How Siddharth Dayama is Taking His Data Viz Show on the Road
I recently had the opportunity to discuss data visualization with a former colleague, Siddharth Dayama, or Sid for short, Senior Manager at Verizon. Sid leads a team of analysts in building dashboard solutions that democratizes data in an intuitive and visual form for business leaders to consume. Sid’s artistic nature coupled with his unique perspective has given him the opportunity to present at the Tableau Conference in both 2018 and 2019. His advice and journey is inspiring as he hopes it will help both novices and veterans in data visualization.
Getting Down to Business
Allen Hillery: Hey Sid! Good catching up with you! Can you give me an elevator pitch on what your team does and how Visual Analytics helps Verizon’s call center operations?
Sid Dayama: I lead the Visual Analytics team in Customer Service Operations in Verizon where we build end-to-end business intelligence solutions that democratize and distribute data in an intuitive and visual form. In turn this enables business experts to slice and dice millions of rows themselves without writing code, extract valuable and actionable insights at will, on a near real-time basis. Through our dashboards we leverage data driven decision-making in the business to optimize operations, reduce costs and enhance the customer experience.
AH: How have you been able to add insights to the business using data visualization?
SD: We’ve had tremendous success in the following areas:
Optimizing Call Center Operations
We analyze Key Performance Indicators (KPIs) such as Call Volume, Handle Time, Call Drivers, Repeat Rates, Call sentiment etc to reduce the need for customers to call us repeatedly. Certain cohorts of customers are calling us 43% less after implementing the data driven solutions we’ve enabled. Our dashboards assist in workforce planning and call routing to get the right call to the right agent.
Optimizing Field Technician Dispatches
We have applied cutting-edge geo-spatial mapping techniques using Mapbox satellite maps and Google street view API’s from Google Developers to survey a geographical distribution of our dispatch rates. This has helped stakeholders in the Dispatch Research Center (DRC) to reduce dispatches by 62% for certain cohorts of customers.
Taking the Show on the Road
AH: I had a chance to check out BOTH of your Tableau Conference talks. What was that experience like?
SD: Absolutely phenomenal! Speaking at TC18 and TC19 is definitely one of the highlights of my professional career. Representing a Fortune 15 company at the biggest conference in the Business Intelligence space and speaking in front of thousands of data enthusiasts was an experience like no other. Getting recognized as one of the top 5 customer speakers was reassuring and encouraging. The Tableau Conference is the best platform to celebrate big-data work.
AH: What was your favorite moment at the Tableau conferences that really impacted you?
SD: Oh, there were so many, but the one that gave me goosebumps was watching my dashboards on the massive LED screens. They looked even more spectacular than I could imagine. It was a super proud moment for the artist in me to present his work to the world in this fashion. I would say this one was the most satisfying and impactful.
Apart from that, Tableau conference is the epicenter for networking in big-data. It was awesome to meet and share ideas with folks from different industry verticals and companies trying to solve the same kind of complex business problems using similar tools and techniques. Amazing how we all face similar challenges each day! IronViz is so inspiring! It is the world championship of data visualization and the showcased dashboards just reinforce how powerful data vizzes can be in storytelling.
AH: I liked your lessons learned that you outlined in the Tableau Conference. Can you comment briefly on each below and give advice to our readers from what you’ve learned.
SD: Sure! Here’s a takeaway of my lessons learned from building analytics dashboards:
Know Your Audience
As a dashboard developer this one really resonates! I can’t stress enough that it is absolutely imperative to not only know who the end client is but also how they are going to use the dashboard to extract information. This helps (the developer) to incorporate the right types of slicing and dicing functionalities in the view. It also mitigates over-engineering visuals while answering simple key questions.
Flow of Information
Most effective dashboards are the ones that tell a story. They guide the user towards data-driven decisions that they can act on. The most relevant information should be easily accessible. For some analyses it may make sense to show pure volumes first and then the derivative metrics such as rates and percentages, while other times the opposite would be more appropriate.
Minimize UI Cognitive Load
We think about this a lot and use certain neuro-aesthetic principles with respect to color pallets and canvas real estate usage to ensure information is consumed intuitively and is retained longer (as seen below). Always consider accessibility and the right color pallets to accommodate stakeholder needs.
It is important to structure data right at the onset of development. Optimized queries enable scalability and promote an exemplary customer experience. Unfortunately, this foundational step is often overlooked due to deadline pressures, which as time passes, becomes a detriment due to the sheer accumulation of data.
Anticipate the Next 3 Questions
Once you know your audience’s comfort level with analyzing data and the KPI’s they’re looking to measure, you can anticipate what questions they are looking to answer. We can showcase those insights proactively.
Annotations with tool-tips, for example enable executives to gain a deeper understanding of trends on a primary graph. The pop-up feature allows them to hover over the trend to answer why or when something occurred.
Evangelize and Train
This is especially necessary in companies that have not yet fully adapted to a data driven culture. It is important to convey the tremendous value automated dashboards bring to the table by democratizing data, removing redundancy and empowering stakeholders at all levels to drive change in the organization.
A Journey of a Thousand Miles
AH: You’ve had an interesting career journey working for the Women’s Tennis Association (WTA), hedge funds and now Verizon? Can you step the audience through that journey and how they all tie in with forming your career today?
SD: Sure! My non-linear career path illustrates a gravitational pull towards working with numbers in one way or another.
I have an academic background in finance with a specialization and Masters in Entertainment Finance and Music Business from New York University. I learned the nuances of valuation of intellectual property, Hollywood accounting, copyrights & licensing, etc. During that time I was teaching financial accounting to undergrads at NYU-Stern. It helped me build a strong foundation of corporate finance principles with respect to analyzing business problems.
