All Things Data

Top Priorities: Growth and Business Intelligence (BI)

According to a recent Gartner study, CIOs see their Number One business priority for 2013 is to increase enterprise growth.  Along with growth, CIOs see their Number One technology priority being Analytics and Business Intelligence (BI).

Chances are you have similar priorities to harness and use your data for meaningful analysis to help guide and grow your organization.

The challenge is how to get this done and to do it within a reasonable budget. 

Our Blog, AllThingsData.com, focuses on practical approaches to harnessing and using data.  Take a look at the various posts.  If you would like more details on any of these ideas, please let us know.  We will be glad to help.  Feel free to contact us.

Also, over the next several posts we will outline additional practical steps for getting the most out of your data. 

Make sure to use your data to help you reach your goals.

Discover Data Quality

We are often asked to help with data quality issues. 

Data quality efforts are closely tied to important business needs.  Accurate, consistent, rationalized and coordinated data records are keys to a useful database, a database that can be employed to help reach business objectives.

One of the most obvious indicators of data issues is duplicate records.  This can include duplicate contact records, as well as duplicate organizational location records.  In addition to duplicates, there are several other dimensions of data quality that exists in a database.  These include the accuracy, completeness, timeliness, interconnectedness and usability of the data records.

To help our clients to get a full understanding of their data, we suggest undertaking a Data Discovery.  This involves

* identifying all the data sources that are needed

* profiling the data from all data sources, to identify data characteristics and issues

* determining what data actions are needed

* planning how to work with the data to meet the underlying business goals 

Today tools like SingleVision® are available to help undertake effective, practical and affordable Data Discovery.

Improve Access to Important Business Information

In our work helping clients to develop and use meaningful data for analysis and action, the issue often arises as to how to effectively distribute this insightful information to the various people within the organization.

Many organizations today are using Microsoft^®SharePoint^®for collaboration and communication.  However, many SharePoint implementations are not actively or enthusiastically used.  This is frequently because of the lack of useful information available through their SharePoint system.

We have seen how SharePoint implementations can be made more useful and productive with the addition of meaningful marketing and sales data.  Various people within an organization can have timely access to important information to help effectively do their work.  Permissions as to who has access to what data can be controlled.

We suggest that you should start practically in the delivery of data through your SharePoint system.  Start with basic libraries and reports.  Then, as you expand your SharePoint capabilities, walk down the path of adding Business Intelligence functions, including Dashboards, Scorecards, and Analytic Charts and Grids.  As you proceed, you will gain feedback from the users of the data.  You will also be able to expand your SharePoint knowledge and skills.

Furthermore, this approach will enable you to evaluate, determine and expand the back-end data that is needed by your users.  Often that data will be developed by matching, linking and correlating data from various sources utilizing a data integration tool like SingleVision^®.  That data might also be needed to be organized into an actionable Data Framework for efficient analysis and action.  The quality, accuracy, timeliness and availability of your data is of critical.

The key is to start practically distributing actionable information, gain feedback, build your data competency, extend your SharePoint skills, and then expand over time the depth and sophistication of the delivery of data.

Data Quality for All

We have submitted several posts about the importance of data quality.  Ensuring that your data is accurate, organized and accessible enables better analysis, strategic decisions and tactical actions.

Recently we attended Gartner’s Webinar called “Best Practices To Ensure BI Success Through High-Quality Data.”  Ted Friedman, Gartner VP Distinguished Analyst, did an excellent job in articulating the benefits of data quality initiatives, along with providing ideas for building organizational commitment for data quality.

A great deal of data quality discussion today focuses on how large organizations can organize to improve their data quality.  We know that smaller organizations have the same issues.  Disorganized and inaccurate data can impede proper analysis and hamper effective action steps for any organization.  Small and midsized businesses (SMBs) need to recognize how important data quality is to them.  As such, it is important for those organizations to properly scale their data quality programs to their needs.

SMBs need to commit to Data Quality.  Data quality programs can be managed in SMBs with practical tools like Microsoft Access, Excel, and SQL Server, along with a practical data integration tool like SingleVision®.

Link Related Organizations to Retain and Grow B2B and B2I Customers

The old sales adage is that “your best prospect is your current customer.”  Today this is especially the case, concerning organizations, and their contacts, that are related to your current customer.

It is important to identify within your databases the organizations and their contacts that are related to each other.  This will provide you with valuable sales retention and growth opportunities, including

• Widening your communication network within your customer organization
• Providing additional contacts to help retain existing business, as well as develop new business.
• Expanding your sales with your current customer
• Identifying cross selling and up selling opportunities
• Identifying organizations and contacts that can be approached with references from a related entity

The key here is to identify and link organizations and their contact records together, across all of your data sources, with Common IDs.  This way you will be able to view the data together.  Views should be available to the Sales and Service teams in their CRM systems for their efforts, along with Marketing for analysis and data mining.

Remember, your best prospect might be related to your current customer.

If you need assistance in linking related records together to retain and grow customers, consider using SingleVision®.  SingleVision will match and link records, along with assigning a Common ID to keep the records linked.

Measuring Marketing Outreach Success

In order to properly analyze and measure the success of a marketing campaign, the resulting transactions must be assigned to the outreach activity that is believed to likely to have influenced the action.

