The Center for Data Innovation spoke with Rogayeh Tabrizi, co-founder and CEO of Theory+Practice, a data analytics and strategy company based in Vancouver. Tabrizi discussed the importance of developing a customer-centric data strategy and the effects of the pandemic on retailers’ use of data.
The interview has been edited.
Morgan Stevens: You started your career in physics and then moved to economics. How did that trajectory lead to you co-founding Theory+Practice and what does Theory+Practice hope to achieve?
Rogayeh Tabrizi: Even the name of the company is informed by my background and the way everything has unfolded. As a physicist, especially the kind of physics that I studied, you wake up to a lot of data every day. I used to work at CERN, and we would wake up to 7 terabytes of data, 20 terabytes of data. My education and career at such a young age were informed with a very different perspective of looking and being exposed to different kinds of data, especially big data. From there, switching to economics was less so about the solutions to very complex, complicated problems but rather more about what questions come first. So, throughout my education, I was exposed to these two very different ways of thinking about the world. On one hand, you really want to figure out the right question, and you need to be extremely transparent about your assumptions and methodology. And then, on the other hand, you have tools in order to go solve very big and complex problems that relate to data. When I started Theory+Practice, the company was sitting very much at that intersection. The question was, how can we use behavioral economics and AI and machine learning and all these different techniques that come from fields such as physics and computer science, and apply them to very real customer-centric problems? If you want to go after those problems, you cannot approach them with just algorithms, recommendations models, or segments and, unfortunately, a lot of people do. But the most important thing is to find and understand the question. There are tools in other sciences such as economics that help you think about those questions differently. The work that Theory+Practice does sits at the intersection of AI and behavioral economics and we build end-to-end solutions for our customers. It is again very focused on customer-centric industries.
Stevens: How are data strategies changing for industries that were highly affected by the pandemic?
Tabrizi: One of the industries that we’ve been working with is retail. We primarily work with retail and financial services, two very customer-centric industries with a lot of questions and unknown factors. As you know, retail was massively disrupted by the pandemic. The interesting thing about data strategy for a lot of retailers is that there was already a wave of disruption. Everybody was talking a little bit more about digital and everybody had plans for digital strategies. Those plans all became very much accelerated, in that a five-year plan became a six-month plan, because the only channel that you can actually talk to your customers on is now online and digital. So the question becomes how are you going to think about what information you’re collecting from your customers that you can use to remove barriers and put the right product in front of your online customers. Everyone is affected, from grocery stores to clothing brands, and the pandemic and subsequent digital acceleration have resulted in retailers like Walmart, Amazon, and Walgreens relying on mobile apps to improve customer experiences. I think retailers are realizing that customer loyalty is something that can’t be taken for granted and are focusing on creating ease for the customers instead of having the cheapest prices. When it comes to data strategy, a lot of retailers realized the importance of creating a 360-degree view of the consumer and using digital to improve consumers’ lives instead of just creating a new app. Businesses started asking what we actually know about our customers and what we can understand and learn.
Stevens: What does the process of working with a specific client and formulating a data strategy look like?
Tabrizi: Let’s say we are working with a massive retailer. Again, the first question is always what are our use cases? For example, we need to understand our customers better, or we need to create a better search algorithm, or we need to improve our loyalty program. There are a lot of different use cases that people have. But the same set of data that is going to help you develop deep behavioral segmentation of your customers is the same data that will help your customers, help you understand their price sensitivity or promotion affinity, and help you build a proper loyalty program. Instead of thinking about them as siloed, you really look at the ecosystem of different use cases that you have. There’s data, your raw data, that needs to be available and consistent. There are pillars when it comes to actually thinking about your information architecture, from your ingestion to identity resolution to the availability of data to governance systems and mechanisms around it. Within and informed by those pillars, you look at your use cases and identify 10 different priorities. Within the context of those use cases, you go back to your pillars, and very iteratively and in an agile way, try to understand what we are prioritizing differently and what actions can you take to make sure the data is available, consistent, accurate, and properly governed. Finally, you have to think about who is using the data and whether or not they have the tools to use it properly so that there are no negative unintended consequences.
Stevens: What do you expect to be the biggest opportunities and challenges to using consumer data in the future?
Tabrizi: The two biggest challenges will be data security and privacy. It is very important to have proper considerations with the policies and framework around data. There are also more technical challenges when it comes to online and digital, as the amount of data that you can collect is exponentially more than the transactional or product data that you can collect through a point-of-sale (POS) system. People have to think about consistently making different kinds of data collected through different sources available. And that data needs to speak to each other to create a proper 360-degree view of customers.
Stevens: How can businesses use a data strategy to not only increase efficiency and profitability but also promote social good?
Tabrizi: Start by asking the right questions and really examining any assumptions and KPIs (key performance indicators) you want to maximize or optimize. I don’t think any organization is just purely evil or thinking only about how to maximize profit. Profit is a very important KPI for businesses but in order to become profitable, you have to make sure you retain your customer. Businesses will be much more profitable if they take care to make sure their customers are happy instead of extracting the last marginal dollar from their customers. In order to make customers happy, you need to ease their shopping experience and remove any obstacles or barriers in their way. There are a lot of ways that businesses can use data to understand barriers and frictions and identify places they can do a better job. In order to truly become profitable, you have to have longevity, and in order to have longevity, you need to have your customers. That said, there are a lot of unfortunate biases and ways in which data is not used properly. That’s why I always advocate for going after asking the right questions because if you start with the right question, the odds are at least you have a higher probability of success.
