. McDonald say, “Yea, but we should focus on the cleaning power of the detergent”? Of course not. Look at the whole package, which customers do.
Service Quality Dimensions
A well-accepted academic model of service quality posits that there are five dimensions to service quality:
- Responsiveness — prompt delivery of service
- Reliability — delivering service with good outcomes
- Assurance — having confidence in the service provider
- Empathy — showing concern for the customer
- Tangibles — leaving the customer with more knowledge of the situation
People evaluate their overall experience based on all five of these factors. The core service could be reliable and of high quality. But if you have to wait a long time for it, or you are treated badly, or you lack confidence in the abilities of the service provider, your opinion of the service will be impaired.
Interestingly, Mr. McDonald’s comments about the lack of importance of responsiveness demonstrates a lack of empathy and could probably reduce the assurance veterans will feel about their VA health services.
Professor David Maister has a highly cited Harvard Business Review article on the “Psychology of Waiting Lines.” He discusses in his book ways that the perception of wait times can be managed, but the actually wait times still need to be managed. Just ask the TSA.
Yes, Mr. McDonald, Disney does measure and manage wait times and wait perception. And they’re only delivering entertainment. In your service world of health care delivery, treating wait times that can affect clinical outcomes as unimportant is truly mind boggling.
TSA (US airport security screeners) frequently tell us during busy times to arrive at the airport extra early. Early arrival does NOT increase capacity. If everyone arrives 2 hours early, then you are no more likely to make your plane than if everyone arrives 1 hour early. Arriving early just creates more congestion, but now TSA can turn the blame to you for missing your flight. Airport capacity management is the issue!! If only journalists would ask the right questions…
The goal of service recovery is to identify customers with issues and then to address those issues to the customers’ satisfaction to promote customer retention. However, service recovery doesn’t just happen. It is a systematic business process that must be designed properly and implemented in an organization. Perhaps more importantly, the organizational culture must be supportive of the central tenant of service recovery strategies — that customers are important and their voice has value.
Summary: Effective service recovery at airlines should be second nature given how much practice they get, but at United, the bromides were plentiful, but meaningful explanations were sorely lacking. This article examines United’s service recovery efforts in the contexts of a service recovery model.
~ ~ ~
Service recovery actions intrigue me. What actions do companies take when an upset customer complains? I’ve written elsewhere about the three principles against which the fairness of service recovery acts should be judged:
- Fair Outcomes: Distributive Justice. The adequacy of any compensation offered for the problem the customer experienced.
- Fair Processes: Procedural Justice. The policies applied and speed with which the issue was handled.
- Fair Interactions: Interactional Justice. How you were treated during the process to resolve the problem.
Stephen Tax and Stephen Brown presented these points in their 1998 Sloan Management Review article, “Recovering and Learning from Service Failure.”
There’s no industry that has more practice in service recovery than the airlines, and I guess this disproves the adage that “practice makes perfect”. As a frequent traveler, I enjoy The Middle Seat column in the WSJ. (Okay, I’m partial to it since Scott McCartney, the author of the column used an idea of mine for a column about how TSA has screwed up its Pre-Check security lines.) A recent Middle Seat column focused on the apologies that airlines provide when customers complain. The gist of the article was:
Airlines say they try to make responses conversational and personal. They aim to apologize and acknowledge the problem, providing more information about the particular situation after research, then offering some compensation as a goodwill gesture, such as some frequent-flier miles.[Note: the Wall Street Journal is a closed site, but if you search on “Trouble Selling Fliers on the Fast Airport Security Line” and“The Art of the Airline Apology”, you may be able to access the articles.]
By pure happenstance I had 2 complaints awaiting response from United, both from a flight in June to deliver a survey workshop in Dubai. And before you get the idea that I’m a constant complainer, these I believe are the first complaints I’ve filed with United. The responses from United border on comical in their hollowness. Let’s look at them against the benchmark of good service recovery principles.
~ ~ ~
My first comment to United was about the brain-dead service design of its brand new service counter in Terminal B at Boston’s Logan airport. My flight to Newark was delayed so that I would miss my overseas connection — not a trivial issue. The physical design of the queues violates basic queuing principles by making it difficult for agents to serve any customers other than the customers in their immediate queue, and the supervisor on duty didn’t compensate for the poor design by shifting workers or customers.
