By Brynn Smith-Raska - October 27th, 2015

Article Event Banner

An interview with Ramkumar Ravichandran

Ramkumar Ravichandran is the Director of Analytics at Visa and a thought leader in building the strategic, effective use of data into your company's daily processes. We recently sat down with Ram to get his advice and hear his personal experiences with text analytics.

Brynn Smith-Raska: Is financing still a concern for analytics programs or is executive buy-in more of the norm?

Ramkumar Ravichandran: Executive Buy-in is critical for many reasons beyond financing. The support and push helps with ongoing investment, strategic deployment and constant reinvention. Now we have strongest executive support in history – they realize the value of Analytics and the impact it has on the bottom line. Executives today are able to read/understand insights better and are able to leverage it in their day-to-day and long term decision making. They have also started demanding more and more of Analytics leading to a virtuous cycle of demand → investment → return → demand. 

Brynn: Has data volume become too large? Can systems handle what’s there or is narrowing the focus difficult?

Ram: Data volume spiked: Not only has the volume spiked, but also the velocity and variety. It is only going up with the arrivals of more smart devices, better instrumentation on the front end, demand for more insights, etc.

However it has led to a problem of too much data (sometimes noise, too). With decreasing storage cost and potential “data science” opportunities, firms are operating on “Get it now, figure out analysis later” mode. This sometimes slows down the analysis, because lot of time is spent in getting to what is needed (data cleaning, preparation, etc.)

What is needed, then, is an outcome-focused, top-down approach. It should start with high-level strategy, the Key Performance Indicators to monitor performance against the strategy, and then move down to the various drivers to help in narrowing down the deluge and focusing on key initiatives.

Brynn: Are off-the-shelf solutions ready for the market or is in-house development still the way to go for proper text analytics?

Ram: It depends on the problem at hand – many of the common problems have been attacked well by the players and have good tools against them (KNIME, RapidMiner, Megaputer, etc.). However if it’s a nuanced problem, typically in-house development has to code it in R or Python.

Brynn: Do you still face challenges in hiring and forming an analytics team? Have skillsets changed as technology eases (automates) the creation of metrics and data modeling?

Ram: Text Analytics skills are shorter in supply when compared to other Analytics skills. However the real opportunity lies in training the Non-Text Analysts on Text Analytics skills – it’s a vast pool of trained resources with Analytical skill sets (Problem solving, Statistics, Programming and Presentation). Tech advancements have helped with increasing speed of turnaround, ability to solve a wider range of problems and presenting the findings in business-friendly ways. However the brains are still needed and that is where the cross-training opportunity lies.

Brynn: How does the customer come into play for text analytics?

Ram: Customer is front and center of all Analytics initiatives, more so for Text Analytics. Text Analytics helps firms understand Customer Feedback (Sentiment Analysis, Operational Reporting) , Connect with them (Brand Awareness, Share of Voice, NPS), Customer Needs (Product/Feature Research, Pain points) and Early Warning Systems (identifying customers at risk of attrition).

Brynn: Are they part of your BI and storytelling or is that separate?

Ram: NPS, Share of Voice, Brand Awareness, and Mentions have become business goals. Initiatives to improve these metrics are framed and monitored over the year.

Brynn: Are there any areas you’re excited about in TA?

Ram: There are a couple.

One is the tighter integration with other analytical methods: Leveraging insights from Text Analytics in association with other Analytical methods have gained steam and are being integrated closely.
Second, Natural Language Understanding in Analytics: PowerBI has revolutionized the way analytics is used by business community.

Brynn: What piece of advice would you give someone who is starting or expanding their TA program?

Ram: Please keep the insights simple, quantifiable and actionable. We have reservations about creating dollar- impact projections from Text Analytics for various good reasons (small sample size, self-selection bias of responders, etc.) However when the insights aren’t quantified, the business community will not be able to relate to it or prioritize it properly. And even more importantly, you should go as far as having actionable recommendations (what you think the business should do) vs. stopping at insights/impact. It is only then that the User will know what you want to achieve with these insights and be able act on that.  

Ram will be speaking at the upcoming Incite Text Analytics Summit: West, taking place November 5th and 6th in San Francisco. He’ll be joined by experts from BlackRock, Twitter, AARP, GE, HP and many, many more. Check out the full event program here: 

Incite Text Analytics Summit: West 2015

November 2015, Hotel Nikko, San Francisco

Learn to tie your data to business goals from over 25 of the brightest minds in text analytics. The 15th Text Analytics Summit: West will show you how.

comments powered by Disqus