Beyond merely asserting ‘Quality is Key’ | Article

If you can’t quantify the benefits of quality, you are only paying lip service to it

Top quality sealExecutive Summary:

As companies rush abroad to cut back-office costs, one fact gets forgotten in the fray. Costs of small increases in error rates are enormous and can easily wipe out cost reductions from cheaper labor. A data-entry error in a document such as a loan application may make it impossible to process the document automatically, incorrect loan documents may be created, the end customer may call with complaints, and there might be soft costs as well.

Based on interviews with more than 50 financial services firms, BeyondCore determined that if the data-entry associated with one document costs $1, the downstream costs incurred for one document with a data-entry error can easily add up to $300. This means that if outsourcing increases the error rate by just one-tenth of a percentage point, the resultant costs wipe out the benefits of a 30 percentage point reduction in labor costs. Thus, in back-office outsourcing, quality is 300 times more important than labor costs.

BeyondCore recently evaluated an offshore 100 business process outsourcing (BPO) vendor serving a US insurance firm. We found that over 17 percent of claims this supplier processed had errors in critical fields such as social security number, claim amount, and diagnosis code. Furthermore, over 40 percent of claims had at least one data-entry error. Ironically, the BPO vendor had not violated its service level agreements (SLA).

If executives truly want to save money from outsourcing, they need to place far greater emphasis on quality than on processing cost.

Quality is Key. Really?

If quality is key, why do companies spend 10 times more effort on price negotiations than on quality discussions? In conversations with financial services executives, we found that they consistently listed “quality” as one of the top three decision-making factors for BPO. Yet, when we spoke with vendors, we heard that at least 10 times more effort was spent on price negotiations than on quality discussions. In fact, the few executives who were willing to discuss the cost and error rate of their BPO vendors always knew precisely how much they were spending on their BPO vendor while only one was able to precisely specify the error rate of his BPO vendor. It turned out that this particular executive had just received the annual quality audit report for his BPO vendor and thus we had asked the question at a particularly convenient time.

Of course, due to the low response rates and the small sample size we were working with, our experience did not allow us to make any scientifically valid conclusions. Thus, we turned to a recent survey by Deloitte to determine what its analysis revealed regarding the true importance of quality in BPO.

“Outsourcing … is still dominantly driven by cost-related objectives” An April 2005 Deloitte white paper entitled “Calling a Change in the Outsourcing Market” states: “Outsourcing, which originated as a popular cost-saving strategy during a recessionary economic environment, is still dominantly driven by cost-related objectives.” The article goes on to specify that “83 percent of participants mentioned cost savings as an expected benefit of outsourcing; 70 percent of participants cited cost savings as one of the key drivers of outsourcing; and 43 percent of participants declared cost savings as the primary criterion for choosing a vendor.” While Deloitte does cite two third-party articles that suggest that outsourcing is not primarily driven by cost, its own data highlights that outsourcing “is still dominantly driven by cost-related objectives.”
If quality and cost savings are seen as tradeoffs, and cost savings are still the primary driver for outsourcing, then the focus on quality will invariably suffer. One of the cost-related risks the Deloitte article mentions is “trade-off between quality and cost savings.” Accenture’s 2004 Executive Survey titled “Driving High-Performance Outsourcing: Best Practices from the Masters,” also states, “Establishing business outcomes is all about the outsourcing company answering a few critical questions, beginning with ‘What do I want the provider to optimize?’ Is it price or speed or quality or a combination of these?”These comments highlight an underlying assumption that there is an implicit tradeoff between cost and quality. However, if quality and cost savings are seen as tradeoffs, and cost savings are still the primary driver for outsourcing, then the focus on quality will invariably suffer.

This situation is only made worse by the fact that quality is usually more difficult to define and monitor. Thus, it is easier to evaluate the outsourcing initiatives and the executives managing them based on cost savings rather than quality improvements.