The result was very longer waits for United’s premier customers — and no wait in the other customer queues at the counter. Some agents were standing at their counters doing nothing while a queue built up 25 feet away. A basic service principle is to take special care of your better customers — D’oh — and to build flexibility into the system to accommodate the random arrival pattern of customers. When they designed the physical counter, they should have thought about the implications for wait times the design could cause. Alas, they did not.
But here’s the response I got to my comment pointing the gross shortcomings
Thank you for contacting United Airlines regarding your June 6 experience in Boston.
Our goal is to provide a seamless experience from the moment you book your flight until you arrive safely at your destination. Based on your comments, and the inadequate handling you report we have room to improve. Please be assured we understand your concerns and that your comments have been forwarded to division senior management for internal review and necessary corrective action.
Mr. Van Bennekom, you are a valued MileagePlus Gold member, and we appreciate your business. In recognition of your loyalty, I will gladly add 7,500 goodwill miles to your MileagePlus account. Please visit united.com/MileagePlus to verify your mileage balance in five to seven business days.
We appreciate your loyalty and the offer of your expertise. After all, a great customer deserves a great airline.
A bit over the top, and more importantly, I never got any follow-up from the management at Logan airport about how my concerns were being addressed in their system design. But that response is great compared to the response to my other issue.
On the day of my return flight home from Dubai, I saw that my reservation online was missing a segment — the flight from Dubai to Zurich on Swiss. (The entire itinerary was purchased through United.) So, I confirmed my itinerary with a United phone rep just 8 hours before departure. When I got to the airport to check in, my return itinerary had been cancelled. In my comment to United, I asked in all sincerity how this could happen and what I could do as a customer to make sure it didn’t happen to me in the future.
Here’s the vapid response.
Thank you for contacting United Airlines.
I appreciate your patience and offer my apology for a delayed response to your message.
Please accept our sincerest apologies for the many inconveniences that you endured with your reservation. We truly appreciate you taking the time to point out the issues that you faced as we work hard to correct problems brought to our attention. We understand how disappointing this was for you and we apologize for the factors that contributed to your overall dissatisfaction. Although we can’t explain or undo what you experienced, [my emphasis] we can work to correct it. This situation will be reviewed and addressed internally with our Senior Managers. Your concerns will be taken very seriously.
It is never our intent to inconvenience our most loyal customer, and I am very sorry your valid expectations were not met. Your feedback will help us evaluate what has happened and allow us to make necessary adjustments as it tells us what areas we need to improve on. We appreciate the time you took to share your experience with us and the opportunity to apologize for any inconvenience.
I understand that we can’t undo all that you have experienced but as a gesture of goodwill I will be crediting your account with 7,500 miles. We ask for your patience as it can take up to 14 business days for the miles to be credited to your account.
We at United apologize for the inconvenience and thank you for bringing this to our attention. As a loyal Premier Gold Elite member your business and satisfaction does matter to us.
Interestingly, Scott McCartney’s article says
United said it tries not to go overboard on the apology. “Generally we tell the customer we are sorry they did not have the experience they expected on United,” spokesman Rahsaan Johnson said. “We try to be empathetic to the customer but not sound insincere.”
Apparently, this agent didn’t get the message. I followed up on my key request.
Per my note, I would still like to know *how and why* my reservation was cancelled so that I can take steps as a passenger to make sure it doesn’t happen in the future.
The response to this was an auto respond. How can they not be able to explain how a reservation gets cancelled? It’s a transaction that was initiated by a computer program or an individual action. Just tell me the conditions that led to the action so I can try to identify and compensate for such a mistake in the future.
In fact, I learned that if you reply to an email from United Customer Care, you get this auto respond that has been mail merged with information from your case number.
Thank you again for taking time to let us know about your recent experience with United Airlines.
7500 bonus miles have been added to the MileagePlus account, XXX.
The reference for this item is: YYY
For current information on your MileagePlus balance, as well as information on the latest services available to you as a valued member of MileagePlus, please visit our website at www.united.com.
I did get a response two weeks later from the agent, telling me:
I am still trying to find out the answers to your concerns. Our Tech Support is still looking into the issues.