Fresh out of grad school, my first job was at a hedge fund analyzing securities in the media & entertainment space to identify suitable investment positions. After about a year at the hedge fund, I joined as a first year employee in a start-up that later grew and merged to form WTA Networks. We built mobile apps and generated exclusive behind-the-scenes content for some of the world’s best tennis players. As the Head of Finance & Analytics, I worked with some exclusive athletes & brands including Maria Sharapova, Novak Djokovic, WWE, Porsche and Nike on monetizing their fan base using Web & Mobile Analytics. I also led the acquisition due diligence of GamerDNA, the second largest gaming information network at the time.
AH: Sounds like you were exposed to a lot of amazing opportunities at WTA. What did you take away from that experience?
SD: It is during my work at WTA Networks that I truly understood the power of big-data and fell in love with analytics and data science. In an early growth stage environment, we employed a data-centric methodology to acquire users, increase subscribers and engagement on our apps. This experience ignited my mission to further develop skills in analytics and data science.
AH: Let’s dive into the next chapter post WTA.
SD: After the WTA merger, I was hungry for a company with a mature big-data tech stack and large amounts of data. When a recruiter reached out, about a job at Verizon, a tier 1 ISP powering a large amount of the world’s Internet traffic with 118M subscribers, I jumped on the opportunity.
I feel these collective experiences have culminated in my current adventures in data viz. My corporate finance background has equipped me with a lens to measure the impact of analytics projects. Sometimes, the impact is the top line, increasing revenues or the bottom line by reducing expenses.
AH: What commonalities and distinct differences do you see data-wise across these three distinct industries?
SD: Data analytics applications such as anomaly/outlier detections, A/B testing, classification and regression techniques etc. can be incorporated in each of these industries and one would use the same set of principles to perform the analysis.
With respect to commonalities, at the end of the day statistical concepts and analytics techniques are sector agnostic.
The economics of supply and demand is ubiquitous. In financial securities, a trader would look at the ask and bid prices to analyze the pricing pressures on executed orders. Similarly in call centers, analyzing the incoming calls (demand) vs. calls being answered (supply) can be tremendously helpful for efficient and effective workforce planning.
With respect to differences, I feel there is a stark contrast in workplace cultures across these industries.
In a tech start-up, there’s limited infrastructure and lean teams delivering at a fast pace of execution. In a Fortune 15 company we work with best-in-class infrastructure with teams that have a long chain of command. Driving change in this environment can be challenging.
One Person’s Data is Another Person’s Noise
AH: What companies do you think have the most interesting data to play with today and why?
SD: Great question! In my opinion, Netflix and Spotify. I feel both of these companies have the best recommendation engines in the content space. I follow both Netflix’s tech blog (Netflix Technology Blog) and Spotify’s Technology and Engineering blog frequently. What I like most about both of these companies is how they tie in data from user behavior on their app/site and content consumption patterns to the metadata behind audio and video content. These companies have A/B testing in their DNA and regularly conduct hack days. Apart from state-of-the-art recommendation systems some really cool applications that I read about are in the areas of search indexing, improving streaming quality, artwork personalization and even driving content production decisions using data. With its cutting edge playlist classifications, based on users listening patterns and commonalities in song metadata, Spotify is literally redefining the concept of genres in music. Another amazing thing is that Spotify not only uses data science to improve their service but also builds products for the users to help them understand their individual content consumption patterns.
Above All Else, Show the Data
AH: How do you see data viz helping call center ops in the future?
SD: I am deeply fond of geo-spatial mapping and many times I think of solutions via spatial coordinates. Jokingly, if the matrix is real the programmer is animating on a schema with X, Y and Z coordinates.
Spatial mapping need not always be geo-spatial mapping. It occurred to me that we could take spatial mapping indoors to monitor the customer service ops activity on the call center floor on a location by location basis. Enter Project Omicron! Imagine a dashboard showcasing a floor plan of a call center where you can view operational efficiency on a workstation-by workstation basis in real-time. The dashboard will show in real time how many calls are coming in, how many calls are getting answered, where they are getting answered and by which agent and also highlight the workstation on the floor if an agent is going beyond a threshold for any operational metric. Now imagine executives in the headquarters analyzing this information for any call center across the country anytime they want in real-time. It’s like a data driven ring or canary for the enterprise world!
Based on my understanding, future trends in BI are moving towards solutions that help identify bottlenecks and permute the root cause of issues faster to solve optimization problems. Visualizations that provide location specific insight and render almost real-time data will help drive this.
I envision some consolidation happening amongst companies in the BI space and the different layers in analytics tech-stack such as tagging, ingestion, ETL, storage, querying, analysis and visualization getting integrated as a one-stop-shop cloud based solution.
With respect to front-end I feel there would be more and more applications using Augmented/Mixed/Virtual Reality which would allow developers to visualize several dimensions of data in an integrated and intuitive way. This will lead to a much more immersive experience.
The Goal is to Turn Data into Information, and Information into Insight.
AH: Any more Tableau roadshows in the future?
SD: If the opportunity presents itself, I would love to speak at TC again! This year, I am organizing a company-wide internal analytics summit in Verizon. I believe strongly in sharing viz best practices and plan to start a YouTube channel to do so! I would also love to meet with analytics and data science students in universities to share how we apply visual analytics to solve complex business problems in the industry.
AH: Thanks Sid, it’s been my pleasure talking to you.
To keep up with what Sid is up to and to learn a little more about him check out his website!