It is not about lower costs vs. higher quality; it is about achieving lower costs through higher quality. This debate regarding the tradeoff between low cost and high quality is nothing new. In manufacturing, the same debate raged until the Toyota Production System (TPS) proved that the best way to reduce overall costs was to focus on high quality. It is not about lower costs vs. higher quality; it is about achieving lower costs through higher quality. Until TPS, people did not realize that the eventual cost of bad quality from product rework, defective product write-offs, and resultant customer service costs completely overwhelmed the cost of production. By focusing on preventing defects, TPS actually significantly reduced the overall cost of production.

While this lesson is generally accepted in manufacturing, in the services sector the same old debate regarding quality vs. costs rages on. This article will demonstrate that the same paradigm holds in BPO; therefore we need to end this debate and focus on reducing costs primarily through improving quality.

First, we need to confirm that the TPS lessons regarding the primacy of quality are also relevant in BPO. Second, we need to find a way to quantify the cost of errors and thus the benefits of high quality.

Quality is 300 Times More Important Than Costs (Case Study)

In an October-December 2004 Deloitte Consulting Outsourcing Study, 64% of respondents stated that they had brought some outsourced services back in house. A US financial services company outsourced a data-entry process to a leading BPO vendor. The BPO vendor met all of its SLA targets and reduced the processing costs by 30 percent.However, the BPO customer soon found that its overall operational costs had actually increased after it outsourced the process. Eventually, the customer brought the process back in house. This turn of events is not rare: In an October-December 2004 Deloitte Consulting Outsourcing Study, 64 percent of respondents stated that they had brought some outsourced services back in-house.

The following simplified and disguised case highlights the importance of quality in BPO engagements. In this insurance claims scenario, the client’s cost of processing each claim was $1, but the average downstream cost due to errors in a claim was $300. The $300 average downstream costs included manual exception handling costs, costs of customer support calls initiated due to errors in claims, and costs of reissuing corrected documents for any claims processed incorrectly the first time.

The customer also faced significant soft costs from regulatory risk, lost revenues due to low customer satisfaction, and had to bear the costs of overpayment on claims due to claims processing errors. These soft costs were not included in the $300 number. The scenario faced by the customer was as follows:

In-house cost of processing 1 document $1
Outsourced price for processing 1 document $0.70 [30% cost reduction due to outsourcing]
Average cost to the company due to 1 incorrectly processed document $300 [unaffected by outsourcing]
Number of documents processed annually (millions) 100

This client only outsourced the back-office processing. All of the downstream activities which constituted the “average cost to the company due to one incorrectly processed document” remained in house and thus this cost was not affected by outsourcing.

Because the cost of correcting errors is so much higher than the cost of processing a document, the error rate of the vendor became the true determining factor for the total cost of ownership (TCO). Unfortunately, the insurance company had placed far greater focus on ensuring the greatest reduction in unit processing cost. As a result, the firm managed to reduce its direct processing costs, but the TCO remained unchanged or even became worse than before.

In this scenario, just a 0.1% point increase in errors wipes out the benefits of a 30% point reduction in processing cost!

In-house

Outsourced

Direct processing costs (millions)

$100

$70

Error rate (percent of errors not caught by direct operations)

1.0%

1.1%

Number of documents with uncaught errors (millions)

1.0

1.1

Eventual cost of uncaught errors (millions)

$300

$330

Total Cost of Ownership (direct cost + cost of errors)

$400

$400

Gross outsourcing benefits (difference in direct processing costs)

$30 million

Net outsourcing benefits (difference in Total Cost of Ownership)

