Expert Interview with Ajay Ohri on the Importance of Big Data
“It has been proven comprehensively that data-driven decisions will beat instinct-driven decisions,” notes Ajay Ohri of DecisionStats. Ajay was kind enough to answer our questions on his work and the importance of big data. He also shared some insights on success.
What is your professional background? How has it served you in your position at DecisionStats?
I have a degree in Mechanical Engineering from Delhi College of Engineering (DCE 2001). I then did my MBA from Indian Institute of Management, Lucknow (IIML 2003). Both these institutes are generally considered in the top ten institutes of their type in India - a sort of Indian Ivy League - with the added difference that there are about 10 times as many applicants because of India’s huge educated English-speaking population.
I also did a year of MS (Statistics) at University of Tennessee, Knoxville; but I dropped out to return to India for health and personal reasons. I picked up some graduate level computer science courses and graphic design courses from the Vols.
Regarding work life, I have worked in analytics since 2004. Since DecisionStats is basically an analytics startup in terms of content, training and consulting; each one of my educational institutes and my work experience has helped me to fine tune the quality of my delivery and provide me with a mental database of experience, contacts and networks I can leverage for professional reasons.
I founded DecisionStats.com in 2007, and in 2015 we have spun off its training and consulting division into DecisionStats.org.
I also wrote two books on statistical computing for Springer that are called R for Business Analytics and R for Cloud Computing.
Please tell us more about DecisionStats. How does it serve small businesses and entrepreneurs?
DecisionStats.com is basically a content and thought leadership forum where I explore tools, techniques, and strategies in analytics. I have interviewed 100+ people in Analytics Industry on DecisionStats.com
Small businesses and entrepreneurs can pick up the helpful tips, techniques and content for free. All content on DecisionStats.com is cc-by-sa.
I do training and consulting. Most of my clients have been small and medium size IT providers. For example, I helped with creating a social media analytics software for a small Noida-based IT Services company. Recently, I helped optimize R code for a spatial analytics startup (which is a part of a big media research company). I also helped a New Jersey-based staffing company with analytics and data science content creation for helping them pitch for such work.
I help small businesses leverage analytics and start up analytics services, but sometimes I also advise them on social media analytics, including blogging and leveraging social media since these are skills I picked up while running my analytics business.
DecisionStats is platform agnostic; we train in SAS and R and Python, even though I wrote two books in R. I just passionately love software, and I think no one software can fit everything the customer needs.
So you see, I do work a lot with small businesses because I am a small business myself.
Who should be visiting the DecisionStats website and why? What can they expect to find there?
People expecting to read about analytics should visit the DecisionStats website. They can expect a lot of original content created to help them in their analytics journey.
What makes big data so important and such a hot topic?
Big Data is exciting and important in terms of the opportunities it throws up. It is cheap to now store data in open source databases for both document and relational storage, and it is cheap to process it in terms of the cloud and internet; plus, you now have software like Hadoop stack, Spark, R and Python to analyze it.
In terms of affordability and democratizing access and ability to process data, we are living in unprecedented and interesting times. Having said that, there is still some time before the ecosystem evolves. Things like wearable computing, Internet of Things and open data for governance just enhance the appeal of Big Data.
Of course Big Data means Big Money. That is why it is hot.
How important do you think it is for companies to make data-driven decisions?
Companies don’t make decisions; people do. Numbers don’t tell lies; people do. There is a problem of agency conflict in terms of decision-making by employees. There is a problem of technology legacy and obsolescence in terms of in house technology teams. Feedback from the market and customers is important data. Opportunity costs in making decisions are rarely taken into accounting software.
Better data visualization in reporting, better awareness of what is out there in terms of affordable statistical computing, and a bit more investment in training about analytics and data-driven decision-making to employees can help companies get better returns with a much lower total cost of ownership of reporting and management information systems.
When DecisionStats was first starting out, how did you use data to help grow your business?
I used analytics data from Google Analytics to help plan out content. I found that people really liked my interviews. Then I found LinkedIn helped spread my content better than Facebook did. Lastly, I found the benefits of sticking to Google guidelines since two-thirds of a website traffic came from organic search engine views from Google.com
I also used location-specific data to help customize my content and to pitch to certain consulting clients. I used to have an IP address tracking software on my website (with a disclaimer in my privacy section) so I could know who is reading what on my website; and it helped me a bit in pitching consulting services. There was no need for anyone to fill out a registration form; I used the IP address lookup service from arin-net and others.
However, once the website crossed 100,000 views annually, I stopped worrying so much and just moved to Wordpress.com, gave up Google Analytics, and depended on Wordpress Stats. I also stopped using IP address tracking since I felt that that was intrusive.
Lastly, the interest in some posts like Business Analytics using R and Cloud Computing convinced me to write two books on these topics, since clearly people were searching for these topics and interested in knowing what I wanted to say.
What plans do you have for further growth now?
We have big plans for future growth! We want to create 100,000 hackers by 2018 in the fields of Big Data, Data Science and Network Security. How do we do that? We want to offer training, but on mobile platforms and in a gamified way including videos and interactivity using chatting. We also think knowing very good English is an unnecessary barrier to learning how to code; so we want to use slang, local languages, and languages that are ignored in the developing world like Haryanvi, Punjabi and those in Africa.
You need minimal English to write code in Python and R or Java and Hadoop, So why are the best MOOCs in English only?
Interactivity is needed. The low pass percentage of MOOCs and the high fees charged by training companies in teaching Hadoop or Python or R proves students need a human touch and customized mentoring, not just lectures.
Also, it needs to be like a game because students remember things that are fun. People say that the internet destroyed kids’ ability to read books. Then came Harry Potter.
We are raising money right and creating content.
What sort of feedback do you receive from other companies who have used your services? What sort of things would you like to communicate with businesses that you haven’t already?
I have got tremendous feedback in terms of companies that have used my services as a trainer and a consultant. Some of these relationships have served me with both repeat business as well as valuable insights to make my services better. The network effects of knowing many people in analytics have helped me to spot trends early on and leverage them.