As I post this update yet another two weeks later, I still have not learned how to become a better customer.
So let’s evaluate these on the 3 requirements of good service recovery.
- Fair Outcomes: Distributive Justice. The compensation was fine. I really didn’t care about any compensation. But what I really wanted was an explanation of the cancellation issue. (Is that Distributive or Interaction? A bit of both.) I try to be a smart customer and smart traveler. I really wanted to know how to avoid having this happen again in the future. Here’s the black eye for United from me. They haven’t given me any explanation whatsoever. Shame on United for that.
- Fair Processes: Procedural Justice. Not an issue here.
- Fair Interactions: Interactional Justice. I came away with the feeling that the replies were perfunctory and my feedback just vanishes into the ether. The replies didn’t instill greater loyalty in me for United. I have a fairly sensitive BS meter from 27 years of teaching, and it was flashing red reading the emails, especially the second one.
I know that United is not unique in hollow service recovery. But what they’ve taught me is to expect little in the way of a substantive reply.
Fewer bromides. More real explanation.
Summary: Employees strike against CEOs all the time, but have you ever heard of employees going on strike in support of a CEO? Employees at the Market Basket supermarket chain in greater Boston have done just that in support of Arthur T. Demoulas, and customers are supporting the boycott. This is an odd event.
~ ~ ~
I live in a state — Massachusetts — known for its quirks. And right now we have a really quirky event transpiring. A local company’s employees are on strike because of the CEO. What’s odd about that? They’re striking (or boycotting) in support of a fired CEO. And they’re not union employees, so they could all be fired today with little or no legal recourse.
The Market Basket supermarket chain has a long history of intra-family squabbles. The company is owned and run by cousins, grandchildren of the founder. Such dissension is not uncommon in family businesses where ownership is split through the generations. In mid July the board of directors, controlled by cousins on one side of the family, dismissed the CEO Arthur T. Demoulas. (Arthur S. Demoulas is a cousin on the other side.)
The reaction of the employee base was immediate and strong. Not just store managers, but rank and file employees stopped showing up for work. Food wasn’t getting moved from the warehouse, so the store shelves have become more and more empty, especially of produce. Employees have been collecting petitions outside the stores for weeks on their own time and held mass rallies at the company’s headquarters.
All this because the CEO was fired.
I should add here that this is my preferred supermarket, especially after they opened a store quite near my house. Before this store opening, I would take advantage of any trip that took me near one of the stores. So, I’m invested in this store surviving, but I’m honoring the employee boycott as is most of the customer base. (Maybe that fact is not so quirky for Massachusetts.) The nearby photo is of a sign outside “my” Market Basket in Hudson, Mass.
Unfortunately, the media has only covered the story that makes good headlines — employees striking for Arthur T, the CEO. The media has not really covered any of the back story. Arthur T.’s side of the family is not exactly wrapped in glory, and we haven’t learned what exactly led to Arthur T.’s dismissal. There are rumors of sweetheart real estate deals by Arthur T. Market Basket is a privately-held company so the financials are not public. The media therefore have to work harder to get all the back story. I guess it’s easier to cover employee rallys then really present the whole story.
But I have talked with the employees outside “my” store. They say that “Arty T.” wasn’t solely focused on profits. He paid employees well and is the driver behind taking lower margins for the products they sell. The “other Arthur” et al. who fired Arty T. want to drive up profit margins by raising prices and probably being less generous with employees.
Of course, maybe Arty’s approach is the real profit maximizing one. Maybe his lower margins have created a hugely loyal customer base, leading to greater volume, and maybe his employee practices mean that customers get a more positive experience, garnering greater loyalty. I recently wrote about a similar story at Hilltop Steakhouse, which went for profits and sunk the company. Maybe if current management gets its way Market Basket will lose its competitive advantage and be less profitable.
The nearby photo is of the entrance to “my” Market Basket. All those slips of paper are register receipts that customers have put up from other supermarkets. Note that store management hasn’t taken them down or driven the employees from the store entrances. Those who run the company in the corporate offices are an island with little to no support from the employee base, including store managers. Corporate management may win the battle, but it may be of a bankrupt company.