$0

BeyondCore has developed a framework called the Total Cost of Errors (TCE) to identify and quantify the costs incurred due to low quality In this case the customer had a bad experience with outsourcing. However, top outsourcing firms usually adopt advanced quality management practices and often achieve higher quality standards than their customers. In these cases, the value created due to outsourcing may be severely underestimated.For example, a BPO vendor that halves the error rate but reduces direct costs by only 15 percent actually creates almost three times the net benefits compared to a BPO vendor that delivers 30 percent reduction in direct costs, but reduces the error rate by just 0.1 percentage points. Moreover, the net benefits created by the BPO vendor that halves the error rate are 11 times greater than the difference in direct costs. Thus, traditional measures of outsourcing benefits would underestimate the benefits of such an outsourcing relationship by an order of magnitude.
Scenario 1 Scenario 2
Direct processing costs (millions) $85 $70
Error rate (percent of errors not caught by direct operations) 0.5% 0.9%
Number of documents with uncaught errors (millions) 0.5 0.9
Eventual cost of uncaught errors (millions) $150 $270
Total Cost of Ownership (direct cost + cost of errors) $235 $340
Gross outsourcing benefits (difference in direct costs) $15 $30
Net outsourcing benefits (difference in TCO) $165 $60

This case certainly demonstrates that in BPO, quality is significantly more important than cost. However, we still need a framework for establishing and quantifying the importance of quality. BeyondCore has developed a framework called the Total Cost of ErrorsTM (TCE) to identify and quantify the costs incurred due to low quality.

Measuring the Importance of Quality: The Total Cost of Errors

Total Cost of Errors = per transaction average costs specifically incurred due to errors in the output of the back-office process The best way to measure the benefits of quality is to define and quantify a “Total Cost of Errors” (TCE). For transactional business process outsourcing, (such as claims processing, application processing, HR recordkeeping, etc.), BeyondCore defines this as the per transaction average costs specifically incurred due to errors in the output of the back-office process.For example, in a credit card application processing scenario, buyers incur costs to detect and correct errors in the applications processed by the vendor; errors in an application may prevent straight through processing at the customer side and force expensive manual processing; errors may remain undetected and affect end-customers who generate expensive customer support calls. The buyer would not incur any of these costs if the processing were completely error-free.

These costs, depicted in Figure 1, constitute the TCE, which has four major components:

  • Quality Control Costs: Costs of QC systems/staff to detect and correct errors. This cost is usually included in the price charged by the outsourcing vendor but is ultimately borne by the customer. Some customers also engage in-house staff to sample and confirm the quality of the BPO vendors’ output.
  • Downstream Error Costs: Costs incurred by BPO customer in downstream processes because of errors caught and corrected before the end-customer is affected. One large bank’s loan officers spent up to 25 percent of their time catching and fixing data-entry errors. Due to errors, a midsize healthcare services company has to manually process approximately five to ten percent of the transactions data-entered by its BPO vendor.
  • Uncaught Error Costs: Costs incurred in remediation steps once an end customer is affected by one or more errors. For example, if an incorrect loan document is issued, the customer may call in to complain about the mistakes. The expenses for each customer service call, as well as the cost of reissuing a corrected loan document, would be included in the TCE.A mid-sized insurer found that due to data entry errors in its new application process, up to 10 percent of its customers received incorrect documents and made multiple calls to the customer service center before the errors were corrected. In fact, almost one percent of the applications required three or more customer service calls before all the data entry errors were corrected.
  • Soft Costs: Hard to quantify and not immediately evident costs that are incurred because of back-office errors. Soft costs may include: lost revenues due to low customer satisfaction, operational losses due to bad decision-making based on erroneous data, or regulatory risk caused by incorrect processing or recordkeeping.

Figure 1. TCE: Adding up the true cost of back office errors

back office process
Click on image for larger version.

Simplified TCE Example Based on Global 1000 Bank Data:

Quality Control Cost Downstream Error Costs Uncaught Error Costs Soft Costs
Average Error Rate 5% 2% 0.5% Hard to quantify. Customer ballpark estimated $5 based on historic experiences.
Average Error Cost $20 $75 $400
TCE Component $1 $1.50 $2 $5
For every $1 this firm spent on this back-office process, it spent $4.50 dollars on hard costs related to back-office errors generated in this process Thus, for this process, the per-document Hard TCE is $1+$1.50+$2=$4.50, while the per-document TCE including Soft Costs is $9.50. The per-document Direct Cost of Operations (DCO) for this process was only $1. Thus, Hard TCE was 4.5 times greater than the DCO. In other words, for every $1 this firm spent on this back-office process, it spent $4.50 on hard costs related to back-office errors generated in this process.While each company’s TCE ratio is unique, TCE is almost always several times larger than DCO.

