Map Your Big Data Journey: Five Key Phases to Watch For
By Brian Kocsy on April 12, 2017
Share this article:
In this article, I’m going to guide you through every step of thebig data journey. You’ll know what to expect, where you should be and the benchmarks you should be looking for.
Why is this important?
It’s important to recognize that proper planning and focus can lead to successful outcomes. We’ve worked with countless clients over the years to ensure their big data efforts are successful , and along the way we’ve picked up some valuable insights that we’d like to share with you.
The Five Phases of the Big Data Journey
Big data isn’t simply a project. It will be a journey that involves many different projects to continuously gain more value from your big data. Over time, your big data journey will evolve through five different phases:
- Ad-hoc – the earliest phase where organizations experiment and learn about their big data needs.
- Opportunistic – the second phase where an organization starts to deliver value to the business, building their skills and knowledge.
- Repeatable – the phase where a company creates a replicable model for big data projects and starts to operationalize.
- Managed – a phase where big data analytics becomes a managed service that starts to spread across the organization.
- Optimized – where big data becomes a well-oiled machine, continuously delivering new projects and exponential value to the business.
As you move through these phases, the value your big data initiative to the business grows exponentially as your capabilities expand and ability to deliver becomes streamlined.
Leaps in Evolution
As with any evolutionary process, it’s not a simple, linear growth curve. Just like in life, your big data journey will have major leaps that will hurdle you past barriers to the next phase.
These leaps will require different strategy shifts in key areas of your big data initiative: tools, approaches, skills and interactions. Let’s examine the three major steps to evolving your big data initiative.
1. Deliver Initial Value
The first challenge you will face, is delivering actionable insights. This is critical to the success of your project. If you cannot derive actionable insights relatively quickly, you may find that your project is canceled.
Here are three ways you can make sure to deliver initial value:
- Focus at the department level. This will allow you concentrate your efforts on an individual use case and prevent scope creep – a major threat to big data projects. In addition, it will help you deliver very specific business value and ROI.
- Focus on the edges of the data . This is where analytics start to cross the borders between different applications or datasets. This is often where an initial set of business value can found and is an excellent place to start.
- Team up . Proving initial value requires close teamwork between the IT, analyst and business teams. Run a use case discovery workshop to find a valuable use case and align the teams. Define frequent interaction points to re-assure alignment along the way.
At the end of this process you should have an agile analytics process where you can begin showcasing more value to the business. You have also set the seeds of a Center of Practice (COP).
Want to learn more?
Subscribe to the Datameer blog today.
2. Putting Insights to Work
Once you have showcased some value, the business, and analyst and IT teams need to partner to operationalize your new found insights. With this, the business teams can continue to see value every day.
This requires taking your analytic processes and operationalizing them by hardening them and run them on regular intervals. In addition, you need to define how the data will be used, who will take action, and what outcomes you want to drive. For instance:
- If you have identified fraudulent transactions, what process and systems get notified address the problem?
- If you identified marketing offers for specific customers or accounts, how do you queue up the call center with this up-sell information?
- If you’re analyzing the results of A/B testing on a campaigns, who in marketing gets notified and what actions should be taken?
Having a strong relationship with your IT team is critical for this phase. If you involve them too late, you may find your plans backlogged.
3. Influence Change and Drive Adoption
Big data is more than just technology. It helps an organization see things in a whole new light – customers, operations, risk and more. But with this new vision comes organizational change. New ways of using data, interacting with customers and executing the business.
To drive further adoption, expand your CoP into a Center of Excellence (CoE). The CoE can be the hub of the big data initiative. You evolve to a CoE by:
- Ensuring you have documented and codified your processes so they can be applied to multiple areas of the business.
- Enlisting your key power users as champions and showcase them to display the the impact of the big data initiative.
- Facilitating scalable, formal training so any department who wants to adopt your big data initiative can do so.
As a result of this leap, the big data program becomes self-sustaining, enabling more power users and analysts to engage with the platform. Adoption will move to an enterprise scale
The big data journey isn’t easy. But reaching big data nirvana, getting to that state where your big data program hums along is a worthy goal.
Don’t worry, there’s still work to be done. Of course, optimization should always be part of any stage of the big data journey . There will always be new tools and features to add value to your big data program. There will always be new ways to use big data to inform your decisions.
It may seem like you’ve reached the end, but you’re only getting to the best part .
Brian Kocsy leads the professional services and support teams to ensure Datameer’s customer success. His unique skill set spans web scale data, search, yield optimization, advertising and fraud.
Share this article: