I debated over whether to title this article Data Governance as a Puzzle … or Data Governance is a Puzzle. Both titles seemed to make good sense as something to write about.
I selected the first option and decided to use this article to provide a comparison of data governance and good puzzles; rather than describing the puzzle (defined as a problem designed to test ingenuity or knowledge in Oxford) that IS data governance, leading to many organizations’ programs resulting in limited success.
Recently, I came across a light-hearted article written by Eleanor Robinson – founder of the gaming company 7-128, that outlined the characteristics of a good puzzle. Her article got me to thinking about the similarities between data governance and a “good” puzzle and how this comparison might be an interesting subject to write about. I will use Robinson’s characteristics of a good puzzle as the basis of laying out data governance as a puzzle.
Characteristics of a Good Puzzle (and Data Governance)
Robinson started her article quickly by focusing on the characteristics of puzzle games that she plays often versus the characteristics of puzzles that she plays only a few times or never returns to play. This is a great analogy for data governance in itself as most organizations want their stewards to continue playing (governing data) rather than leaving and never returning.
Many of the characteristics she defined apply directly to the success of data governance while other may require a stretch of the imagination (or lengthy practical experience) to draw the comparison.
These are her characteristics of a good puzzle:
Good Puzzle Characteristic: Game must have re-playability. (Robinson)
Data governance must be re-playable, or should I say, the actions of governing data must be repeatable. The data must serve multiple purposes (and thus be re-usable). The roles must be reusable, governed processes must be repeatable, communications consistent and repeatable, metrics and tools re-playable as well. In fact, just like in a puzzle, organizations improve the value, efficiency and effectiveness of their governance activities through repetition.
Verdict: Based on this characteristic, data governance is a puzzle.
Good Puzzle Characteristic: The puzzle must be engaging enough that you lose track of time and what else is happening around you for at least brief periods when you are playing it. (Robinson)
There are several statements that I make about data governance repeatedly. One statement is that the only way to implement an effective program is through the activation of the data stewards. I have been known to say that “everybody is a data steward” and that organizations must find a way to get past that fact and deal with it.
Data governance programs that demonstrate success to their organizations activate and engage their data stewards, help the stewards to recognize themselves as data stewards, and engage data stewards where they “touch” the data. Just like in a puzzle, finding the most effective way to engage is utilize the data stewards to test ingenuity to solve problems and address opportunities.
Verdict: A key characteristic of a good puzzle is how the puzzle engages you. A key characteristic of a data governance program how you will engage the stewards and the rest of the organization. Data governance sounds like a good puzzle to me.
Requires Strategic Planning
Good Puzzle Characteristic: The puzzle must include some capability to do strategic planning, to plan ahead and modify the outcome. Pure chance games lose their luster rapidly, no matter how pretty they are. (Robinson)
Having a plan for your data governance effort assures that your activities are focused on achieving your target state. Instead of taking a “ready, fire, aim” approach where you shoot before you know your target, I always suggest aiming the data governance program at specific targeted activities. Often, the target activities are determined through an assessment or the critical organizational analysis against industry best practice.
Superior data governance plans are often agile, with the focus on being efficient, effective and able to modify the outcome of your program based on demonstrated success (or lack thereof). The ability to focus (and refocus) your data governance program based on planning ahead better prepares the program to react when the puzzle of data governance gets complicated or threatened.
Verdict: Although a good puzzle provides some capability to do strategic planning, data governance requires the ability to plan and adjust to modify the outcome of your program. This puzzle characteristic aligns with data governance from the perspective that both require advanced planning and the ability to adjust actions to improve outcomes.
Good Puzzle Characteristic: There must be a time factor, either as a countdown clock or as a reward for faster play. That being said, the time must be adjustable for different skill, play mode, or ability levels. (Robinson)
Data governance programs are not typically timed or on-the-clock, so to speak. However, program leaders are not given infinite time to demonstrate value to the organization or to their leadership. The value from a data governance program comes from improvements in data definition, production and usage that require planning, execution and measurement.
Organizations often measure program success in two ways. The first way focuses on measuring the business value that comes from improved governance of data. This method of measuring program success takes time. The second way of demonstrating program success results from measuring how well, or the rate at which, the program is being accepted and engaged by the enterprise. The second way results in measurements that can show improvements in governance in a shorter period of time. The clock is always ticking. Plan to demonstrate success quickly and often to appease management’s focus on the time factor.
The next characteristic focuses on improvement through repetition. Repetition leads to improved skills, and entrance into different data governance puzzle play modes (based on the organization’s maturity level).
Verdict: The good puzzle characteristic of the time factor impacts every data governance program when demonstrable success must come quickly. Hopefully you will not reach the end of the time factor (5-4-3-2-1-done) before providing demonstrable reward to your organization.
Improvement Through Repetition
Good Puzzle Characteristic: Increase in skills through repetition should result in achieving higher scores, reaching higher levels and solving more difficult puzzles. Which means it should not be solvable all the way through on the first try, but improving skill, not chance, should result in increased success. (Robinson)
Maturity comes through experience. Many published maturity models lay out a progression from an initial level of maturity to a defined level, then a repetitive level, before becoming managed and ultimately optimized. Organizations that plan for data governance success learn from their experience and make improvements in how they govern data.