What does this mean for the importance of quality versus costs, especially in the case of business process outsourcing? First, if TCE is greater than DCO, then quality is more important than processing costs. Second, if outsourcing increases your error rate even slightly, it can wipe out significant reductions in processing costs. This is because traditional outsourcing only affects DCO and quality control costs and does not affect the other components of TCE. Thus, in an outsourcing situation, the TCE-to-DCO ratio actually increases, making quality relatively even more important.

Once we establish the importance of quality in BPO, we are faced with a different challenge: how do we reliably monitor the BPO vendor’s quality and enforce quality-related SLAs? This is not an easy problem to solve and companies have lost millions due to badly designed or laxly-enforced quality SLAs.

The Quality Oversight Trap

A 1999 Everest Group white paper titled “Avoid a Multi-Million Dollar Mistake” states:

“If you can’t measure it, it won’t get done.” “In any outsourcing relationship, it is important to remember that ‘If you can’t measure it, it won’t get done.’ Therefore, detailed service level agreements are not only necessary, they are critical to the success of the relationship. Typically, these agreements are not defined in as much detail as they need to be.”

This is excellent advice. Unfortunately, BPO customers often equate the word “detailed” with “fine-grained.” A detailed SLA is not just fine-grained; it measures those metrics that are of greatest operational importance to the customer, and it covers multiple possible contingencies.

One US insurance firm developed a very fine-grained quality SLA for its BPO vendor. The company insisted that less than 0.5 percent of document fields could have errors. However, given that the document had almost 200 fields, it was possible to meet this fine-grained SLA while having critical errors in every single document processed.

On average, more than 17% of claims processed had critical errors, and more than 40% had some errors In fact, when BeyondCore analyzed this vendor, we found it was in perfect compliance with the SLA, but on average more than 17 percent of the claims it processed had critical errors, and more than 40 percent had some errors. The fine-grained SLA was clearly flawed!

Other BPO customers indicate that in an effort to improve quality, they try to control which BPO vendor employees can work on their process, the salary paid by the BPO vendor, the ratio of supervisors to operators, the bench strength, and similar factors down to the finest detail.

At this point, the customer is so involved in the day-to-day operations of the BPO vendor that it ends up squelching any initiative on the part of the vendor and stands to lose the soft benefits (e.g., process improvements, improved executive focus on core business) of outsourcing. In fact, if the buyer encounters problems, the vendor might be right in claiming that the customer’s micromanagement was a major component of the problem. An outsourcing relationship is like a cross-cultural marriage where unstated resentments often sabotage the whole relationship.

If you don’t precisely know your quality goals and how to define quality in terms of those goals, then a detailed measurement / monitoring system will only give you a false sense of security and control It is imperative that the customer clearly defines its quality goals and then sets up SLAs based on those goals but refrains from dictating the tactical means necessary to achieve them. If you don’t precisely know your quality goals and how to define quality in terms of those goals, then a detailed measurement / monitoring system will only give you a false sense of security and control.

For example, if the primary focus is to reduce manual exception-handling costs, then you need to measure the percent of documents with critical errors. Once a document has to be manually handled because of one field with critical errors, then the incremental cost of fixing other errors in the same document is not as high.

However, if the primary concern is expected regulatory risk or operational risk, then every field counts. If there are two different errors in the same document, they may lead to two independent regulatory/operational risks. At the same time, each field does not lead to the same amount of regulatory/operational risk. Thus, an error in a field such as Social Security number has to be weighted higher than an error in Zip Code.

Of course, in many cases there may be multiple quality goals. However, the relative importance and tolerances for these goals have to be explicitly agreed upon. A TCO-oriented view of cost, quality, and operational risk, based on a detailed TCE analysis, should form the core of any such comprehensive quality measurement framework. Once the goals are clear and quality has been defined precisely, the customer should monitor quality on an ongoing basis while letting the vendor achieve the agreed-upon quality goals in its own way.