As a business school professor, I watch what’s happening in the business community pretty closely. I can never recall a situation quite like this where employees put their jobs on the line like this to support a fired company president or CEO.
I’ve never met Arthur T. Demoulas, but the employees all speak about him in very high terms. He certainly has some leadership skills that are seldom seen in businesses — even if we don’t know the whole story. I guess the good news for him is that should he not get reinstated or successfully buy out his cousins, he could go on the lecture circuit.
~ ~ ~
Update: August 28, 2014
The Market Basket board agreed to sell the company to Arthur T. for $1.5 billion. But this is not the end of the story. How will Arthur T. manage the employees that have backed him so well? He will be hailed as a conquering hero right now. What happens if he has to make a tough decision down the road? At least, I can shop there again. But I’ll wait a week for the likely influx of returning patrons, and I suspect a lot of new shoppers will go there to see what all the hoopla was about.
Update: Early 2015
Chats with cashiers confirm that lots of people who had never shopped at Market Basket came to check it out because of all the news. So, the store is more crowded than before. Drats! I don’t think the whole contentious confrontation what an advertising gimmick, but maybe that’s how it’s worked out.
Leading companies evaluate customer support requirements at the design stage. Do your R&D engineers accept the need for “Design for Supportability”?
During new product development, R&D engineers need to consider a wide range of often-conflicting requirements, including product features, cost, quality and manufacturability. Therefore, it is not surprising that support requirements are often neglected. However, support is an essential factor for achieving customer satisfaction in many industries, and so ensuring that products are easy and economical to support is a priority. Customer support managers know this but often have problems in convincing their R&D colleagues. This article looks at the issues involved in achieving Design for Supportability-the full evaluation of support requirements at the design stage.
Product managers and R&D engineers have a wide range of requirements to consider at the design stage. Typically, these include product features, cost, quality, and manufacturability. Since R&D resources are always limited and fast time-to-market is essential, it is not surprising that customer support is often forgotten during new product development (NPD). However, neglecting support concerns at the design stage is a missed opportunity because ease of support is essential to customer satisfaction in many markets. In addition, support can be a major source of revenue and sustainable competitive advantage. Therefore, it is important that both R&D engineers and product managers fully understand:
- The role of the customer support organization
- How support can lead to long term competitive advantage
- How product design influences support requirements
- How to evaluate support requirements at the design stage, using what are called Design for Supportability (DFS) techniques.
In many companies, manufacturing initiatives have led R&D to think beyond their immediate domain of responsibility-or, to use the popular term, think “outside the box”. For example, concurrent engineering is now common practice and this dictates that manufacturing processes should be developed simultaneously with the design of the product. However, this “box of thinking” ends with product shipment, whereas DFS requires R&D to extend their thinking to cover the lifetime use of the product.
Understanding the Role of Support
Recognition of the importance of support has grown over the last ten years. Support used to be considered as just service and repair, a backwater not worthy of management attention. However, attitudes are now changing for three reasons:
In many markets, support plays a key role in achieving customer satisfaction, and strongly influences repeat purchasing behavior. For example, a key reason why automobile owners do not remain loyal to a brand is because they have received poor after-sales service from dealers.
Customer support is a key source of revenue. In addition, it can be very profitable area-with margins exceeding those made on the products themselves. A McKinsey study concluded that “in most industrial companies, the after-sales business accounts for 10 to 20 percent of revenues and a much larger portion of total contribution margin” .
New technologies have changed many aspects of support. Today’s products are more reliable, and this has reduced the relative importance of maintenance and repair. On the other hand, the complexity of equipment has often increased, particularly if it is software-based. This means that aspects of support such as user training and telephone support are now fundamental to many business models. Consequently, designers need to think how support can be simplified because of the costs involved. For example, support costs in the software industry are typically 5-10% of revenues (not to mention the costs customers incur themselves in support product usage).
Leveraging Support for Competitive Advantage
Many leading companies gain a competitive advantage from support . For example, the Caterpillar Corporation’s support of their earth-moving equipment is legendary-guaranteed delivery of spare parts anywhere in the world within 48 hours and machines that can be repaired quickly and economically. (For more details of their approach to “negative downtime”-predictive software that prevents failures-see .)
