Marco Bressan is the Chief Data Scientist for BBVA and the Chairman of BBVA Data Analytics. Marco will be speaking at CON.VERGE.
Marco is a researcher and mathematician and was a “data scientist” long before the term was coined. He has a PhD in artificial intelligence and researched how to teach machines to see as well as machine learning.
We got the opportunity to interview Marco and chat about his experiences with data and how it is changing the customer experience at BBVA.
Marco: I started working at BBVA to lead the Innovation Labs, but soon after arriving I realized that BBVA had many banks, and that they were all over the place in terms of the office of the future, ATM of the future, the role of banking, currency, and virtual currency.
I realized that if I had to narrow it down, there were 2 things that I wanted to focus on. One was data; capturing the value from the huge amounts of data the bank had. They have the richest source of data. It’s not the size of Facebook, but the richness of the data is just amazing. The other was open platforms.
The data activity grew so much that the bank decided to increase its commitment to data. Then soon after I joined the bank in 2012, I created a center of excellence for big data skills. Eventually, the role of the Center of Excellence for doing big data projects grew, transforming the whole analytics practice.
Here are the other questions we asked, along with Marco’s answers:
1. How Does Real-Time Data Play A Role At BBVA?
This is one of the largest opportunities, but also one of the largest transitions for banks. Banks are one of the earliest adopters of real-time data. However, they also have very large and expensive legacy systems and in many cases those legacy systems are incompatible for real-time data storage and processing requirements.
What we’ve done is have some very specific solutions for some very specific products that enable us to do real-time analytics on top of the data. Our focus is working at the same level we’re working with batch data with real-time data, but in very specific domains.
2. What is the connection between big data and personalized customer experience?
This is huge. The origin of every project right now is around customer experience. The bank reorganized itself around a year ago to resemble more of a software company than a bank. If you look at the organizational structure of the bank you have the engineering organizations, a solutions development organization, design, and data.
We work using the triangle structure, where in order for a project to exist it has to have the business, technology, and experience sides of the triangle aligned. By experience, we mean the design and the data. This is interesting because usually data falls under technology, but in our case, it is really related with the experience.
What that means is that when we are building algorithms from the data, what the data scientists have to keep in mind is what kind of experience those algorithms have to deliver. The metrics to which those algorithms are evaluated and developed, like whether I can tolerate this precision or tolerate this error. I can deliver it one way or another depends on that experience.
So, in order to make this happen, we are doing a lot of work in making our designers, our experience designers, our service design team work in alignment with the data teams, so each one knows and understands the other world.
To improve alignment, we have been co-located for specific projects and we try to also find people that have experience with data and also experience with design. The reason for that is the data allows you to model based on behavior as opposed to a survey, etc.
If you are able to understand behavior, it should influence your design. That is the underlying rationale.
3. How is BBVA customizing the customer experience through personal touchpoints?
BBVA is quite advanced in terms of the digital channel development that we have. We have applications, mobile, and web, and many of these have enabled these kinds of personal touchpoints, so the customer can reach us any time they want. For us, the key is not so much the touchpoint itself, but that the experience is seamless across all of these channels.
In order to do that, we use data to get a 360-degree view of what the customer is doing. Just two years ago, a customer could go to a branch, connect over our mobile application, or use the wallet application, and we had no way to link these experiences.
When the customer wanted to contact us through the branch or mobile application, we would only have a very fragmented view of what the customer wanted because we would not know that just 2 hours ago, she was in the branch. Now that we’re able to connect the back end of this, we are able to look at more of a complete customer journey and customize the experience.
4. What are some best practices for promoting a data driven culture in an organization?
First, you need to have some internal skills as we have with our Center of Excellence. There are many layers, between the other analytical teams. We need to build strong communities, so these teams are able to share the knowledge and leverage that knowledge between different countries.
For example, if Spain is working on a specific model, dashboard or metrics, then France can leverage that knowledge too. We have a very successful training called From Data Miner to Data Scientist that promotes a data-driven culture among practitioners, the executive team, and the people who use data-driven decisions.
We are trying to move from the traditional way of working with decisions based on opinions to actually helping those people understand the advantages of making decisions based on evidence.
Most importantly, in building any type of culture, you need to first convince the people that there are advantages in embracing that culture. Right now, it is about showing advantages of making decisions based on data.
What Will You Be Talking About At CON.VERGE?
- How BBVA is leveraging value from data
- Where data can really help change the customer experience
- Challenges with data that people underestimate and overestimate
- My experience with changing culture about analytics
- Sharing what we’ve done in starting to build a data driven organization