Collaborating on Quality: Moving from Oversight to Insight

In traditional BPO, vendor quality oversight involves periodic quality audits of the BPO vendor, followed by sessions focused on problems identified during the audit. Such quality audits are usually quite confrontational and ineffective.

First, if the quality audits are carried out only once or twice a year, they merely provide widely-separated snapshots of quality and do not provide an ongoing means of monitoring the vendor’s quality. Second, they are often confrontational in nature, and if mishandled, can seriously hurt the outsourcing relationship.

As outsourcing becomes a key corporate strategy, it is imperative that customers move from arms-length quality oversight to ongoing BPO Quality InsightTM. BPO Quality Insight incorporates three major transitions from traditional oversight:

BPO Quality Insight Quality Oversight
Ongoing real-time quality visibility Periodic quality audits
Focus on the root causes of errors irrespective of whether the vendor or the customer controls the root cause Focus on what the BPO vendor could do to improve quality
Predict and prevent problems Fix problems after the fact
  • Ongoing real-time quality visibility: Leading customers have been manually conducting quality confirmation sampling for some time now. Automated software systems that periodically sample the quality of BPO vendors and provide real-time objective quality visibility constitute the next phase in this trend. These automated solutions are overall more cost effective and more accurate that traditional methods. Moreover, these automated systems are not subject to human bias and thus diffuse any conflict arising out of the inherent subjectivity of a manual quality sampling.
  • Focus on the root causes of errors irrespective of source: Most organizations do not yet outsource every component of the business process. For example, a BPO vendor may be responsible for data entry, but the BPO customer maintains control of the software systems into which the data is entered. Thus, the systems-process-people root causes of errors can stem from components under the control of either party.A large international financial institution wanted to improve the quality of its captive BPOs in India and Ireland. When it analyzed the captive using a leading quality insight software, executives identified easy-to-implement opportunities for improving quality by 30 percent in days. While a significant component of these quality improvements were under the captive’s control, two of the largest quality-improvement opportunities were in the systems and processes under the control of the US parent. By collaborating with its Indian captive and creating a “blame-free” environment, the customer was able to quickly identify far greater opportunities for quality improvement than it could have achieved using traditional oversight.
  • Predict and prevent problems: The Deloitte “Calling a Change in the Outsourcing Market” white paper states: “One in seven participants expressed frustration with determining and measuring SLAs, and experiences have indicated that vendors can ‘twist the numbers’ to protect themselves.”A significant source of the number-twisting problem is that these discussions tend to happen after a significant problem occurs and the focus is primarily on assigning blame. Customers need to identify and monitor early indicators of quality problems, because it is far more difficult to fix problems after they occur. For example, increased variability in output quality can be an early predictor of future quality problems. Preventing problems is an inherently collaborative activity because the vendor can’t be blamed for problems that have not yet occurred. By creating a blame-free environment, customers and vendors can truly collaborate to reduce quality problems.

An Executive Checklist

In summary, this best-practices checklist focuses the reader on BPO quality and the resultant TCO-based vendor management paradigms. We leave for another time discussion of other significant outsourcing benefits such as business flexibility, financial structuring of assets, best-practices acquisition, and disaster recovery support, to name a few that are out of the scope of this article.

  1. While evaluating whether or not to outsource:
    • For each in-house process you are considering outsourcing, you must first understand the relative magnitude of the hard TCE compared to the direct cost of operations.
      • Best Practice: Conduct a detailed TCE analysis of your current processes.
  2. Before outsourcing:
    • Conduct a detailed quality and output benchmark of your existing process.
      • Best Practice: Enterprises often have existing benchmarking organizations that compare core business processes with those of industry peers. That same organization can conduct this kind of benchmark.
    • Identify the systems-process-people root causes of existing errors and what these imply for the scope of the outsourcing project. For example, if systems are a significant source of problems, then outsourcing only the people-intensive functions may turn out to be a disaster. Once the responsibility is split between you and the outsourcer, it will be even more difficult to identify and fix the problems. Hence, you should either fix major systems and process problems before outsourcing, or outsource responsibility of all the problem components to the same outsourcer and hold it responsible for fixing the problems.The key here is that correcting the major root causes of problems should be the primary responsibility of one single entity: either you or the outsourcer.
      • Best Practice: Do not outsource a problem. If you can, fix the “low-hanging fruit” sources of quality problems before outsourcing. This is likely to dramatically improve the economics of the process and will save you the major headaches that would occur when the same problems are encountered after outsourcing.
  3. Vendor selection / contract negotiation:
    • Conduct a quality benchmark of your short-listed vendors based on a process that is substantially similar to your own.
      • Best Practice: During vendor due diligence, understand the typical transition process the vendor has utilized with other clients such as yourself, and benchmark the effectiveness of the transition processes. For example, if the vendor claims that it requires three months to ramp up a process to full quality, try to measure an existing process that they have been managing for just about three months.
    • Before detailed contract negotiations, provide the vendor your quality and output benchmarks for the process to be outsourced. In the vast majority of cases, vendors do not have detailed customer baselines on which to structure their cost expectations and SLAs. As such, the contract language can be difficult to agree upon. Moreover, the lack of visibility into customer baselines leads to a classic information asymmetry problem. This lack of information may lead the vendor to underestimate the cost of meeting the contract SLAs, and it may eventually end up burning significant amounts of money trying to meet unrealistic SLA terms. Alternatively, the vendor may assume that the customer is underestimating its quality problems and may quote a higher price anticipating the expense of meeting difficult SLA terms. In either case, significant distrust may be created in one of the very first steps of a five-year outsourcing relationship.
      • Best Practice: Proactively provide candidate vendors your quality and output baselines. Work with your outsourcing advisor to ensure that you address the information asymmetry problem as best as possible.
    • Define contract SLAs based on the TCO of the outsourced process.
      • Best Practice: Create significant value-sharing opportunities based on TCO reduction. If the mutual quality goals are well understood and measurable, and the TCO framework is well defined, a flexible value sharing agreement is the best way to incentivize your vendor into allocating the best resources to your process. It is far easier to manage a vendor based on incentives rather than through micromanagement via contract language.
  4. Vendor management:
    • Set up ongoing visibility of BPO vendor TCO. Monitor early indicators of quality problems.
      • Best Practice: Use an automated system that provides ongoing objective visibility of the BPO vendors’ quality.
    • Periodically benchmark the BPO vendor’s quality against other vendors.
      • Best Practice: Collaborate with your peers who are also outsourcing similar business processes to benchmark your BPO vendor’s quality against those of your peer group.

Author Profiles:

Arijit Sengupta is the CEO of BeyondCore, Inc. BeyondCore’s software is the state-of-the-art in business service operations management for in-house and outsourced back-office operations. Arijit previously worked at Oracle, Microsoft, General Motors, and the Yankee Group. He has spoken at conferences in a dozen countries, including XML Europe, Oracle Openworld, and Java One. Arijit was a member of the ebXML Joint Coordination Committee and held leadership positions in several other eBusiness initiatives. In 2003 he received the RosettaNet “Outstanding Leadership in the High Tech Industry” award. Arijit holds an MBA with Distinction from the Harvard Business School and Bachelors degrees with Distinction in Computer Science and Economics from Stanford University.

[email protected]

George A Logemann is an Executive Vice President and Co-Founder at BeyondCore, Inc. George has over 20 years’ experience in the field of outsourcing, as an early customer at Eastman Kodak where he was part of crafting one of the first outsourcing megadeals; as a vendor at IBM, Compaq, and EDS where he led $250M+ sales opportunities; and as an advisor at the Yankee Group, a leading technology advisory firm, where he directed its outsourcing research / consulting organization and was personally involved with over 40 outsourcing transactions while providing counsel to most major outsourcing vendors.

[email protected]

Post a Comment

Your email address will not be published.

( required )

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>