Another example is EMC, the leader in data storage. This company places such a high emphasis on giving excellent customer support that it has decided to offer only one level of service contract. Only offering one level of service — and the company aims for this to be premium quality — goes against the standard marketing doctrine of defining a tiered set of service contracts. In fact, EMC’s whole service organization is structured as an “investment center”, with the sole mission of bonding customers to the company for future hardware purchases. Consequently, EMC’s service organization offers innovative approaches, such as an extensive “phone home” capability built into their systems to detect problems before a hard failure .
A third example is VendorC. This is a pseudonym for a market-leading company that designs, manufactures, sells and supports complex vending machines. Vending companies buy large numbers of machines to provide self-service sales of a wide range of goods, some of high value. Modern vending machines — also known as vending terminals — are a complex mix of mechanical, electronic, security and display technologies and a model can cost in the region of $15,000. Due to the large number of mechanical components and their high levels of usage, regular maintenance and repair is required. Vending terminals can now be linked via modems to a central computer, which remotely monitors performance, sales activity and stock levels in chains of vending machines. Using modem links, VendorC offer full goods management to their customers i.e. ensuring that machines are both efficiently maintained and replenished with sales goods in a timely fashion. This incremental service is a new and important source of revenue for VendorC, and it arose from their consideration of support needs at the design stage. VendorC’s management assigns significant resources to support as it generates 35% of sales at margins of typically 25%. For VendorC, customer support is the source of competitive advantage-so much so that the company does not want to attract their competitors’ attention and consequently wants to remain anonymous.
How Product Design Influences Support Requirements
Product design influences both the degree of support necessary and the way it can be delivered. For example, decisions taken at the design stage affect product reliability and consequently the level of maintenance required. Similarly, modular design can reduce repair costs, and troubleshooting is made easier by good diagnostics. In addition to repair and maintenance, design also influences user training and product upgradability. For example, remote support avoids the need for costly on-site visits and on-board features can significantly reduce the amount of training that users require on complex products. However, the impact of design upon support requirements typically does not receive enough attention during new product development.
Four factors inhibit companies from developing products with high supportability i.e. products which are easy and efficient to support. These are:
- Customer support requirements are considered too late in the product development cycle.
- Field support engineers and managers, who know support problems first-hand, have little or no influence over product designers.
- Product cost models only consider design and production costs, ignoring support and customer usage costs.
- Product features take priority over product support considerations.
Several studies have found that support is neglected during NPD. A survey conducted by one of the authors, showed that many high-technology companies do not consider support until well into the product development cycle-on average after two-thirds of the development cycle has passed . At this stage, it is too late to ensure high supportability of product, because most of the design decisions have already been made. With fast time-to-market the norm, it is essential for support to be considered right from the start.
In most cases R&D engineers do not have sufficient knowledge of the issues facing field support personnel. However, field support engineers who know the issues first hand are seldom consulted on the suitability of proposed designs. Research in the US has revealed that field personnel are only “occasionally involved in new product work” . Many companies make big mistakes because they do not use the knowledge base of their customer support organizations. For example, the authors are familiar with a company that developed a new product line based on the architecture from a previous product. This was despite the fact that there was a whole body of evidence from the repair depot that this architecture had led to frequent failures in the past, which were extremely difficult to fix.
A recent survey by one of the authors has found that customer support’s influence early in the product development cycle has increased over the past few years, but it is still not the norm . Leading companies such as NCR, HP, GE and AT&T recognize the importance of panels of field experts reviewing the supportability of new designs . All aspects of product support should be considered, whereas research has shown that many companies limit their evaluation at the design stage to repair and maintenance issues only, thus omitting many other key aspects . The evaluation of all aspects of product support at the design stage is called Design for Supportability.
If companies are not careful, decisions taken to lower production costs may increase support costs. An example is memory boards for products. The goal of minimum material costs often leads production personnel to choose non-reprogramable memory. However, when products need to be field upgraded, this can increase costs-as whole memory boards need to be exchanged. The authors know of two major international companies where this was the case-a few dollars were saved in production but the costs of field upgrades were enormous when the companies had to upgrade their installed base (to solve software problems). It is essential that both R&D and production engineers consider lifetime cost-of-ownership. This is because higher development and production costs may lead to lower lifetime costs!