For the analysis to be efficient, the organization of both the outreach and the transactions data is important.  One example would be a mailing list, utilizing keycodes that represent the offer given to the contact in question.  Sometimes a respondent will include the keycode of the offer, often they will not.

Most frequently when a customer does go forward with the offer, be it attending a meeting or purchasing a product, the outreach record and the transaction record remain independent, with no straightforward link between the two.

To link transaction records to outreach records, some matching may be necessary.  The business rules for linking these records will vary in each case, but will often rely on the matching of the contact’s name in addition to some other piece of information to confirm that they are not different people with the same name.  The matching algorithm may give varying confidences on the records to be considered a good match.

For example, offer AB1 was mailed to 1,000 people, and, based on matching the mail/outreach file to the transaction file (specifically the transactions in question), 50-60 customers went forward, depending upon the match confidence level to be accepted.

The analysis may then be taken a step further to determine the demographic of the customers that went forward in contrast to those that did not.  This may affect what “type” of customer is mailed a certain offer, as some evidence may show which campaign is most effective for any given variety of contacts.

When marketing campaigns are properly analyzed, the campaigns that follow will benefit from the intelligence and the insight gained.  Over time campaigns will become more effective, garnering higher percentages of success, and saving time and money.

Foundation for Sucessful Data Analysis

In these challenging economic times there is a greater need for data analysis to understand what is going on and to determine what to do about it.

There are a plethora of data mining and business intelligence tools on the market today.  Even Microsoft’s SQL Server has business intelligence multi-dimensional data cubing and data mining algorithms built-in as part of its capabilities.

Still, while there are lots of great analysis tools available, it is a fundamental step that you need to first have your data properly organized to gain the benefit of these tools.

For example, if you are analyzing transactional data for multi-entity organizations, the first step is to identify and link all of the separate, but related entities and individuals that are part of the same organization in order to get the complete analysis of an organization.  Also, the data needs to be organized in an efficient manner, so that the analysis tools can work in an efficient manner as well.

Further, a continuously updated data framework is going to enable analysis on an on-going basis and not just a moment in time snapshot.  Without first organizing and coordinating the data, you either will be unable to perform the necessary analysis or your analysis will be incomplete and inaccurate regardless of the analysis tool.

Master Data Management Data Format Goals

After profiling the customer’s client files, the next step in a Master Data Management initiative is to profile the data at the record level.  

We profile the Master Data Management records by breaking them down to their most basic elements.  For example, take a look at elements for a sample record at my Alma Mater, Rutgers:

Institution – Rutgers University
Campus – New Brunswick
Unit – Business School
Department – Accounting

With this breakdown, we can begin the discussion with our client as to what are the criteria for Master Data Management match at each hierarchy level:
1.  How many hierarchy levels do we need?
2.  What are the definitions of each hierarchy level?

We deem Level 1 matches to be the most granular and the higher levels to indicate more overall relationships.  

Depending on the business goals, one client may choose the hierarchy levels as:
Level 1 – Rutgers University, New Brunswick
Level 2 – Rutgers University

Another client may need further levels in their Master Data Management match:

Level 1 – Rutgers University, New Brunswick, Business School, Accounting Department
Level 2 – Rutgers University, New Brunswick, Business School
Level 3 – Rutgers University, New Brunswick
Level 4 – Rutgers University

Profiling organization records for the MDM match helps with  building corporate hierarchies, developing uniformed reports, and properly organizing the data when building a new database or incorporating data into an existing database

Similarly, we also breakdown Contacts records for the Master Data Match by its elements such as Company Name, Name Elements, Address Elements, Title, and Email.  While Contacts do not have hierarchies, the breakdown helps with the other Master Data Management matching goals.

Multiple Representations of the Same Organization or Contact Within a Single Data Source

When we initially start a data integration project, most of our clients realize that they have data “all over the place” and are focused on the issues that are caused by the lack of integration of separate data silos.  However, one of the areas that are usually overlooked is the issues that are within the each separate independent data silo.

Typically, a client will have multiple representations of the same organization and same contact within a single data source.  These multiple data representations can be duplicate records, the same organization, but a different location, the same organization, same location, but a different contact, similar, but incomplete information in one or more of the records, and many other possibilities and combinations.

These multiple representations can cause an entity many issues.  A Gartner study says that poor quality customer data costs U.S. business an estimated $611 billion dollars a year in postage, printing, and staff overhead.  When you factor in the many other negatives of poor data quality besides mailing costs, then the costs are much higher.  For example, if I am a sales rep, am I looking at the complete history of communication with the customer or prospect when I am on a specific record in my CRM system?  If I am a marketing person, how do you properly analyze how many accounts and contacts that you have for a certain criteria if you are counting the same organization or contact more than once?  How do you avoid not just the expense but the potential alienation of the contact by not sending out the same communication multiple times to the same contact that is listed multiple times in your database?

The answer is a numbering system that can coordinate the multiple representations of the same organization and same contact so that you know that these multiple records are actually the same organization or the same contact.  Once you can coordinate the data, a single customer view can be set-up so that a representative can see the records that have been matched and linked by the numbering system.  Marketing can count the multiple representations of the same organization or the same contact as being only one organization or one contact for analysis purposes.  Further, marketing can now send just one communication to the one contact.