Saturday, June 1, 2013

Big Analytics Starts Small

Becoming analytics driven is a big undertaking. But that doesn’t mean you need a big project. Big Data can start small. A pilot is a perfectly acceptable start and often the best way to gain momentum.

  1. Becoming analytics driven is more a state of mind than an activity. If that’s the case then the key thing is creating simple 'Eureka' experiences which change mindset by letting people experience in lots of small ways just how big difference it can make to performance if you know what your key measures are , know how you are doing on them each day and keep them in your face at all times. 
  1. Pilot programs are useful when they demonstrate tangible value. Take something small and manageable and test it within the organization. Two other key pieces are education and marketing the WIFM (what's in it for me). Sometimes a senior team member does not understand one of the key data analytics used by the company and won't be comfortable asking, so it can really pay off to have a quick one-on-one session to discuss the analytics and how they can benefit that executive. 
  1. Connect analytics projects with BPI-business process improvement project, to pick the right initiative either base on re-purposed business goal or low hanging fruit, to deliver the manageable result with visible metrics. Keep clear goal in mind, either to optimize business processes or to improve customer satisfaction, with manageable metrics and incentives to inspire the participation. 
  1. POC or POV helps in building that acceptance from Pilot to Full Scale purchase of software. Establish small pilot study / analysis to show the efficiency of the procedure / economics of purchasing the software that is where proving the value of analytics is required, as many times businesses are not ready to invest on big bang tools yet.     
  2. Cultivating the analytics culture is sometimes even more critical than project itself, Analytics culture or mindset is required among senior management team to accept the future. A lot of input data and lot of understanding is must. Then only it can show true output of a POC to impress the customer. Still, people are the key, not only the knowledgeable analysts, but also enthusiastic power users, the top-down leadership sponsorship, and supportive end customers 

  1. Start with Driver (Drivers) mattering the most for the company. In a perfect world, analytics
    should embrace the whole organization and should be built around the business drivers that matter the most for that company. In reality, if you're going to get greater acceptance by starting with just one driver and measuring some or all of the KPIs relative to that driver, that's the way to start. It’s important to learn as much as you can from small programs and use that knowledge in developing subsequent programs 
  1. Even starting with a proof-of-concept or a focused pilot, for the large organization, the BI framework should be blueprinted via the holistic view and long term perspective, to avoid the multiple BI platforms with inconsistency in the future, to ensure the effective data governance/ management. 
Developing and maintaining a culture that values "analytics" is a journey. It often starts from and grows with leaders consistently seeking fact-based recommendations and decisions. Leadership support and executive sponsorship are critical success factors to becoming an analytics-driven organization.


Post a Comment