TRILLIUM SOFTWARE INSIGHTS

Data Governance – Where are the Teeth?

 By Jim Orr, Director, Enterprise Data Strategy, Harte-Hanks Trillium Software

I enjoy hearing the terms that so often describe the data governance discipline, especially those that represent unity and harmony. Words such as;

Collaboration, Cooperation, Joint ownership, Consensus, Partnership, Group effort, Teamwork, Alliance, Ambassador, Facilitation
 
I myself have used many of these terms because we all know that success in data governance requires strong cross-functional participation. What I don’t see and hear enough of are words such as authority and accountability.

Why? One reason may be because many organizations choose to build their data governance programs solely on the principle of collaboration rather than institutional authority. Ideally, you really need a good balance of both to make it work, but many companies are currently falling short on the balance thing.

Companies that error on the far side of collaboration and consensus are struggling to secure, implement, and keep control over fundamental areas of data governance essential to developing and implementing standards for data models, quality, business rules, process, etc. Yes, these overly collaborative and typically project based models can drive small incremental progress but are falling short of overall expectations. They are also more likely to fall apart over the course of time.

On the other hand, companies that invoke more institutional authority over data management are starting to see accelerated levels of success. By no means am I suggesting a highly autocratic bureaucracy, however I do endorse a model that clearly defines roles, responsibilities, ownership, authority, and accountability. In addition, one that provides for quick and effective issue resolution at a tactical, strategic, and political level.

So I ask the question, is your data governance program heavy on collaboration or institutional authority? Is it helping or hurting you? Where does it need to be? What do you need to do to get there?

Share your thoughts, as this is a critical subject for your company's future.

Managing Metadata = Managing Data – Is There a Difference?

By Michele Goetz, Director of Product Marketing, Harte-Hanks Trillium Software

I had the great experience of participating in Information Management’s DM Radio panel discussion on metadata this week.  Listen to the recording. Information Management’s Eric Kavanaugh and Jim Ericson asked myself, Rob Karel of Forrester Research, and other data quality providers, some pretty provocative questions that moved metadata from ‘data about data’ to defining its value in driving operational success in business processes and clarifying its role as the backbone to describing the business. 

As we talked about metadata in this context it really hit home that managing metadata is not unlike managing data.  Get your metadata wrong, and you will quickly feel the results of inconsistencies, gaps, and risks in your business.  Quite the data quality issue!

Why is that?  Jim Ericson brought up the point that metadata can be served up in a service that initiates an action.  Think about the status of a claim, the threshold for a stock trade, the instruction to call out to a cloud service.  These can all be metadata elements.  The implication of poor quality metadata in these instances can mean the rejection of reimbursement for healthcare services rendered, investment loss, or a breakdown or bottleneck in a business process.

When building out your data quality practices, don’t forget about metadata.  Profiling, and continuous profiling and monitoring, is key to maintaining alignment of metadata to your business.  As your business, market, and processes change, so too will your metadata.  Things to consider when determining how to define, measure, and monitor metadata are:

• Make sure your profiling solution is robust enough to account for the quantification of metadata conditions both as a way to check against defined standards or glossaries as well as help discover new metadata. 

• Your profiling solution should be available and easy to use by your business data stewards that understand how metadata supports and drives processes, reporting and linkage between their business applications

• There are reporting capabilities that address both context and data structure to ensure there is linkage between what the business needs and expects and how your system will support and serve to that expectation

Metadata is as much about providing context to the business as it is to help data fit within your infrastructure.  Keep in mind the quality of your metadata and apply the same methodologies for data governance and stewardship to be successful and drive value to the business as much as to your data management platform.

The Winds of Financial Data Policy Change are Blowing in the UK

By Richard Wilson, Vice President, Product Strategy, Harte-Hanks Trillium Software

Change is a difficult fact of life for every organization right now.  At Trillium Software, we’re challenged every day to adapt and innovate in response to radical shifts occurring in the information quality market.  At this point, nearly every industry sector is experiencing a similar metamorphosis in their landscape of business partnerships, competitors and customer profiles.  The realignment, happening in the wake of multiple disruptive waves of financial, policy and business change, is impacting everyone’s traditional marketplace and all our business engagements.  Change is everywhere, and all at once!

One interesting aspect of these new ‘business data’ engagements to note is the change in the makeup of the audience interested in the topic of data quality.  Good executives are always very keen to understand and exploit the opportunities arising from detection and remediation of costly errors, whether they occur on the ‘shop floor’ or in the risk office.  So it’s not unusual to see very senior officers intensely interested in data quality issues once the consequences of poor quality data are made clear to them. 

The impact of gaps in data accuracy elevates dramatically when the ‘corporate focus’ on data moves from instances or rows in operational systems to elements  ‘measured’  in risk or performance metrics.  When it impacts at the level of a key performance indicator, bad data produces measurable financial pain, manifested in costs such as increased risk reserves.

Executive interest and understanding is a cornerstone event in developing the foundation for true corporate data governance.  It is very interesting and exciting to see that ‘transformation dynamic’ for data quality play out in many of the sessions we are privileged to participate in.

