Prototype! Don’t PowerPoint!
Drawing of Wright brothers’ first prototype of an airplane.
Strategy without action is the slowest possible route to a win. Yet how often in business do we confine strategy to the pages of a Microsoft Office document?
We also know that you never get it right first time either; strategy evolves with time and circumstance. First and foremost, strategy is a process. Therefore, it can (and should) be layered up over time, allowing for course correction as you progress.
Prototyping is a tool for progressively increasing the certainty of outcome for a particular course of action – reducing the final cost and refining the overall end quality. In a product-led world, an organisation would produce a succession of prototypes to test the idea, gain insights and then improve the next version until the concept’s value becomes a clearer reality or it is discarded.
We know that the same principles behind prototyping can, and should, be applied to strategy definition and application. Often, strategy remains theoretical and never moves into action. Or moves into action too quickly and wastes millions of pounds pursuing the wrong thing.
Data strategy lends itself very well to the principles of prototyping
Provided you have clear alignment on the direction of the overarching business strategy, a vision and value proposition for where you want your data capability to be in the future are good enough for the data teams to get started.
The core components of good data strategy will cover processes, models, tooling, skills, org structures and more. Each of these components is likely to be well-defined within their teams but the key to realising the value proposition lies in the cross-team engagement and in the dependency management between activity.
It is very easy for data teams to become focussed on increasing maturity of vertical, functional competencies, often driven from some form of data maturity assessment. Just as easy is to be seduced by the latest data technologies, often these are seen as fast-acting antidotes to the problems of the day. The reality, as we all know, is more complicated. These two tendencies repeatedly ignore the horizontal integrations, resulting in incomplete solutions or broken process dependencies down the line.
“Treating the implementation of data-defined business value goals as iterative prototypes will help turn concepts into reality very quickly”
Defining measurable business value supports strategy prototyping
By starting with a definition of the business value the data teams are targeting we can begin to break these tendencies apart. Taking the business value realisation cases down to the next level into a series of goals with data points to measure goal progress against value achieved, you can dramatically improve horizontal integration.
This happens because you are dealing with what matters to the organisation as a whole, not as a discrete set of functional issues. Data projects can then be prioritised by highest business impact and value delivery. It also enables an assessment of the data organisation’s ability to deliver the business value rather than a maturity assessment against an arbitrary set of industry competencies.
Treating the implementation of data-defined business value goals as iterative prototypes will help turn concepts into reality very quickly. If something isn’t working then fail fast and move on to the next area. If you need to rapidly test and develop an uncertain piece of metadata software and business process, then do it on a small scale before committing to the full project.
These principles have already benefited our customers many times over and are particularly relevant to the Chief Data Office function.
Remember, strategy is a process and can be tested through prototyping. A few early models or initiatives will help you quickly learn if you’re going in the right direction or if you need to consider alternative courses. A static, two-hundred-page long document won’t!
Besides, we’re not very good at PowerPoint round here…