Data Governance: Key to Data Management Cover Page

Data Governance: The Key to Data Management (Part 2), by Tom Marine

Part 2 in our popular four-part series by Tom Marine on how to set up a data governance initiative

Thanks for coming back for Part Two of our data governance series by Tom Marine. In Part One, we talked about setting up the data governance charter, which involves building your team, RASCI development, the maturity model, priority factoring and maintaining your DG momentum.

In this installment, we’re going to focus on the challenges and drivers that businesses need to understand to make data management changes and why that fits into data governance. Then, we’ll discuss capturing the current state, and why this can be an eye opening exercise.

Part 2: Process Improvements – Reasons to Change + Current State Capture

Data governance plays a key role in improving processes. When doing the groundwork of defining responsibility, the workflows will have appropriate reference. First, let’s review the main reasons that data management improvements typically are desired.

Reasons to Change – Marketing

  • Speed to market – getting products introduced and ready for sale faster.
  • Single Source of Truth – having one place to house data eliminates multiple points of entry, which means reducing the opportunity for errors.
  • Vendor Integration – Putting the onus on your vendors for providing content is an ongoing challenge for many manufacturers and distributors. By providing the tools they need with an intuitive interface, you create “endless aisle” opportunities. This helps in a “one stop shopping” scenario.
  • Content is King – Besides having direct input from vendors, which, in some cases, can be used without modification, generating content using a PIM is versatile for customer-specific, channel specific and other possible uses. Additionally, the nature of a central source of the truth allows for access remotely with both internal and possible outsourcing models.
  • Omnichannel Marketing – By being able to have specific targeted content (marketing copy, images, associated content), a PIM gives you the opportunity to have coordinated campaigns – not just multichannel, but omnichannel. “Omnichannel marketing is just multichannel marketing done correctly.” The coordination of content is typically the obstacle when developing campaigns – finding it, preparing it and delivering it. The PIM provides that structure for channel/market/customer distribution.
  • Data Syndication – On the back end of the data cycle is the delivery of data to retailers and other distributors. Unfortunately, those targets have unique requirements, whether it’s the data, the mapping or naming conventions. Therefore, having an agile system to deliver that data dramatically decreases the customization and custom development usually associated with these challenges.

Reasons to Change – Business

  • Speed of working – being able to work faster by eliminating redundancies.
  • Speed of responsibility – by defining accountability and responsibilities, proofing cycles and touches can be reduced.
  • Single Point of Entry – similar to the single source is having a single point of entry. But the key is to integrate and map those single points to the single source. For instance, an ERP may still be the point of entry for some product data, but it needs to also be in the single source of truth (PIM) if required. That should not be a redundancy, but an integration flow so that any entry in the ERP makes its way to the PIM without having to be touched.
  • Pricing Efficiencies – This is a common issue. These can come in the form of where prices are entered, how many times they’re entered, how they are calculated, tracking customized customer-specific pricing, etc.
  • Increase Productivity – Although some companies try to justify expenses by reducing overhead, the best use of productivity is to increase the output of the current resources. By giving your team the appropriate tools, you’re able to meet more customer/retailer requirements, create/deliver more products, and increase revenue through volume.
  • Improved Cross-Training of Resources – Although this is not a direct requirement, allowing different resources to be exposed to the entire process (sometimes for the first time) can create both respect and knowledge gains. Others won’t see their colleagues as mysteries and may have more comfort in how to deal with requirements that cross department borders.
  • Retain Leadership Position in the Market – Maybe the best question here is, if we don’t do it, will our competitors do it? Because, if they do, then they may suddenly become more formidable.
  • Improved Communication – Having everyone working together on a common cause brings people together. They typically have “ah hah” moments where they recognize their dependencies on others and develop respect, create accountability and improve relationships.

Current State Capture

You need to know where you are, before you can set your course for where you want to go. That’s why the creation of a current state diagram is critical. There are many ways to approach this – whether it’s building a large diagram on a white board to stimulate conversation, or sticky notes posted on a wall in the sequence they occur.

