Big Data Governance Needs Robust Business Process Management

A number of clients have been talking about the importance of aligning key business processes with their information governance programs as well as emerging initiatives around big data. It is almost like BPM, MDM, information governance, and big data exist in their own silos. In this blog, I will endeavor to provide a framework to align these discrete initiatives.

An organization is built around its business processes, so it makes sense to start there. My IBM colleague, Ken Jacquier @kenjacquier, introduced me to IBM BlueworksLive, a SaaS-based BPM offering:
https://www.blueworkslive.com/#!gettingStarted:overview

 Overview of the business process
Using BlueworksLive, I was able to create a very simple process “snippet” that describes the management of social media at a retailer with physical store locations. You can view this process below:

 Detailed business processes – milestones and activities
We discuss each milestone and activity in the process below:

1. Customer sets up RFID card
Retailers can now leverage RFID cards to enable their customers to interact with social media in the off-line world. In the past, customers would have to stop what they were doing, and use a smartphone to post a message on Facebook or Twitter.

1.1  Customer obtains an RFID card
The customer obtains an RFID card from the retailer

1.2  Customer links RFID card to Facebook page
The customer links the RFID card to their Facebook profile prior to the shopping experience.

1.3  Customer okays Facebook wall postings
The customer sets up their profile and allows the retailer to access their Facebook page and make short postings to document their experiences with the retailer.

2.  Customer completes store purchases
In this milestone, the customer completes her shopping experience at the retail store.

2.1  Customer places items in shopping basket
The customer places items in her shopping basket. Some of the merchandise may also contain RFID tags that help the retailer track inventory throughout the supply chain and on the store shelf.

2.2  Cashier asks customer for phone number
Some U.S. states allow retailers to ask for phone numbers at the point-of-sale. Retailers use this information to better understand their customers.

2.3  Cashier scans items
The cashier scans the merchandise in the customer’s shopping basket.

2.4  Cashier deactivates RFID merchandise tags
The cashier deactivates RFID-enabled merchandise at the point-of-sale. We discuss this topic further under big data governance policies.

 2.5  Customer scans RFID card
The customer scans their RFID card, which is the trigger for the retailer to make a posting on her Facebook wall.

2.6  Facebook: “I got 30% off shoes at XYZ”
The retailer posts a message to the customer’s Facebook wall that says: “I got 30 percent off shoes at XYZ.”

3.  Data aggregated in marketing data warehouse
The marketing department now pulls all the relevant content into the data warehouse for analytics.

3.1  Reverse append customers from phone numbers
Even if the customer pays in cash, marketing can do a “reverse append” to obtain the customer identifier based on her phone number. Marketing can then use the indentifier to add the customer’s purchases to her overall transaction history in the data warehouse.

3.2  Pull in data from Facebook friends
Marketing pulls down a list of the customer’s Facebook friends from her profile.

3.3  Marketing analyzes friends, locations, etc.Marketing uses information from the RFID card to understand the customer’s shopping behavior by location. Marketing can also determine if the customer’s Facebook postings resulted in incremental purchases by her friends.

 Mapping of big data governance policies to business processes
The next step is to map big data governance policies to the specific activities. We list the sequence number, the activity, and the associated big data governance policy below:

1.3  Customer okays Facebook wall postings
The retailer obtains informed consent from the customer to post messages to her Facebook wall and to use key information in her profile such as her friends.

2.2  Cashier asks customer for phone number
Marketing depends on high quality phone numbers to improve customer insight. Even if the customer pays in cash, marketing can do a “reverse append” to obtain the customer identifier based on their phone number. Marketing can then use the customer ID to add the customer’s purchases to their overall transaction history in the data warehouse. Store operations needs to train and incent the store associates to capture accurate phone numbers at the point of sale. Of course, there are several challenges with this approach. Some customers will decline to provide a phone number. In addition, some retailers have found that store associates try to meet their targets by entering the phone number of the store or a local hotel. Information governance needs to work with marketing and store operations to establish real-time validation rules to ensure that store associates enter appropriate phone numbers at the point of sale. It should be noted that phone numbers are not “big data” but they do affect the ability of marketing to associate social media to the customer profile.

2.4  Cashier deactivates RFID merchandise tags
RFID tags could potentially be used to profile and track individuals. Retailers who pass RFID tags on to customers without automatically deactivating or removing them at the checkout may unintentionally enable this risk.

3.1 Reverse append customers from phone numbers
This activity is closely tied to 2.2 above. The big data governance program also needs to establish metrics around the percentage of “poor” phone numbers, something that has a direct impact on the establishment of a customer identity that is important to marketing.

3.2  Pull in data from Facebook friends
Marketing needs to relate a person’s Facebook profile with their internal record. This may not always be easy. For example, a customer may have a friend named Susie Smith on Facebook. Marketing needs to determine if “Susie Smith” is related to the six instances of “Susan Smith” in their internal systems. This entity resolution may be even more difficult for Twitter because users often have cryptic handles. The big data governance team needs to establish policies to relate Facebook identities with internal customer profiles using attributes such as relationships, name, and address.

This example is for a simple process but the overall framework should also apply to more complex scenarios.

About sunilsoares

Data Governance Practitioner
This entry was posted in Uncategorized. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s