How do data warehousing practices affect BI efforts?

How well do an organization’s data warehousing practices support enterprise business intelligence (BI) efforts? In April 2004, Cutter Consortium conducted a survey of 145 worldwide end-user organizations on the adoption and use of data warehousing and BI technology in an effort to answer this question.

In this Update, I focus on the following issues:

The degree of confidence that organizations have in their customer data

Corporate use of “data stewards” for implementing and enforcing policies for ensuring data quality and integrity in the data warehouse and BI environment

The extent to which organizations are using “privacy czars” to enact and enforce policies and techniques for ensuring the privacy of customer data

The effect of complying with new customer privacy regulations on corporate data warehousing and BI practices

Survey Methods

In examining corporate analytic adoption trends, I’ve decided to focus specifically on how end-user organizations are implementing the technology. Thus, I filtered out survey responses from software vendors and software services organizations that market BI and data warehousing software products or assist in developing analytic applications. Vendors and services firms typically apply advanced technologies (in assisting clients) to a greater extent than do end-user companies. This tends to skew survey results, giving the appearance that more end-user organizations implement the technology than is often the case.

After filtering out BI and data warehousing companies’ responses, the survey findings are based on the responses of 145 end-user organizations of various sizes. Survey findings for enterprise analytics adoption are based on a broad sample of organizations located in various geographic regions and consisting of small, medium, and large companies with yearly revenues ranging from less than $1 million to more than $1 billion.

Confidence in Customer Data

For years, consultants, vendors, and analysts (including myself) have harped on the need for organizations to take steps to ensure the quality of their customer data. Consequently, I decided to ask survey participants just how confident they really are with the quality and integrity of their organizations’ customer data.

However, this is the kind of finding that politicians like to spin their way to make their policies appear as successful as possible. A far more useful finding is that just 15% of organizations are completely confident, with another 56% indicating that they are only somewhat confident. Put another way, almost 30% of organizations express some degree of dissatisfaction in the quality of their customer data.

The fact that 15% say that they are completely confident with their customer data is somewhat surprising; I expected this number to be about 10%. Regardless, 15% is nothing to crow about either; it means that an awful lot of organizations (approximately 85%) are basing important decisions affecting their customer-facing operations on data whose quality they question. Simply put, the quality of the analyses generated by customer analytics and other customer relationship management (CRM) and BI applications can only be as good as the data used to feed them. And if you’re not completely confident in the quality of your customer data, you will always have that nagging feeling that your analyses are weak.

Based on these findings, I estimate that corporate confidence in the quality and integrity of customer data will continue to increase slowly but gradually — as has been the trend for the past five years.

Corporate adoption of data quality stewards

Data quality issues have plagued corporate information systems since their inception. But the proliferation of data warehouses, and the general trend of companies to disseminate BI and other analytic applications for CRM and other purposes throughout their organizations, has revealed the extent to which poor data quality affects the ability to conduct business operations.

Some organizations have attempted to solve their data quality issues by throwing technology at the problem. But as I’ve said in the past, data quality is more than a technology issue. It requires that organizations build data quality procedures into their business processes. And this requires implementing procedures and assigning responsibility for data quality to those responsible for generating and/or handling the data. One way for organizations to do so is to assign a “data steward” with the authority to plan, implement, and enforce data quality practices. The assumption here is that organizations with one or more designated data stewards are in a better position to successfully enact and enforce standards for data quality across their organizations. Of course, this raises the important question: how many organizations actually designate data stewards who have the authority to enact and enforce policies and techniques for ensuring data quality for data warehouses, BI, and other applications?

The findings suggest that although just one-quarter of surveyed organizations currently have designated data stewards, the percentage will increase considerably (to approximately 53%) within the next 12 months. But that’s only if these organizations’ plans pan out. I’m going to play it conservatively and bet they will not. Consequently, I estimate that use of designated data stewards will increase by no more than 5% by mid-2005. However, we won’t know for sure until we conduct another survey in that time frame. Although I believe that data stewards are an important part of ensuring organizational data quality, various issues complicate their use. First, most organizations today have IT environments consisting of a multitude of different vendors’ operational (i.e., enterprise resource planning) systems, supply chain systems, data integration tools, and BI applications often spread out among geographically dispersed divisions.

