Imagine a centralized marketing data mart containing a wealth of information on all of your customers’ preferences, hopes and dreams. Everyone in the organization could benefit: Operations would have a better handle on logistics; merchants could glean trends for optimizing purchasing; management would unerringly trend for the future; and marketers could analyze and optimize every contact with customers.
Many doubters still consider it fiction. Your task is to make the concept real. Attaining this is easier said than done. The truth is that legacy systems don’t talk to each other in real time, and a little secret that many data system providers may not let on is that their solutions only deliver information.
Understand technology shortfalls. Middleware will guide the transformation process and deliver data from doorstep to doorstep, but will not inject it into the previous or next system. Further work or customization is required to get that data into the heart of the next tool to allow that next system to understand what should be done with the information.
Custom development groups in most organizations often are too consumed with building or supporting system requests from the last century to deal with this issue and manage it in an ongoing manner. To further obfuscate this issue, outside vendors and consultants might be licking their chops while viewing your organization and project as a chance to pad their bottom line.
Hedging your bets. Though technology keeps advancing, we are in the nascent stages of mastering our marketing data collection and delivery systems. Remember that it has taken the automotive industry decades to improve the quality of their products.
Half-done efforts on any information-oriented system projects will result in a mess of data rather than a mass of data. Still, though caution is advised, progress is advocated. Yes, many CRM efforts fail, but no reward is without risk. So to start on the right path or to right the road your organization is on, the prudent course is to follow some basic tips and constantly measure your results.
Develop a data roadmap. Short-term and long-range planning should follow your understanding of your data. A thorough assessment of existing data and processes is crucial to bring data together into a cohesive, functional system. Even matching batch processes to each other could cause schisms in hygiene, much less the addition of real-time data feeds from sources such as call centers.
By developing a staged roadmap with tangible milestones, you can track progress and change course if necessary. Separate your roadmap into data, process and functionality. Within these three areas, gauge your needs by assessing your current state and compare it with your desired state. This gap analysis can be used to drive your project, your measurement of success and the division of your projects into milestones.
For each actionable area in defining the current state, the gap analysis and the future state, divide discrete efforts into milestone tasks and assign a measure for success. One might start by mapping out high-level existing data types within buckets like customer, transaction and product data. The next step would be to list ideal data types and locations.
Match your current data state to the future state in a spreadsheet organized to accommodate details such as current location and optimal destination, source systems and processes that affect the data. With this in place, one can determine the broader tasks required to move toward the future state. List current state processes that touch the data today and future state processes and touch points that will access or manipulate the data later on in the evolution.
These efforts should start to clump into logical groups that can be mapped into distinct projects. When forming this list of projects, assign a high-level mission statement to each. When this list is complete, you can map the groupings into a meta-project roadmap. Tie to this a quantifiable mission statement for the milestone such as “Creation of a cohesive, high-level data entity diagram” or “Outline of system functional requirements for brand managers.”
Where are your data now? Hygiene, security and privacy are major process issues that must be addressed when assessing current or future systems. Understanding and mapping these issues puts you in a better position to understand the implications to your data mart.
Laws keep changing and are becoming more restrictive. An understanding of current policies and how they affect data sources, processing and use will come in handy as system builds or reviews come into play.
As you map these issues, consider who is responsible for each component. Collecting their positions, roles and responsibilities for data naturally will start to build out a directory of users and owners. This, too, can be a huge effort as we are considering three aspects of data here: data, process and processed data. Each component obviously is interrelated, and all have their own legal implications.
Though the whole of mapping hygiene, security and privacy could require additional resources (or a dedicated full-time staff), at least addressing and understanding top-line owners and components may save you valuable time as potential issues arise. Try to focus on a few select questions such as “What happens to data as it flows from place to place?” “Who touches or has access to the data?” and “Who ultimately is the owner?”
By performing some of these baseline mapping and tracking efforts, you will create a measurable, quantifiable base to build your platform and technology. These tools can be used for periodic reference checks to ensure your road or building blocks have not shifted, and if they have, the tools will let you adjust and better ensure that you are on the path to success.