Documentation Cost Savings: A Pyrrhic Victor
Product management at CompanyX, a maker of desktop software tools, decided to lower the costs of documentation shipped with the product. Since their sophisticated software products only retail for about $200, the cost of printing the documentation exceeded the cost of manufacturing the software medium and other packaging. So product management set a goal to cut the documentation costs in half for the next product release. The support organization, which at the time was not part of the “core team” that made these product decisions, only found out about this decision when asked to review the documentation (one task where support is typically consulted). When they saw the abridged documentation, support was flabbergasted because they knew that the information omitted would lead to a huge increase in calls to technical support. Since these calls cost the company on average $20-25 apiece, the documentation cost saving would be more than offset by increased support costs. They built a business case showing the trade-off and suggested ways to embed more of the documentation information into the product through templates and on-line error messages. Thus, they helped meet the goal of reducing documentation cost without increasing support costs.
A major reason why Design for Supportability is neglected is that R&D engineers tend to give product features a higher priority, as these may be more interesting to develop. Product marketing may push new product features and “bells and whistles” over functions that could make products easier to support since they feel new features allow selling the product to a broader customer base. In addition, product managers may forget the importance of support and the associated revenues. This is particularly common at companies where product-development entities are a separate organization from the field organization, which supports released products. In this case, the field organization’s revenues are separate from the manufacturing division’s and so there is little financial motivation for the product groups to develop products which are easier and more profitable to support. Implementing DFS may require significant investments by R&D for which no return will be seen in the product-development entity’s bottom line. However, leading companies evaluate the return on investment from product features and compare this to how investments in supportability can give returns over the working lifetime of the product. In essence, supportability is treated as a feature.
Design for Supportability at the Leading Edge
Some companies such as Rank Xerox, Hewlett-Packard and VendorC excel at DFS. They achieve this through a clear understanding of customer needs and an evaluation of product support early in the design cycle.
Rank Xerox was one of the first companies to fully evaluate support requirements at the design stage. They are particularly strong at surveying customers’ support needs-something that many companies miss when they conduct market research that focuses on product features alone. Rank-Xerox sets goals for all aspects of support at the design stage; an example is their analysis of the access time to various modules of a photocopier. They found that more money was saved by considering field repair times than from speeding up assembly time in production. Why? The reason was simple. A copier is assembled once in the factory but repaired in the field (including disassembly and re-assembly) many times during its working lifetime by field engineers. Rank-Xerox now has a company design goal that the field access time to any component/module in a photocopier must be less than five minutes.
VendorC, the vending machine company excels at Design for Supportability. Their new products typically require 18 months development, and the team working on NPD includes R&D, product management, manufacturing, suppliers and customer support specialists from the start. At the design stage an analysis is made of the RASUI of products-the reliability; availability; serviceability; usability; and installability. In each of these five categories design goals are set and the Quality Department has responsibility for ensuring that RASUI goals are met during NPD. VendorC’s focus on customer support has enabled them to gain a significant competitive advantage and the company has achieved good returns on its investments in DFS.
One of the world’s largest software companies improved DFS by creating the role of “Supportability Engineer”. This engineer is responsible to both communicate the lessons learned by support personnel to the product designers and to develop new data collection tools. The software company also made product support a cost center and instituted Activity Based Costing in its accounting system. As a result, product designers quickly became interested in improved supportability features that drove down the product’s bottom-line costs (costs which product managers now also saw on their P&Ls, which affected their personal bonuses).
Hewlett-Packard has always had a strong reputation for good product support and has focused on DFS since the early 1990s. Two key elements in their approach to DFS are their lifetime costs model and their financial reporting . To identify the issues and costs involved with support, a lifetime model is used and Diagram 1 shows an example for a medical electronics product. The diagram illustrates the “events” during the working lifetime of a product, from installation, through use and maintenance, to removal from service. By determining the frequency of each event and the associated costs, better decisions can be taken about the potential returns on investments in making support more efficient. For example, installation normally only takes place once but a product may require several upgrades over its lifetime.
Hewlett-Packard use a lifetime costs model to analyze both the frequency and associated costs of all support “events” (such as installation or upgrading).
