All Things Data

Data Understanding

In speaking with Steve LoCastro from BDS, he pointed out at an interesting phenomenon when we are engaged in a customer data integration application to develop a single customer view, data analysis, data quality, or other solution.

While there may be a technical person or a team of technical people who know the company’s data in detail, many times the business end-user who is asking for assistance does not know their company’s data with specificity.  They do not how certain pieces of information are represented or calculated in their databases.

Consequently, the business user will ask a question and the results may be unexpected.  Sometimes, after some research, it is discovered that although the results are correct based on the question, the question itself needs to be changed.  The question needs to be more specific and take into account the subtleness in their data in order to obtain the results desired.

Therefore, one of the great side benefits from a business user’s perspective in getting involved in a customer data integration, data analysis, or other related initiative is that for the first time they start to gain an understanding of the company’s data with some specificity.  Sometimes, this side benefit is simply learning about data that they did not know even existed.  This understanding enables the business user to start thinking about how they can use their data in ways that they had not thought of before leading to better analysis and decision making.

Marketing Analysis, Data Quality, and Customer Data Integration

As mentioned in my last post, you can get value out of your data systems even if you have bad data.  In fact, one of the values of a customer data integration application is that it can provide a single customer view of your data based on the data “As Is”.  This single view of the customer could be provided for just one data silo or it could be based on multiple data sources.

Of course, having good data quality is a great benefit for your organization for reasons beyond a customer data integration project and developing a single customer.

However, what is data quality?  As you can see by the following definition posted in Wikipedia, data quality can mean different things to different people.  “Data quality is the quality of data.  Data are of high quality "if they are fit for their intended uses in operations, decision making and planning."

Here are some sample marketing analysis data quality questions:

1. Understanding how many unique and duplicate organization and contact records
2. Understanding how many stand alone organization records and how many unique corporate families
3. Understanding how many unique locations

The marketing analysis data quality questions provide insight as to what is the potential within a client’s customer and prospect base and how much resource is needed to market to this potential.

The marketing analysis questions are very different from the more tactical data quality questions such as how many records have or a missing a certain data element, such as how many contact records are missing email addresses.

For most organizations, both of these types of data quality questions are important.  The marketing analysis data quality needs to be addressed in order to help with strategic planning.  The tactical data quality issues need to be addressed to assist the effort in carrying out the strategic plan.

CRM, Data Quality, and Customer Data Integration

Every customer relationship management system (CRM) has bad data.  Further, sales representatives do not want to be burdened with the responsibility of data hygiene.  They strictly want to focus on the selling effort.

However, even if you have poor data quality, this does not mean that your organization cannot utilize the data in the CRM system.  As long as you are getting some information from your representatives, such as company, contact, phone, and/or email, you have a start.  Using a customer data integration solution, you can match and link on company and contact data even if you have sketchy details.  You can correlate companies so that representatives can have a single customer view and marketing can run reports.  Then over time, you can address the data quality issues.  You do not have to wait to complete your data clean-up efforts in order to get value out of the data.

Insurance Direct Marketing Forum

Just a quick reminder, Bob Bloom, President of Business Development Solutions, shall be speaking about Customer Data Integration at The 6th Annual Insurance Direct Marketing Forum in Philadelphia on Monday, September 15, 2008.

So if you happen to be attending, please stop by and say hello to Bob. 

Customer Data Integration and Contact ID, Customer ID, Account ID, and Client ID

In speaking with Steve LoCastro from BDS, Steve pointed out that many companies have “Native Intelligence” in their data sources.  Native Intelligence is information in the data that indicates that separate records have some relationship with each other.  The native intelligence could indicate that the records are part of the same corporate family, indicate the hierarchy relationship of the records, or some other relationship.  Many times these relationships can be indicated in the data’s Contact ID, Customer ID, Account ID, and/or Client ID.  An external marketing list also might have data that indicates the relationships of separate records.

In developing a single customer view, Steve suggests that the data should be examined in order to find any information that could be used as native intelligence.  Further, the use of native intelligence is not something that should be looked as independent of a customer data integration solution.  Instead, a company’s native intelligence is maximized if it is incorporated into the customer data integration match. 

Combining CDI matching algorithms with native intelligence will enable records to be matched, linked, and coordinated by their corporate hierarchy relationships to develop a single view of the customer that could not be otherwise presented if these two processes were completely independent of each other.

Top 8 Steps to a Successful Customer Data Integration Solution

Here are the top 8 steps to a successful customer data integration solution:

1.  Top Management Commitment – Without top management being on-board, the project will go nowhere.

2.  Dedicated Project Leader – Someone who sets aside the time and has the skill to manage both internal resources and work with your external partner.

3.  Clear Goals – Define the objectives of the implementation.

4.  Phased Approach – Do not try to tackle everything one time.  Set priorities and work toward reaching your goals and then add additional goals once the initial ones are met.

5.  Set Time Line – Put stakes in the ground regarding expected deliveries.

6.  Project Flexibility – No matter how much you plan, there will always be things that come up that cannot be anticipated.  Be flexible enough to work through these issues.

7.  Experienced Partner – Work with someone who has been through the territory.

8.  Appropriate Tools – Use a Customer Data Integration software solution that is suitable for your goals.

Top 8 Reasons to Implement a Customer Data Integration System

Here are the top 8 reasons to implement a customer data integration system:

1.  Report and Data Mine Utilizing Separate Data Silos
2.  Develop a Single Customer View for Up-selling and Cross-selling
3.  Consolidate Multiple Databases After a Merger or Acquisition
4.  Correlate Data for Corporate Hierarchies
5.  Identify Data Quality Issues
6.  Discover Duplicate Records
7.  Append External Marketing Data to your CRM System
8.  Build a New Database System

Customer Data Integration Matching

I recently spoke with one of the members of our team, Steve LoCastro, about the different types of matching for a customer data integration application.  Steve is one of the SingleVision developers and also is a key member in working with our clients in developing their single customer view.

In speaking with Steve, he stated that in order to develop a complete single view of the customer, you need the ability to modify matching criteria based on the specifics of the data. 

For example, if you are matching 2 internal files, such as a CRM system and an accounting system, the match rules are going to be different than if one of the files included external information such as a marketing list.  When the files are fairly homogeneous, such as only internal files, you may be able to use fairly loose rules.  When external files are included in the mix, then you mostly likely will have to tighten your rules.  Otherwise, you will get false positives. 

The key to a successful customer data integration match is to be able to review the data, understand the objectives, and have the ability to craft the rules based on the data and objectives.