Bala Gopalakrishnan on Leveraging Data-Driven Insights
Bala is Chief Data Officer at Pelmorex Corp, the parent company of The Weather Network, MétéoMédia, Eltiempo.es, Clima, Otempo.pt, Pelmorex Data Solutions, and Pelmorex Audience. Pelmorex is a leading player in Digital Advertising and Weather B2B/B2C solutions. Bala leads the Engineering and technology talent at Pelmorex Data Solutions and Pelmorex Audience, which boasts one of Canada’s best GIS, Data Science, Big Data Engineering, and BI talents. Bala heads up industry-leading research into AI/ML with internal teams in collaboration with top Canadian Universities.
What is your day-to-day currently like and is there anything exciting that you’re working on?
Just to give some background, Pelmorex, which owns The Weather Network, operates in the B2C space as a leading consumer brand that people know and consumers come here to get weather-related information. Being a weather company, we are sitting on massive amounts of historical weather data as well as weather forecast data.
Because of the fact that over 28 million Canadians interact with our apps, with our websites, and on TV frequently, we also get enormous amounts of behavioral contextual data, what consumers are consuming weather for, in a very privacy-safe aggregated and non-personally identifiable way.
We want to use weather and all the insights that we have to innovate and apply this data along with data science to derive actionable insights for solving business problems. An example of this could be the restaurant industry taking weather into account to help them manage take out optimally, especially during the pandemic. Another example could be involving hurricanes and snowstorms. How do we optimize the impact for the insurance industry and optimize their claims process or for marketers who are spending money on advertising, how do we take this data and optimize it?
What kind of data-related trends are you noticing right now?
The increased adoption of data is now accelerating this velocity of decision making so the value of data deriving actionable insights has tremendously increased.
The pattern that we are beginning to notice, is that data and analytics used to be an interesting opportunity for companies to go into but most decisions were being made in the traditional way. Today, I think we can all agree that with data and analytics, every organization is using data to make decisions. All organizations have invested some amount, whether it be big or small towards data analytics teams as well as visualization dashboards and metrics.
The difference in what we are noticing or the trend that we are noticing is this used to be like a waterfall, an industry would create the insights or a company would create insights, it would go to their managers and stakeholders who will then use these insights to create tactics for the industry or the company.
The increased adoption of data is now accelerating this velocity of decision making so the value of data deriving actionable insights has tremendously increased. Where the industry trends have moved towards is small bite-sized actionable insights, so you can quickly experiment with them, test and confirm the hypothesis, and if those yield good results then utilize those insights. It strengthens the overall analytics process so what we’re seeing in our industry is this trend of going from data and analytics to quicker actionable insights and then experimenting and learning from it.
Has the pandemic influenced any of your data strategies?
We had these distinct patterns prior to the pandemic of why people checked the weather. And historically, people need to plan around the day, and they start in the morning by checking the weather. With the pandemic, while some of these groups of customers & consumers still exist, our behavior in day-to-day life has completely changed. Many of us work from home, many of us have a lot more time in the day and our time is utilized in various other ways than before the pandemic
The reasons why people check the weather has changed but now we see more people are checking the weather at 10 am because that’s the time they want to go around for a walk in their neighborhood or they want to take their pet out and so on. The context of why the weather was being checked has changed during the pandemic and it’s put emphasis on various types of content. People used to look at long-term weather forecasts, for trips, vacations, etc., but now we see a lot more emphasis on people checking hourly forecasts to see what’s going to be the weather two hours from now. This is the change from the B2C side.
From a B2B side, the Pandemic has changed lots of industries and operations and whether it’s e-commerce delivery or whether it’s a retail company or even government trying to deal with this pandemic. We can definitely agree that the impact of weather and behavioral patterns of consumers has definitely changed. From that perspective, we’re seeing a lot of demand from various industries wanting to understand how the weather has impacted their industry in the past and use our data science and data expertise to derive actionable insights on how they plan out the next six months.
What is your greatest challenge right now when it comes to data? What is keeping you up at night?
