Big Data Lessons the East Can Take from Facebook, Wal-Mart and LinkedIn

by   |   October 2, 2013 2:47 pm   |   0 Comments

Venkat Viswanathan LatentView 228x228 150x150 Big Data Lessons the East Can Take from Facebook, Wal Mart and LinkedIn

Venkat Viswanathan of LatentView.

Like in the West, the number of big data initiatives in the East is quickly growing as more and more companies begin to understand its power. According to the Forrester Research report, “Big Data Adoption Trends in Asia Pacific: 2013 To 2014,” China and India tie as the leaders in big data adoption with 21 percent of respondents reporting they have implemented big data already. Indonesia has the highest percentage, 25 percent, of all countries in the region planning to implement big data soon. Respondents in Malaysia are at the rear of the pack with the highest number of respondents, 45 percent, reporting no big data plans at all. However, Forrester found that “most big data projects currently underway in the APAC region are small in scale or proof-of-concept projects.”

By comparison, North American technology vendors lead the big data revolution with 54 percent share of the market followed closely by Europe, according to Transparency Market Research. It’s easy to understand why that is. One glance around and you see giants like Google, Facebook and Twitter that had to work with big data early and figure out how to make it work.

Related Stories

Opinion: Avoiding pitfalls in predictive analytics models.
Read the story »

Podcast: Civic hackathon aims to enrich lives with data.
Read the story »

View from Mumbai on recruiting and training top analytics talent.
Read the story »

In awarding prize for analytics, Netflix failed to predict it wouldn’t be used.
Read the story »

As these pioneers came to master more of the work, ecosystems formed around their efforts which added value via collaboration. The ecosystems provide the opportunity for newer products to get access to big data and mature faster. A recent example is the integration of Bing search from Microsoft with Facebook. With more than 1 billion global users, Facebook provides a great opportunity for Bing to significantly scale the number of searchers using its engine, and is a natural ally to counter Google. And the relationship is symbolic of the U.S.-centric nature of such alliances in the big data world.

The East is currently missing the diverse and massive experience and the opportunities to collaborate that the West now enjoys. In our experience, very few companies in the East are able to leverage data from partner ecosystems, and tend to be quite local and internal data focused. That said, the increased use of cloud-based technologies is allowing companies in the East to access platforms like Amazon Web Services and Google Apps, and quickly narrow the gap in adopting technology that can make them globally competitive. By leveraging global talent and attracting global investors, some companies in the East may further bridge the ecosystem gap as well.   Moreover, many companies in the East are not bound to legacy IT network and other infrastructure systems like those in the West are, and can more easily leapfrog the adoption process.  They can also learn from the mistakes early adopters have made.

So what can the East learn from the West about big data?

1. Get the leaders on board.

Companies in the East are on the right course in starting with pilot projects that they can then use to secure C-level buy-in. In our own experience of working with companies in the East, while leaders see the potential for data analytics, they tend to start small, test the waters, are wary to bet large sums and would like to build a business case. One of the biggest obstacles to project success in the West is the lack of top management support and involvement. For each of the waves in IT investments around ERP, CRM, eBusiness in the last 20 years, executive management support and alignment has been an often cited as an obstacle. Make sure senior management is on board early before you start the big projects.

2. Don’t try to restructure the entire organization.

Follow the example of companies like Wal-Mart which created a subsidiary, Walmart Labs, to figure out how best to use big data and then how best to weave it into the existing company. In this way, innovation has free reign and change is far less disruptive to the main business.

3. Collaborate often and with members of diverse industries.

Follow the example of the Wal-Mart and Facebook collaboration where the two giants learn from the strengths of the other. As they are not competitors, they could share information freely and partner on initiatives that leveraged their combined strengths.

4. Remember that the effectiveness of predictive analytics is not about the math.

Maintaining a practical viewpoint on how much model accuracy is needed in an analytics program for it to be effective, rather than spending months in the pursuit of accuracy, is a very sensible way to gain business impact. Some big data innovators in the West have proved that sometimes prioritizing speed and business value over accuracy can help avoid the hunt for unicorns.  One example: Netflix demonstrated with its “Netflix Prize” challenge that within days of launching they could get significant improvement in their rating algorithm. More accurate, but only marginally better forecasts that beat the benchmarks came in almost a year later. That’s much too late.

5. Yes you can (and should) run with incomplete and/or imperfect data.

Companies in the East often feel that they don’t have enough clean and reliable data for them to initiate an analytics initiative, and in our experience tend to wait for major IT data warehouse implementations. The reality is many companies in the West have demonstrated with their “Hacker Day” approach to innovation that incomplete data could still lead to meaningful and effective initiatives that lead to positive business impact. Companies that are good examples of the Hacker Day approach include LinkedIn and Facebook. Both have had multiple features (such as the Facebook “Like” button, and LinkedIn’s “People You May Know” feature) emerge from these kinds of events.  This approach shows that it is important to make progress with what you have today and deliver business impact now rather than wait for the perfect data infrastructure to arrive.

7. Bring data closer to decision making.

The democratization of data that visualization has enabled clearly can bridge the gap between decision makers and data stewards. For effective adoption of big data analytics across the organization, and gaining sponsorship by senior people, it helps to get them closer to the action and help them see for themselves the business impact that is created. In the West, increasing adoption of “visual story telling” is pushing analytics closer to business decision makers. The increasing popularity of tools like Tableau, Spotfire, QlikView and Flavors addressing this need is a clear trend. In our experience, the East has yet to adopt this trend, and still operates in the hierarchical past where you would need to make report and data requests to IT, and then wait until they are delivered rather than make dynamic business decisions enabled by visual data.

These lessons are not just for companies in Asia, of course. Analytics organizations in the West could also learn from these lessons and improve their data analytics functions. There is ample opportunity for all. The key is learning from one another and collaborating at every turn. It is more important than ever to build strong partnerships.

Venkat Viswanathan is the CEO and founder of LatentView, a business analytics services firm. The company has offices in Princeton, N.J., Mumbai and Chennai, India.





d22d61d6 ff77 4565 867f 1a8c909bf112 Big Data Lessons the East Can Take from Facebook, Wal Mart and LinkedIn



Tags: ,

Post a Comment

Your email is never published nor shared. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>