Of course projects are good, but the roadmap approach is usually better. The old saying “if you don’t care where you’re going, then any road will do” applies here.* Projects are good when you have an issue, a roadmap is key if you have a destination in mind.
If you want both longer-term financial certainty of your investments and to be sure of the benefits you can expect, then a Data Science Roadmap is the answer. A roadmap usually looks out over one to three years and sets out a broader agenda. You get a forward-looking view of:
- how to develop your data science capability,
- what the important deliverables and capabilities are along the way,
- what it might cost,
- what it needs to deliver in terms of benefits and ROI, and
- where and how to get started based on where you are now.
You might think of it as a data science program consisting of several data science individual projects.
We’ve followed companies that have well-developed data science roadmaps. Consistently, they have generated significantly more value, and transition faster into becoming data-driven enterprises with all the benefits that come along with that – being able to “know”, which leads to more profitable and more accurate decisions. Good data science, as we argue in our “Competing on Data” blog post, is all about knowing sooner, acting faster, and winning more.
One of the key values of a roadmap is that it can help you establish how to begin to treat data as a critical business asset – as something that grows in value and becomes more useful every year. How last year’s analysis is a foundation for this year’s analysis. With projects, everything turns into a one-off.
A really good roadmap will explore numerous areas, likely including the following (though the actual topics should be tailored to you):
- What questions you want to answer
- What outcomes you want to achieve
- How the answers you seek and the outcomes you seek will advance the business
- What data you actually have
- Where the data is stored and what you can reasonably get
- Data quality and what might need to be done with the data to make it useful to the business
- What data you might need over and above what exists
- Your existing data analytics skills
- The current state of infrastructure and source systems, security, applications, and tools to support analytics
- Current data governance practices
- Gaps in people, processes, and technology to fulfill your objectives
Collectively this will help to define the “destination.” We all know that data is increasingly becoming a competitive weapon. But most companies aren’t there yet – so you may not be very far behind. But data sophistication is a curve with accelerating benefits. You want to be on the curve, not behind it. Catching up on an accelerating benefit can be extraordinarily hard work. Once you have the roadmap, you will have a plan that pays off many times over – more so than planning in many other business areas. With a data science roadmap, risks are reduced, and objectives more often achieved.
Sometimes we simply help our clients to create the roadmap, and sometimes we help them to implement it. Other times we act as a trusted partner and advisor, helping to shepherd a program along.
If a roadmap sounds like it can help you, contact us.
* Lewis Carrol, in Alice’s Adventures in Wonderland, is the source of the famous quote “if you don’t care where you’re going, then any road will do.”