We are living in a time where the data behavior of businesses and consumers is being changed forever. A lot of these changes and trends will be everlasting or at least long-term. They are not going to change again. But we are also seeing that some of these changes in behavior and the consumers wanting to leverage all things digital and data to make decisions.
This also opens up opportunities to firms like us, because we are right in the middle of that. Our challenge and opportunity is to really figure out what are these massive shifts in consumer behavior, where are the right opportunities to serve consumers, and what value can be derived from that. Then aligning this with the exclusive weather and behavioral data that we have and the capabilities that we have in order to create solutions that will both impact consumers as well as businesses and deliver value to them.
The pandemic has kind of moved us to a strategic point where every value prop can be questioned and therefore we are trying to identify what is the purpose and solve the problems in the right way.
What kind of strategies do you have in place to make sure that you are maximizing the use of the data that you have?
I think we are constantly learning as well, but we are fortunate that in our company, data was always considered to be very important and we had a data analytics team set up several years ago when data was the largest buzzword in the industry.
We need to understand thoroughly how consumers use our services, what do they use them for, where do they see the value and that’s what our strategy is. To understand the enormous amounts of data that we get from the interactions our consumers have with our apps and website to derive this context. Our strategy is to utilize that and to continually personalize and improve products for them.
A good example of that is over the last few months, we’ve had a product come out called Storm Centre. Our app automatically changes when there is an active weather update, there is basically a television studio on the app. There’s a continuous video and continuous tweets and information to consumers who can really use that type of one-stop place to understand what’s happening because we were looking at how our apps and the data that was being consumed from a B2B perspective.
Our end-to-end applications and properties are all free to consumers and we monetize by advertising. We do need to improve the value of the ads that we show in our app or our website while delivering better value for advertising clients while still maintaining privacy. A lot of our data is to make the advertising work more effectively. We are living in a big data world where there is enough horsepower and computing that we can combine this with enormous amounts of other business data and AI to unearth key patterns of these industries.
How are you solving your data quality issues?
It starts at the collection itself. I think data quality issues are very specific to certain industries and also the impact on the accuracy of the data is also specific to certain industries but to us, I think the basic stuff we do around collecting data, which is through our apps or websites, we make sure we build it the right way from the start. It starts at the collection itself, that there are no gaps there and then the second is that we make sure that there are enough rules and procedures in place that we are able to remove poor quality data.
It’s the same thing when we have data from other vendors, we go through this extensive scrubbing process to remove poor quality data, and then finally when we apply this data either through data science algorithms or applications, we make sure the data is understood correctly and that we don’t have things that are built out of context.
Depending on the data we have, other factors come into play, such as user experience, we make sure whatever we create using that data is making sure that the data is trustworthy, but also that what we create has enough context.
What innovations do you use around data protection?
It is such a big issue these days especially if you’re an organization that’s collecting data touchpoints from its consumers. We are constantly evaluating the market and improving our data protection to the gold standard every day, and we have put in place fundamental principles that we follow because our company takes consumer trust very seriously. Data protection is not just regulatory, and we are not just doing this for compliance, millions of consumers are trusting us.
First, it starts with the consumer so we are transparent in all our properties in our apps on what data we are collecting. It’s not hidden in a 50-page privacy policy, it’s in one paragraph right in front of the consumer and they say yes or no. We give them the ability to opt out and respect that if they don’t allow for the collection of certain types of data. We’ve made sure that we don’t mix any PII and non-PII data so we can keep it completely separate and any data that we use for insights is from non-PII and it’s completely anonymized and aggregated so this data is never available at an individual user level.
Do you have any final words or topics you think that the public needs more awareness around?
The more people adopt these great tools, there will be a lot more information can be used for intelligent decision-making.
I would say there is a lot of misgivings about data in general. I think the public is afraid of any data sharing or in general the word data. I think data has two sides of the coin. I think there are definitely things that we need to be worried about when information like our credit card goes out but there’s also great information about weather, about movements of users, about the information that people share, it can be used in a lot of great ways for decision making whether it’s the pandemic or whether it’s things like the COVID alert. The more people adopt these great tools, there will be a lot more information can be used for intelligent decision-making.