When solving a good puzzle, players improve in their ability through the experience of advancements and setbacks they experience when attempting to solve the puzzle. The same can be said for data governance. Organizations learn by doing. Organizations mature at governing data over time. The holy grail of data governance is optimization though, just like being able to solve a puzzle is the ultimate result.
Verdict: Data governance programs are good puzzles because organizations show improvement through repetition. Measurable success does not always come quickly but organizations can learn from experience and improve at the same time.
Easy Early Success
Good Puzzle Characteristic: You must have some success on the first try. This means it must be easy enough at first for everyone to achieve at least the first several screens at the easiest setting. (Robinson)
The expression “reach for low-hanging fruit” means that organizations should look to address opportunities that provide value quickly and without the most complex level of execution. Low hanging fruit means success in something that adds true value to the business without a prolonged period of execution.
Data governance programs that demonstrate success early are often given the chance to continue to demonstrate success. Shoot for the streetlights before you shoot for the stars. The streetlights are within reach on the first try while the achieving the stars require planning, execution and measurement – and success will not come quickly.
Verdict: According to this characteristic, data governance is definitely a good puzzle. Success must come with some challenge although the demonstration of early success entices organizations to keep working on the puzzle.
Definable Levels of Success
Good Puzzle Characteristic: If it has levels, it should have variability between levels to add interest. Just making it faster or increasing numbers is NOT enough. This may include changed playing fields, addition of new hazards, changed rules of play and new graphic types. (Robinson)
As I mentioned in 5. Improvement Through Repetition above, there are maturity levels that organizations use to define their success in data governance.
Organizations start at the Initial level when they are just getting started. Organizations move to the defined level when they record and formalize the components of their program. Organizations achieve a repeatable level of success when the components they defined prove to be successful. Organizations achieve the managed level when the repeatable actions they take demonstrate value. And organization reach the optimized pinnacle level when they are constantly improving their data governance score.
These defined levels of success are used widely when organizations conduct data governance readiness assessments.
Verdict: Data governance is a good puzzle according to this characteristic because organizations that focus on continuous improvement often get better at demonstrating levels of success over time.
Ability to Break Through Barriers
Good Puzzle Characteristic: If chance locks out the possibility of solution, there should be something – a bonus gained previously or some item of value that can break through the lock-out at least sometimes. Otherwise, it becomes boring if you always fail at a particular level or pattern. The use of a “bomb” is an example of such an item of value. But these objects should not be too available or else it will become boring if you always can have something to use to win the game. You need to fail occasionally. (Robinson)
This may be the characteristic that only slightly can be used to assess whether or not data governance is a good puzzle. Data governance programs reach obstacles and barriers to success all the time. Whether it is a change in leadership, a change in resources, a change in organizational focus, or any other change to the organization, typically the resources (for example) of a data governance program are impacted.
These barriers to success can be considered puzzle roadblocks, or those things that cause you take a step back and re-evaluate the approach you are taking to solve the puzzle – or implement your data governance program. Barriers are common. Some people may even say that the foreseeing the barriers of success, and tacking those problems head-on, are what make the life of a data governance program administrator (puzzle solver) an exciting challenge.
Verdict: Oh yes, data governance programs like good puzzles, face barriers all the time. Successful program leaders demonstrate success in addressing those barriers continuously making their program-oriented efforts “interesting” and challenging.
Components and the Approach
Good Puzzle Characteristic: There should be a large variety of different piece arrangements that occur at random when you start the game. The same setup should not always appear. One of the factors that make a game of cards so appealing is that basically no two layouts or hands are the same. (Robinson)
I utilize a Data Governance Framework that I developed years ago and have enhanced over time. The framework consists of six core components (data, roles, processes, communications, metrics and tools) viewed from five organizational levels (executive, strategic, tactical, operational and support) to demonstrate all of the pieces of an effective data governance program. I consider these to be the pieces of the complete data governance puzzle.
Organizations typically do not focus on the entire framework at one time. They work on a specific component or two to start, improve by learning from experience and improve in their maturity, before they move on and attempt to tackle the next component. If you fit all the piece (components) together over time, your chance of solving the puzzle (demonstrating data governance program success) increases dramatically.
Verdict: Data governance is a puzzle according to this characteristic because it has several components that can be improved through experience and can be completed incrementally.
In this article I have compared data governance to the characteristics of a good puzzle. I debated writing about the puzzling and challenging aspects of data governance (of which there are many) but decided instead to write about how data governance and puzzles share features.
There are many characteristics of good puzzles that are also characteristics of effective data governance programs. I hope that the comparison I drew between puzzles and data governance made sense and are helpful when you consider the work that is necessary to implement data governance for your organization.
Next time, perhaps, I will write about the puzzle that is data governance. We all know that data governance can be complex and challenging, and cause restarts and improvements through experience. Data governance is just like a puzzle. Enjoy solving data governance!