Turning briefly to another place where changes to the way data is managed, take a look at the United Kingdom financial services industry.  Data is having a huge impact on the environment in which businesses operate in the UK.  

The link was provided to me by independent consultant Ken O’Connor, a very keen observer, blogger, DQ industry expert and author of an E-book on the issues and challenges of creating a single customer view.  It relates to a new mandatory initiative impacting banks in the UK, which will be expanded to Asia, Europe and other regions around the globe.

For those of you unfamiliar with the back story, here’s a brief outline.  In the UK, banks do not have the same deposit insurance that we have in the US, where deposits in banks are guaranteed by the Federal Deposit Insurance Corp (FDIC)

The Financial Services Authority (FSA) in the UK is attempting to address that gap in depositor protection this year, with a project known as the Financial Services Compensation Scheme (FSCS.)  The scheme was designed to protect depositors in the event of a single bank failure, or a more universal financial crisis.  What it is intended to do is to provide every depositor in the UK access up to £50k within a very short timeframe in the event of one or several bank failure(s). 

The scheme imposes a service fee on each deposit taking institution to create a pool of insurance for payout in the event that the failing institution cannot make this information available and the government must step into an administrative role.  The challenge is that banks are required to create a single view of each customer (consumer, commercial and wholesale) and produce ‘on demand’ a list that balances the totals across multiple accounts.  

The problem is banks do not fundamentally work that way.  Banks usually administer money in separate accounts and often do not know about how customer accounts link together.  They may demonstrate an understanding of their relationship with you because of their excellent customer service, but their accounting systems do not reflect the knowledge they have in their heads.  So there are many deposit-taking institutions who will face an uphill battle to produce anything approaching accurate in comprehensive customer data, especially because so many of their accounting systems are old and data does not match across systems.

As the year winds on, more-and-more banks are coming to the realization that they cannot achieve the required accuracy, so pressure is mounting to find a resolution. 

For those us who live in the ‘data world’ this amounts to pretty ‘high drama.’ And if you have an account in a UK bank, you should probably be interested as well.

This UK financial services situation is a highly volatile situation that is bound to see day-to-day “changes,” twists and turns.  Stay tuned to Ken’s blog and my blog here to learn more over the coming weeks and months.  Companies better wake up soon to these changes or introduce huge risk into their organizations and firms.

How are you going to deal with these policy changes?  Have you already thought about it?  Let me know.

Proper Communication Leads to Innovation Breakthroughs

By Bernie LeCuyer, Sr. Vice President, Global Services, Harte-Hanks Trillium Software

Every so often, I’m fortunate enough to experience “breakthrough thinking” when working with a customer.  I see innovative ideas just about everyday but rarely see them become something of significant business value.  Too often, ideas never see the light of day because of something I was once told is called “the drift.” You know, statements such as, “that will never happen because the way it is around here is, blah, blah….”  You’ve heard it before.

Well, just a few years ago, a customer asked to me to review the design for automation of their data quality process. The solution, while innovative, is not the part I want to tell you about. What made this a true breakthrough in my eyes was how he got the project accepted by the organization as a whole – from top executives on down.

The organization that sprouted this project idea is one I’ve worked with for nearly a decade.  It’s full of bright people with innovative ideas for everything, it seems. Trouble is they have problems getting the ideas accepted within their organization, and consequently, these ideas are not funded.  This is a common problem in many companies. However, one IT management individual stepped up and took responsibility for this latest idea and effectively championed the value to both the business and IT.

Inadequate communication had been at the core of the organization’s inability to get this sort of thing done in the past.  Not specifically the verbal or written kind of communication, but the ability to articulate, understand and balance the needs and personal agendas across all departments/groups and at all levels of the organization.

Getting IT and the business to understand each other was a major breakthrough for this organization. To make it happen required the individual to acknowledge these points:
•    The business owns the data (but they may not know it);
•    You have to speak business benefits to business people (and it is not as easy as most people think);
•    You have to build and maintain a relationship with your stakeholders through continuous coaching.

Once the above points were acknowledged, then it was possible to do the following in order:

1) Change the thinking within the management of both IT and the business to accept responsibility for the entire process related to the data. This was done by identifying the most important aspect of the proposed changes as it pertained to each stakeholder and obtain their buy-in that the changes were good for each of them.

2) Educate both sides on the benefits (often confused with advantages) of the solution. The advantage may be that a process takes less time to complete but the benefit is cost savings. Don’t confuse the two.

3) Coach the stakeholders to assist with communicating the benefits to their connections, thereby creating a sort of “tipping point” for the solution throughout the organization. It was critical to create a groundswell needed for large, high-cost, and/or radically new ideas as was the case with this project.

4) Convey passion about the benefits of the solution, in turn, making others passionate and supportive of the project and its stakeholders, fueling momentum for the project.

So, sometimes, it’s the means to the end that is most important.  The breakthrough really was the ability of my customer to put himself in other people’s shoes throughout the organization, and sell them on what is important about this data quality project to each of them, individually.  Then, articulate the expected benefits in a way that they all understand to get them to buy in.  It was important for him to listen to the stakeholders first, understand what was important to them, and learn how to talk their talk to articulate the benefits and results.  That’s a breakthrough that all companies should strive for in order to ensure their data management or data governance projects move forward.