Consequently, just defining and agreeing on a set of standardized data integrity processes and tools can be daunting. However, the biggest issues are probably cultural. Different departments and divisions tend to believe that they own the data and the processes that generate them. This makes it very difficult to designate a specific individual or group that actually has the authority to enact and enforce organization-wide data quality standards and practices. In addition, cultural issues are very difficult to overcome when it comes to implementing new policies and procedures because they tend to cross political and organizational (in addition to geographical) boundaries, which means they do not lend themselves to solutions using technology alone.

The use of privacy Czars

Over the past few years, consumer privacy has taken center stage in all forms of media. Businesses around the world of all makes and sizes have found themselves the target of government regulatory agencies, consumer groups, and disaffected individuals for various lapses in ensuring their customers’ privacy. New regulations (e.g., the National Do Not Call and Do Not E-Mail registries, etc.) have also been enacted. In an effort to placate customer concerns, some organizations have installed designated “privacy czars” who are charged with enacting and enforcing policies and techniques to ensure the privacy of customer data and compliance with privacy regulations. I sought to determine the extent to which organizations are actually utilizing privacy czars.

Thirty-nine per cent of organizations surveyed say that they have privacy czars in place. This is very impressive. To be honest, I expected a figure of about 25%. This number — combined with the finding that another 17% indicate that they plan to designate a privacy czar within six to 12 months of the survey — indicates that organizations are taking consumer privacy quite seriously. One could argue that new privacy laws backed up by fines and the threat of lawsuits have forced companies to act in this manner. To be sure, this has had some effect; however, as I’ve said in the past, taking steps to ensure the privacy of customer information just makes good business sense. And appointing a privacy czar indicates to your customers (and the authorities) that you take their concerns seriously and that you are making a concerted attempt to ensure the privacy of their personal information. Based on our findings, I estimate that the corporate use of privacy czars will increase approximately 10% by mid-2005.

Compliance effect

Findings indicate that compliance is definitely forcing changes in organizations’ data warehousing and BI practices. Currently, 35% say that their organizations have had to modify applications and/or data analysis methods to comply with new customer privacy regulations. Another 28% say they will have to make changes to their applications or analysis methods within the next six to 12 months in order to be in compliance. This trend will only increase as more government and industry-sponsored consumer privacy regulations are enacted.


The purpose of this survey was to generate findings to gauge the extent to which companies are implementing/distributing BI analytics throughout their organization in support of their enterprise analytics initiatives, and to identify key application development issues and trends.

Despite common knowledge that data quality is crucial to the overall success of both analytical and operational systems, only a limited number of organizations express complete confidence in the quality of their customer data. The reality is that the majority of organizations (85%) still have a long road ahead of them until they achieve the level of customer data quality they desire (and require). In the short term, this trend will not change or will change slowly. Over the long term, however, this trend will have to change as more and more organizations are forced to conduct studies on their analytic applications in order to measure and optimize their effectiveness and increase their ROI.

My findings covered in this Update point to various trends and considerations:

Data stewards are a vital part of any organization’s data quality initiative because they provide the all-important authority figure that can actually set and enforce enterprise data quality standards and policies. Thus, the more power they have, the better chance they have to overcome the cultural and political issues that are sure to arise as they try to enforce policy. Consequently, a better term for them might be “data dictators.”

Organizations’ use of privacy czars as a way to indicate to their customers and regulatory bodies that they take privacy regulations seriously will remain a popular trend for the foreseeable future. The bottom line is that if your company does not have a privacy czar (and is not planning to designate one within the next six to 12 months), it is lagging behind the general trend.

Compliance with consumer privacy regulations is definitely affecting organizations’ data warehousing and BI efforts. Moreover, the need to modify data analysis applications and practices will become increasingly common as new government regulations and industry practices are enacted. Consequently, organizations should seek to employ BI application architectures and processes that they can easily adapt to meet constantly changing regulatory requirements.