Therefore, the lifetime model ensures that all aspects of support are considered at the design stage and not just repair and maintenance. As mentioned earlier, supportability issues may have to “compete” for resources during NPD with product features or manufacturability. Thus, a clear understanding of the cost implications of support is essential and the lifetime cost model helps. However, Hewlett-Packard has gone further and changed their financial reporting so that product divisions (which design and manufacture products) see financial returns on DFS investments. Previously, field support revenues had been reported separately and so the motivation in the product divisions to invest in DFS was low (because these investments would make the field support organization more efficient but not increase factory margins). Following the change in financial reporting, better decisions have been made during NPD.
Steps to Achieving DFS
Extensive research by the authors in a range of industries shows that companies typically go through five stages in the progression towards full Design for Supportability, as illustrated by Diagram 2. Companies at Stage 1 do not recognize the potential of support business. Consequently, they do not evaluate support at the design stage. Symptomatic of companies at Stage 1 is the customer support organization having a low status and field support complaining about product designs. Poor product design means higher repair costs and can lead to dissatisfied customers. Therefore, failing to consider product support at the design stage often has a negative business impact, as indicated by the position of Stage 1 on the diagram.
The Five Stages, which companies typically move through, until they achieve effective Design for Supportability.
Some leading companies have reached Stage 5, which is characterized by all of the issues considered at At Stage 2, companies consider reliability and repair times at the design stage and typically set quantitative goals for product reliability (mean-time-between-failures, MTBF) and ease-of-repair (mean-time-to-repair, MTTR). However, broader aspects of support are not considered at the design stage. Further progression leads to Stage 3, where companies involve panels of field engineers in NPD reviews. However, often field engineers’ suggestions cannot be implemented because reviews come too late in the development cycle. Therefore, it is essential to evaluate all aspects of support at the design stage i.e. installation times; fault diagnosis times; field access times; repair times/costs, user training times; upgrade times; etc. Integrating this into NPD is difficult, and it often takes companies a long time to reach Stage 4. Companies at Stage 4 set quantitative goals during the design phase for all aspects of support. These goals push R&D engineers to develop new products, which are easier and more profitable to support than previous products. Cost models are used to guide the decisions about trade-offs between features, manufacturability and supportability.
Stage 4 with two important additions. Firstly, financial reporting mechanisms are used to ensure that return on DFS investment is clearly visible to management. Secondly, and fundamentally, companies that reach Stage 5 have management teams, which fully recognize the importance of support to their business models and to their product’s value propositions. Consequently they devote sufficient resources to analyzing a product’s cost structure over its entire life cycle. Achieving this change in attitude is something many companies find hard and few companies have reached Stage 5. However, the competitive advantage resulting from DFS can be stark.
Volkswagen Wins Through Clever DFS
An example of the advantage of evaluating repair issues at the design stage comes from the car industry. During product design Volkswagen evaluates the probability of each part of a car being damaged in an accident and the associated repair costs. This is important because car insurance companies closely monitor the costs of repairing accident damage and use this information in the calculation of which insurance category applies to a particular model of car. Hence, repair costs have a direct influence on the cost of insurance, which itself has a major impact on product sales especially to young purchasers. The cost of insuring the latest model of the Golf has been reduced by full two insurance classes-through bolt-on panels instead of welded ones, and bumpers molded in three separate parts (to allow partial replacement). The lower insurance costs have given VW a competitive edge and forced competitors to make expensive changes in the design of their products and production lines.
Keys to Success
Moving from Stage 1 to Stage 5 is a change that cannot happen overnight. It is a considerable journey for a support organization to increase its influence over the product design stage. Here are some of the keys to a successful transition:
- Make a business case using lifetime cost models. Support organizations know they can deliver value to the product design process, but R&D are not going to listen just because support yells loudly. Consider what information R&D need for their task and develop data collection systems that deliver this information quickly. This will build credibility with R&D, which is critical for success.
- Focus on communication and negotiation skills. For example, the Cutler-Hammer Automation company is just beginning its DFS journey. However, it has already found that engineers from the support organization responsible for liaison with R&D must be first and foremost excellent communicators. Technical excellence is of secondary importance when selecting support engineers for the liaison role.
