Make an appointment

We are looking for a data analyst! Check the job posting.

Big data, means or ends?

Online Dialogue

Online Dialogue

19-09-2012 - minutes reading time

Our Online Dialogue colleague and Big Data expert Arend Zwaneveld has for Marketingfacts wrote an article on Big Data, is it a means, or the end?

Those who keep a close eye on trends have not failed to notice: Big Data is the solution to all our problems, is going to change our world, and not participating is the same as losing. Although exaggerated -and perhaps not fully true for another few years- the rise of big data is undeniable. Those who want to jump on board right away want to know the best way to jump. Those who do not plan to jump on board immediately would do well to at least familiarize themselves with the subject in order to distinguish hype from happening.

Definition of big data: unknown

What is big data? A real definition does not yet exist, but it is rarely short and sweet. Because most companies are a lot smaller than the famous success cases from the U.S., it is technically even questionable whether real ‘Big’ data is already available in the Netherlands.

Is big data actually the answer?

As an interested party, how do you know if Big Data is what you need? You could, of course, read through the list of Big Data success cases below and determine if any of them can be translated one-to-one to your own organization.

Those interested in Big Data would do well to ask themselves a number of questions first. There must be a need, preferably one for which a business case can be made.

After going through these and many other Big Data examples at length, I came up with a set of questions that reveal the opportunities for Big Data analytics within your organization. Answering these questions leads directly to Big Data ideas and need for Big Data:

  1. What if you could fully see individual customer behavior? How would your organization then be able to better serve these customers?
  2. What if I could solidify with data what I have suspected for ages (competition, trends, customer behavior, inefficient colleagues, etc...)?
  3. What if all “unstructured” information “on the hard disk” were structured into one database/table? (word documents, evaluation reports, HR progress meetings, minutes, project schedules, annual reports, etc...)
  4. What if the interest of the entire company becomes leading over that of departments?
  5. What if you didn't have to throw away any more data?
  6. What if the analyses are completed in time to do something with the results?
  7. What if I had 200 interns to read and summarize files on the intranet and Internet?
  8. What available data to improve processes, pricing, products and services are we currently underutilizing?

Whether the information-need that comes with answering these questions is actually ‘Big’ data is not actually interesting: Do you want to beat the competition, or be able to tell a competitor at a get-together next week that you are already doing Big Data!

The Roland Garros fan experience

There are now numerous examples of big data - some anecdotal, others described in detail - that can be used as inspiration when thinking of big data applications within one's own organization. One inspiring example is the fan experience of tennis tournament Roland Garros.
Roland Garros fan experience

The fan experience is a very comprehensive real-time match statistics tool for display at the bottom of the television screen or your second screen during the match: the statistics, metrics and odds provided function to complement the commentator/fan, who can directly conclude from the fan-meter how crucial the ‘hitting in’ of the next service is to the candidates' chances of victory.

During the match, the fan experience continuously calculates the probability of winning for each player. Changes in this real-time probability of winning make spectators immediately aware of the importance of the past, current and next rally for the course of the match. To calculate this information, the big data algorithm cleverly uses very diverse data sources, including those based on historical matches between the two opponents:

  • service rate
  • course of the match
  • nature of relevant decision moments (double faults, tie-breaks, etc...)
  • influence of weather (low sun, rain, rain break)
  • influence of type of surface (gravel, hard court, grass, etc...)

Final

In the early stages of last men's Roland Garros final, Nadal was ahead for a long time against his opponent Novak Djokovic. The fan experience statistics computer had already indicated, based on historical data, that as time passed, Nadal's chances of victory were visibly diminishing. When it started raining and the match was forced to be stopped, the fan experience meter immediately indicated that this rain break was to Nadal's disadvantage ... who eventually lost the match as well!

From data to big data

Have your own answers to the questions above and the practical examples convinced you of the usefulness of Big Data for your business? Then, of course, you'll want to start using Big Data tomorrow! The advantage is: many bright minds have already done extensive research on making organizations ‘data-driven,’ “Competing on Analytics,” structuring and quantifying improvement processes - e.g. Six Sigma, DMAIC, Deming Cycle, etc....

Based on their findings, they each designed their own variant of the original Six Sigma Capacity Maturity Model, those known to me in no particular order:

The only thing they have not yet incorporated into their models is the concept of “Big Data.” Why? Because this term has only been used since 2005 (Web 2.0) - 2008 (Big Data Computing). The term was only really launched by a 2011 McKinsey report: “Big data: The next frontier for innovation, competition, and productivity.”

The road to big data

An organization in which big data can be practiced and actually contribute to the bottom line does not emerge overnight. Before you as decide you want to become a ‘level 5 data-driven organization,’ it is important to realize that a big data culture clashes with the way organizations are organized and operate today:

Data doesn't lie

  • Say goodbye to intuition and gut feeling.
  • Data acts as a strategy consultant: the truth from the shop floor reaches the boardroom in a visible and quantified way.

Big data optimizes globally

  • Big data reaches across departments and silos.
  • Departmental interest becomes secondary to corporate interest: what will you reward managers on? Do you still need managers?

Big data = predictive analytics

  • As an organization, you must be culturally ready to hand over some of your decision-making power to a ‘self-optimizing algorithm in a black-box. An auto-pilot for companies, in other words.

For all these reasons, big data is an end stage that is reached automatically with the step-by-step progression of one of the many Capacity Maturity Models, not an off the shelf software product ordered today and installed tomorrow!

Therefore, the best advice for organizations looking to get started with big data is:

  • Start small.
  • Invest heavily in people rather than software.
  • Make (many) mistakes quickly.
  • Good luck!

Evening Symposium on Big Data

Are you interested in diving deeper into the practical possibilities of big data? During the Big Data evening symposium which Marketingfacts and the Nyenrode Marketing Society are jointly organizing on Sept. 27 a.s., Arend will elaborate on this matter in a presentation. Besides him, Jaap Bloem (VINT), Piet Daas (CBS) and Michiel Rozema (Microsoft) will also give their views on developments in this field.

For more information, see the full program of the evening or register directly. See you on the 27th!

Continue reading

Interested in learning more about the companies that are ahead in analyzing large amounts of (unstructured) data? Below is a list of 25 cases. It's not about big data for nothing!

 
Credits image: Rodrigo Moraes (CC)

Originally posted on September 14, 2012 at Marketingfacts.co.uk

Online Dialogue

Online Dialogue

You may also find these blogs interesting