Here are some key points to follow as you build the current state:

  • Choose the appropriate time – Depending on where you might be or what you want to know and in how much detail, the appropriate time may vary. You may start out with only a small engagement, just to see how the process may fit and then work your way into a larger and more in-depth initiative. The factors may include how close you are to needing a data management solution, and how mature your company is … especially around cultural change.
  • “What” is the question – If there is one thing that will help this process go smoother, it’s the way questions are formed. When at all possible, shy away from the accusatory and defense-response ridden question of “Why?” You’ll only set back the project. Phrasing your question with a “What” will elicit open-ended dialog and answers based on the receiver feeling that you want their knowledge, not challenging it.
  • Typical Flow – Try to avoid the exceptions. Go after the big fish first. If you solve those, then the exceptions have a tendency to go away, or at least become easier to deal with.
  • Don’t Solve Problems – This needs alert monitoring. Remember, this is an exercise in where you are, not where you’re going. If you try to solve issues during this part of the engagement, then you’ll never get through the process and, sometimes, you’ll be playing a game of change / change back.
  • “Aha” Moments – However, there will be some obvious quick wins based on some “Aha” moments. Don’t discount them, but be careful that any interim solution is a no-brainer. Don’t try to do more than one of them at a time.
  • Parking Lot – If a topic arises that creates a lot of energy and the dialog doesn’t approach closure, then it’s time to move the topic off to the side. Create a “parking lot” for these items, with the intention of returning to them after further investigation or when the energy is not as intense. Be sure you do return to such topics to show accountability.
  • Business Requirements –Much like the parking lot, document business requirements for everyone to see and continue to build the requirements as you go. Don’t focus on these, just capture them. This will make the Requirements Gathering go much faster, and the subsequent RFP.

Example of a Current State Diagram

Example of a current state diagram

Developing Business Requirements

An important fact-finding mission should be around identifying the requirements based on the business case and business needs. Getting these business requirements done Right Up Front (RUF) will keep you aligned through the process and will keep you from veering off track. It will also create a cohesive team understanding of the direction of the initiative as a guideline to the end result.

Continue to ask the question – How will the data be used? You’ll have a head start if you’ve been capturing the answers during the Current State process.

Be careful not to base your requirements on a solution provider’s set of features. Solution providers have customers across many markets and industries. This can be a conundrum for an organization: Do you allow someone who doesn’t understand your business tell you how to build your system; or do you ignore their input because you “know” you are unique?

Well, both, frankly. Solution providers need to take the time to understand your business in order to develop the proper consulting approach; and prospects need to realize that a provider may have experience that is invaluable. The key: be open but firm.

Here are some basic requirements buckets to consider as you prepare for a data management / data governance initiative. Many of these will have been accomplished if the data management system is already implemented, but many times, those implementations don’t fully engage the DG group – therefore, leading to the possibility of not having fully developed processes and system utilization techniques. In many cases, these end up being governed by the DG team and incorporated into the DG Charter.