- Be a team player. Support organizations may feel that they suffer because of poor product design. If you want to develop influence, you need leave any anger aside and focus on doing what is best for the company. Recognize that product design is a process based on negotiation. If you go to the designers with a list of demands, you will not go far (as the Hewlett-Packard customer support organizations initially found). Identify which aspects of product design are truly critical for supportability and focus on those initially, recognizing that engineering will realistically only have time for the priorities.
- Look for allies. The whole world is not against you — though it may seem that way at times. There are some natural allies for support organizations amongst the players in the design and manufacturing areas. Quality assurance groups, for example, typically share many of the concerns of support groups. If you can gain credibility with these allies, then they can carry your message, giving you increased credibility.
- Market the concept. Gaining influence within the design process will require that you sell the idea that support organizations can deliver value. Marshall some evidence before you approach senior management or product designers. Perhaps another division in your company — or another local company — has had positive experiences in this area. If so, get someone from that division or company to come in and speak about their experience. Ideally, try to bring in both a product designer and support person to speak. You may also find evidence from your competitors about their efforts in Design for Supportability. This, too, may create receptive eyes and ears.
- Start small. Don’t tackle a project that is so large that you are doomed to fail. In creating these linkages with designers, you will almost certainly confront cultural obstacles, which will be a major challenge to surmount. Pick a product — along with the designers — as a test case for the new ways of doing business. Make sure the product presents a reasonable challenge, one where you have a good chance of success but that the designers will accept as a good test case.
- Benchmark your current state. You’ll never be able to prove how far you’ve gone without knowing where you started. Before you initiate new business processes, benchmark your support organization and specifically any test case products. Do a thorough analysis of the cost of the current (poor) design for supportability practices. Senior managers are most likely to listen to cost arguments, but also develop measures of customer satisfaction and retention as part of your benchmark. Remember to look at all the costs of current DFS practices, including costs incurred by your customers.
Again, don’t expect overnight success. You will be challenging engrained business processes. Look for small successes, market them well, and move a little deeper into the designers’ den, without losing track of the goal. And remember, if you get products developed right through Design for Supportability, the return on investment is clear-more profitable support revenues and a competitive advantage.
What Do You Think?
Have you experienced problems in trying to achieve DFS, or do you have a success story? In either case the authors would like to hear from you.
Get more information about a major research report on best practices in Design for Supportability.
~ ~ ~
Knecht, T., Leszinski, R. and Weber, F.A. “Making Profits After the Sale”. The McKinsey Quarterly 1993, No. 4, pp. -86.
Goffin, K. “Gaining a Competitive Advantage from Support: Five Case Studies”. European Services Industry, 1(4), December 1994, pp. , 5-7.
Fites, D. V. “Make Your Dealers Your Partners”. Harvard Business Review, March-April 1996, pp. -51.
Rao, J. and Reitz, W., “EMC: Rewriting Customer Service Rules”, Babson College Working Paper, 1999.
Goffin, K. “Evaluating Customer Support during New Product Development – An Exploratory Study”. Journal of Product Innovation Management, 15(1), January 1998, pp. -56.
Page, A.L. “Assessing New Product development Practices and Performance: Establishing Crucial Norms”. Journal of Product Innovation Management, 10(4), September 1993, pp. -290.
Van Bennekom, F. and Goffin, K., “Using Customer Support’s Knowledge Base To Enhance New Product Development,” Service Operations Management Association Conf. Proceedings, 1999.
Hull, D.L. and Cox, J.F. “The Field Service Function in the Electronics Industry: Providing a Link between Customers and Production/ Marketing”. International Journal of Production Economics, 37(1), November 1994, pp. -126.
~ ~ ~
Originally printed in The Professional Journal of AFSMI, August 2000
Written with Dr. Keith Goffin, PhD. Dr. Goffin is Professor of Innovation Management at Cranfield School of Management in the UK. For fourteen years he worked for the Hewlett-Packard Medical Products Group: on new product development, managing customer support groups, and as a marketing manager. He was also largely responsible for developing HP’s “Design for Supportability” program. In 1995, Keith joined the teaching faculty of Cranfield and his research interests are all in the field of innovation. He has published over forty articles in management journals and this is his third article in AFSM’s Professional Journal.