Big Data’s “Access-Process” Loop


Our perspective on Analysis
The accepted first steps in any data analysis project begin with a business-worthy questions/purpose.

  • Start with a question
  • Form an hypothesis
  • Get hold of data
  • Run some numbers to prove or disprove the hypothesis.

Right? Maybe not.

In our experience, the highest value questions originate from

  • An understanding of the business fundamentals combined with
  • A free-form exploration of data.

Moreover the first set of questions which any business user might hope to answer using data are generally not the right questions to begin with. Rather they evolve as the data is explored more deeply.

To guide data analysis beyond initial hypothesis phase requires easy access to data and exploration tools. And this is where most data analysis efforts get bogged down before they can provide value – we at MoData call this the “Access-Process-Loop” Conundrum.

The Accepted Roles of Business and IT
Published best practice suggests that business and IT together are required to to solve this conundrum and deliver value: business brings the ideas and IT brings the data and tools.

The inception of analysis projects tends to originate from business processes that need help. Common wisdom is that the most effective analysis requires the deep context that only business users have.

The IT organization is portrayed as consumed with supporting the business with enabling and reliable technology, a full-time occupation simply keeping up and integrating the latest technological developments.

The generalization above happens to be the most common case, but our belief at MoData is that the role stereotypes have been designed into the organizational structure. The downside is that this detracts from the combined set of skills brought by the union of business and IT. The most game changing analysis is born from the melting pot of

  • Business context
  • Probing questions
  • Breadth of data sources
  • Analytic horsepower

Hopefully you now start to see the issues preventing the relevant, accurate and effective analysis of data which could have resulted into “actionable insights

Breaking the Access-Process-Loop Conundrum
As mentioned earlier, most initial hypotheses will be proven wrong in a good analysis project (mostly because of observational bias) the Access-Process-Loop needs to be very efficient both in terms of cost and time.

Deconstructing the Access-Process-Loop:

  • Accessing the relevant data (Access)
  • Performing initial analysis (Process)
  • Identifying initial insights
  • Enhancing the hypothesis
  • FINALLY going back to get more data (Loop).

The process continues till the point where either hypothesis is proven and a meaningful conclusion can be drawn OR its discarded.

An efficient Loop avails business user the opportunities to get involved in “data-driven” decision making and provides them a way to perform what-if analysis and quickly answer “I wonder ?” questions. An efficient Loop enables fact-based decision making critical for modern organizations to survive and prosper.

The MoData team has been hard at work to make Access-Process Loop super-efficient. Years of effort are finally coming to fruition and we have very exciting developments to share in coming weeks and months.

Reference: Image courtesy www.derrickbailey.com

Get The Latest Updates From Mo-Data Straight To Your Inbox