Wasted Talent and Lost Opportunity: Are Your MBAs Just Sifting Through Data?

By Len Dubois, Sr. Vice President, Marketing & Sales Support, Harte Hanks Trillium Software

It always amazes and gratifies me when clients realize the depth, impact and true costs of poor data quality within their organizations. 

In a recent customer meeting, one of the data stewards was diagramming the architecture and lifecycle process of their information needs within an order-to-cash system.  During the conversation, the steward discussed the types of data that each step in the process required, for example; names and addresses of customers for sales, marketing and shipping applications – product information for billing, inventory control, distribution and shipping – credit and/or payment for the finance department, etc.

In each scenario, data points were added or eliminated based on the contextual need for information for each department.  Representatives from MDM, Data Integration, and SCM all added value to the conversation in terms of completeness, accuracy and structure.

It was at that point when one of the business analysts described what her life was like trying to rectify and assimilate information from the various reports she received in order to answer two primary questions from senior leadership:

• How much new revenue was attributed to new customers last quarter?

• How many new licenses/products were sold to existing customers last quarter? 

She explained that because each business function defined customer, product, and yes, even revenue differently, she spent three of every ten days collecting, assimilating and analyzing reports from across the organization, and aggregating the information within Excel spreadsheets and reports.  Then, and only then, does she feel prepared to provide answers to senior managers. 

This was a simplistic boiling down of the facts in the meeting, but it demonstrated to all that a business analyst (an MBA level employee), whose job it is to analyze data in order to provide information for decision-making, was spending 30 % of her time performing data quality checks on information that previously was thought to have been high quality data.

Like Sonny said in the movie A Bronx Tale, “Wasted talent.”   Imagine the costs and wasted opportunity associated with this manual approach to ensuring high data quality.  Never mind the fact that she is paid to strategize about business processes and how to optimize business results and performance, not sift through data looking for data quality problems.  And, realistically, with new data coming in all the time, the data set may never be accurate enough to maximize business results because the manual process is never-ending. 

I know I’m preaching to the choir here, but this is a real live situation involving a company that really did not understand the impact of data quality ON THE BUSINESS.  And unfortunately, this lack of communication between IT and business happens all the time.  There is so much opportunity to utilize data quality tools to solve real, live business problems in so many companies.  But you have to communicate across business and IT lines, constantly measure results, and ensure data is analyzed, cleansed and reported on in real time to answer the big business and revenue questions that senior management will ask you. 

How are you breaking down the barriers between IT and business users to eliminate wasted opportunity?  Would love to hear about your ideas and anecdotes.

When it comes to data quality, it’s the last mile that counts - That last mile is Operational Data Quality (ODQ)

By Michele Goetz, Director of Product Marketing, Trillium Software

Today a significant gap exists between how IT managers or employees, and that of line-of-business executives, view the overall quality of data, according to a recent study by Forbes Insights.  This chasm exists because there is a missing link between what data quality projects are designed to achieve vs. what the business expects them to achieve. 

In light of this, alignment of data quality projects needs to provide tangible business benefits that are tracked through key performance indicators rather than though data quality indicators alone.  Both IT and business users must understand that data quality success is intimately tied to business success.  The context and use of data is required when approaching data quality.

In the past, the easiest route for most companies was to approach data quality within the data warehouse, installing the black box that runs on a server tucked away in a back room. While effective for a specific task, this approach leads to a lack of enablement of data quality policies across the enterprise. 

In fact, we hear from customers that instead of improving their data quality services within their quality solutions, they turn to quality procedures that are implemented within ETL processes. This approach, in many cases, conflicts with policies within the original quality services causing further degradation rather than improvement.  Lastly, the approach is also entirely reactive, causing business processes to be less effective and efficient.

The solution is to employ Operational Data Quality (ODQ).   ODQ practices happen within a business process deploying data quality services at the point of data entry, and usage within an application and between applications.  There are a number of benefits in this approach: 

•    Data quality policies are designed and aligned to business process improvement and results, providing transparency and immediate understanding of the linkage between data quality and business goals;

•    Data stewardship and standards are maintained and remedied by data stakeholders that are most impacted by degradation, as well as having the most expertise on how to improve the data according to business needs;

•    Improvements are made as they happen and quality of data migrates throughout the enterprise, allowing improvement beyond the immediate process;

•    Success in one business process provides a framework and justification for investment across the enterprise in data quality initiatives;

•    Data Governance initiatives benefit through Operational Data Quality practices by addressing data quality policies within the context of a business process that already has roles, responsibilities, and accountability.

The quality of data is no longer just an aspect to satisfy executive level reporting and analysis.  Companies are increasingly relying on their data to make decisions within processes such as customer on-boarding and off-boarding, risk assessment, and supply chain management.  This requires increased certainty that the data available supports next steps in a process and decision to make these steps.  Rolling out ODQ within your business is imperative to your business success.