  • Taxonomies, Hierarchies and the Data Model – How will the data be structured in order to meet the business needs, not the technological functions? Do you need to have multiple hierarchies? Do you have multiple channels and markets? Will you have distributor syndication requirements?
  • Managing Classifications, SKUs and Attributes – What will make your SKUs more valuable and more efficient to prepare for market? What are your global attributes? What the attributes by classification? What are your attribute configurations (algorithmic, character validation)? What kind of attribute inheritances can be established?
  • Technical and Integration Points – How will all the data get knitted together? What and where are all the systems? What are all the operating systems and network protocols? What are the different data bases being used? What are the current integration tools/processes? Where does all the data get used (syndication)? Are there ways to perform mass data imports and exports? Will the infrastructure need adjustment based on requirements? How is administration and configuration handled for current access?
  • Data Quality Parameters – What are the limitations of having data that meets yours and your customers’ expectations? Is there a normalization process to manage multiple data input points? Are there field validations that will need to be adhered to or changed? Do you use Choice Lists to limit data entry? What kind of special characters are used in your data?
  • Workflows, Collaboration, Privileges and Alerts – Are there documented workflows and processes on how data travels during its usage? Are those processes optimized? Are the roles of the users clear and have they been trained to perform those roles in a particular fashion? Has system access been configure to mimic those roles? Are there system alerts set in order to understand when there are problems? Are there structured statuses to monitor and direct the data flow?
  • Asset Management – Where are all of the assets? What is the naming convention? Where are assets used? What are the different file types that are required (PDF, JPG, TIF, EPS, etc.)? What are the versions and renditions requirements? How are assets acquired? What kind of normalization is done? What kind of metadata is associated to the assets?
  • Marketing Support – What are your current channels? What are your current markets? Who are your current customers? What kinds of tools are used to coordinate synchronized releases of pertinent data through your multi-channels to create an omni-channel experience?
  • Security – How deep will privileges and security need to be taken? What are the different roles? Who are the different players? Where is the earliest point of entry? Where should the one-time entry be? What limitations should be placed on access to product groups, products, attributes and taxonomies? What are the rules around duplication of those?
  • Regulatory & Compliance – Are there specific regulatory requirements that need to be woven into privileges, workflows and alerts? Is country of origin and harmonized code maintenance required? What are country, state and local selling restrictions?
  • Implementation Interruptions – What, who, how, how much time and when do the solution providers intend to be able to implement your solution? Aligning those with your own challenges (busy times, resources, other projects) will need to be addressed. Also, do the solutions providers do their own, or do they use a third party implementation group? If so, what is the track record and what does that relationship between the implementer and the provider look like? Where is the responsibility and accountability? How many other projects do they have going on? What is the level of experience? This may be one of the most important requirements, since getting it right or wrong, can make or break an implementation. Dig deep into their histories and get references.
  • Maintenance & Support – What are your expectations? How do they align with the providers? How are break-fixes handled and in what kind of time frame? What is the roadmap of the solution? What kind of upgrade plans do they advise? What are the costs of upgrades? What is the cost of the yearly maintenance (typically a percentage of the sales of the software)? Are maintenance fees prorated on a fiscal/calendar year?

In our next installment, we’ll cover the phase that follows current state capture: developing and optimizing the future state. Unlike the current state, this can be a contentious exercise and may require a delicate touch and, most likely, third party involvement – or at least someone who has no skin in the game.

About Tom Marine

Tom Marine Head ShotTom is a Principal Consultant for MDM at Liaison Technologies. He is an accomplished executive with more than 30 years of marketing strategy integrated with technology. Tom strongly embraces cultural-training techniques. He has facilitated change in many different Data Management environments, from workflows in highly technical e-Commerce/legacy integrations to creative productivity planning in marketing-advertising-product management suites.

Tom’s technical areas of expertise include MDM/PIM, e-Commerce and Multichannel (Omnichannel) marketing with additional experience in data governance initiatives, data quality engagements, integration roadmaps, and taxonomy/hierarchy development. Having been a user, provider and consultant on MDM, Tom brings a 360-degree approach to the specific needs of technology solutions with marketing.

Tags: , , , , , , , , , , ,


  1. Data Governance: The Key to Data Management (Part 1), by Tom Marine | Hub Designs Magazine - 12/14/2015

    […] of a data governance initiative, and we provide guidance for creating the data governance charter. Part 2 focuses on reasons to change and capturing the current state. Part 3 delves into optimizing the […]

  2. Data Governance: The Key to Data Management (Part 3), by Tom Marine | Hub Designs Magazine - 01/26/2016

    […] series by Tom Marine. In Part One, we talked about setting up the data governance charter. In Part Two, we covered capturing the current state. In this installment, we take on the more difficult task of […]

  3. Data Governance: The Key to Data Management (Part 1), by Tom Marine | Hub Designs Magazine | Next Chapter Solution - 01/30/2016

    […] In the next installment (Part 2), we’ll talk about the challenges and drivers that businesses need to understand to make data management changes and how that fits into data governance. We’ll also talk about a fun exercise – capturing the current state. […]

  4. Data Governance: The Key to Data Management (Part 4), by Tom Marine | Hub Designs Magazine - 02/26/2016

    […] series by Tom Marine. In Part One, he talked about setting up the data governance charter. In Part Two, he covered capturing the current state. In Part Three, he looked at the difficult task of […]

%d bloggers like this: