Poor data quality costs enterprises money: by making business processes less efficient, by increasing the cost of maintaining contact with their customer base and through loss of customers due to poor customer service provision. It leads to breakdowns in the supply chain, poor business decisions, and hampers efforts to meet regulatory compliance responsibilities.
Poor data quality is a costly issue. According to a PricewaterhouseCoopers Global Data Management Survey, 75 per cent of those surveyed reported significant problems as a result of defective and fragmented data, more than 50 percent had incurred extra costs due to the need for internal reconciliations, 33 percent had been forced to delay or scrap new systems, 33 percent had failed to bill or collect receivables, and 20 percent had failed to meet a contractual or service-level agreement (Friedman et al., 2004). I interviewed 16 industry experts, in roles ranging from Database Administrator to Data Operations Vice President, in locations as disparate as the United States, the United Kingdom and Singapore, in order to elicit examples of the tangible costs of poor quality data.
Half of respondents said incorrect billing, often caused by failures in creating a single customer view, is a major cost of poor quality data (15.69% of responses to the question). Customer relationships are damaged when companies produce incorrect or untimely bills. Customers may feel that a business that does not present timely and accurate bills is disinterested in producing the quality services and products they expect, with an obvious effect on customer retention. Beyond customer alienation, costs to handle billing complaints increase operational overhead and expend non-value-added customer care time. An employee resolving a billing problem could be better employed time cross- or up-selling, or engaging in other value-added work. In addition, billing systems are normally not biased to over-bill. When they under-bill even conscientious customers are less likely to notice or report this. This incorrect billing directly impacts the timely collection of revenue. The greater number and frequency of billing errors, the longer it takes to collect revenue.
Further to this, inaccurate billing creates difficulties in knowing the genuine financial picture. ‘Billed’ revenues may be false indicators that will be unmasked only when customers complain, and lost revenues from under-billing may never be unmasked. In summary, billing errors cause customer dissatisfaction, increase operating costs, prevents staff from engaging in value-added work, decreases or delays revenue, and masks real financial performance. A further associated problem is that if a customer is incorrectly identified, that customer may be extended credit where they otherwise should not have been. As Respondent 16 states, “if [the enterprise] doesn’t know who the customer is it is a lot harder to establish… risk parameters.”
Half of all respondents also raised the issue of significant business decisions being formulated based on poor quality information (15.69% of responses). Business intelligence and performance management deployments are being relied upon by more and more users with greater business impact. A broad range of transactional, financial and operational data drives critical business decisions in the modern enterprise. Therefore, the impact of poor quality data is a growing concern as organisations more are more reliant on technology. Information intensive applications such as Enterprise Resource Planning and Customer Relationship Management systems deliver value only if the data they use is reliable, complete and accurate. Process automation increasingly means that data is the foundation of critical business operations, and poor data quality can therefore lead to breakdowns in the supply chain and poor business decisions. From a marketing perspective, poor quality data can result in flawed segmentation resulting in lost business, poor targeting, inadequate budgeting and unreliable financial projections. As Respondent 15 suggested, “the ability to use the data to develop your process products, understand your customer requirements, do asset servicing correctly, track your internal performance, your profitability, all of those things,” relies on data that is fit for purpose.
A major issue again identified by half of respondents was time and resource cost of re-entry and/or verification of incorrect or incomplete data (15.69% of coded responses). There is an associated opportunity cost of hours spent resolving issues caused by data quality problems (as stated previously, staff who are tasked with resolving such issues could be employed elsewhere).
43.75% of respondents asserted that alienation of customers as a result of poor addressing was a primary cost of poor quality data (13.73% of responses). Customers will lose faith in the ability of the company to provide high quality products and services if they fail to address them correctly. Failure to properly capture address data also inevitably results in lost dispatches and increased sending costs.
A significant indirect cost of poor quality data is that opportunities for cross- or up-selling are lost, partly as a result of failure to establish a single view of the customer and their order history. Accurate capture of enquiries and transactions facilitates a better understanding of buying patterns that allows sales representatives to better manage the sales process and ensure high conversion ratios. Business intelligence applications armed with more accurate data can accurately indicate customer profiles that are historically receptive to similar sales efforts. This was highlighted by 31.25% of respondents (9.8% of responses).
Another issue that arises as a result of poor quality data is that of lost production through supply chain problems. This was raised by 25% of participants (7.84% of responses). The flow of data is at the heart of modern businesses’ extended supply chains. Efficient flow of accurate data has helped to fuel impressive productivity gains and advances in operational planning. However, there are numerous points in the supply chain where data quality potentially becomes a major issue and can cause logistical collapses; particularly where data is arrives from external sources, as there are often no universal standards for either the format or content of data. Respondent 8 in particular cited an inability to match and integrate data from disparate internal or external sources as a major and catastrophic result of poor quality data. Different systems and users need to see the data in very different ways within very different contexts; product categories can be divergent (with different schemas, validation and business rules, and definitions). Different systems (for instance, an inventory system and a customer relationship management system) utilise the data differently and repurposing the data can be a complex task. This is certainly one area in which automated solutions can make an impact as the more manual methods are utilised to repurpose complex data throughout the supply chain, the more expensive and unreliable the solution is likely to be. An associated cost can be damaged supplier relationships.
Regulatory non-compliance was cited as a key potential cost of poor quality data by 12.5% of respondents (3.92% of responses). Noticeably, these respondents were based in the United States, and one worked in the financial services industry. Data quality problems can result in difficulty in compliance with current and future regulations such as Basel II, Sarbanes-Oxley, The US Patriot Act and International Accounting Standards.
The remaining costs of poor quality data are to a degree interrelated and can impact one another. Firstly, a significant cost of poor quality data can be an inability to deliver data to decision makers in a timely manner, often as inaccuracies have to be reviewed or corrected. Timely data can mitigate the risks involved in making strategic decisions about competitive strategy, pricing policy and launching products and services. Specifically, accurate data delivered in a timely way can give advanced notice of competitor moves, warn key decision makers about issues in production capacity and facilitate rapid response to market opportunities. This was outlined by 12.5% of respondents (3.92% of responses).
From a marketing perspective, problems in campaign formulation, resulting in poor campaign responses can be a significant cost of poor data quality. This was underlined by 12.5% of respondents (3.92% of responses). Typically, marketing campaigns are formulated by segmenting a database on several dimensions – for example, account types and deposit levels in financial services or usage and service categories within the telecommunications industry. Segmentation based on erroneous data will result in people being targeted with inappropriate offers or products. This leads to wasted spend on marketing literature production and postage, as well as potentially irritated customers (for instance, a forty-five year old woman being targeted with products aimed at the recently retired demographic).
An issue in part resulting from those costs of poor data quality mentioned above is a lack of confidence from end users in the reports being generated by the systems. This was a concern for 18.75% of respondents (in 5.88% of responses). This in turn can be one of the contributory factors in low rates of user adoption of systems, which was outlined by 12.5% of respondents (3.92% of coded responses). Poor data quality is only one factor contributing to low user adoption of systems (sales people can be naturally hesitant to share their information with others in the organisation, for instance), but can undermine incentives for people within the organisation to use such systems.
References
Friedman, T., Nelson, S. D., & Radcliffe, J. (2004). CRM Demands Data Cleansing. Gartner.
