Marcia Tal is the CEO and founder of Tal Solutions, a New York based advisory and management consulting firm that powers The PositivityTech Platform. This platform utilises customer listening to identify and implement opportunities for financial institutions to anticipate complaints and their causes and address customer feedback. Earlier this year PositivityTech launched the Bias Index, utilising AI to identify bias within customer and employee complaints.
In this interview, Marcia discusses the inspiration behind this platform and how organisations can prevent bias in the first place.
What is PositivityTech?
PositivityTech is an intelligent platform that integrates both human insights and artificial intelligence with different types of advanced analytic capabilities. It uncovers hidden opportunities that are sitting inside of customer complaints that can help with growth, risk management and much more.
The platform itself provides not only an understanding of current complaints but also gives early warning indicators. Leading indicators usually come about from what we call outliers – things that are outside of the norm. The whole idea of using complaints as a data source is outside the traditional box so the whole goal is to apply leading indicators.
What was the inspiration behind creating the platform?
I spent 25 years in banking and spent the majority of those years in different roles related to analytics. Over the course of all of that time, I developed a passion for listening to customers and I would do that in various ways by going to visit customer-facing centres. One of the things that I took away from all of this is that your customers tell you so many things that you really need to know, but we don’t really have a systematised automatic way of using that information as a credible data source. Additionally, I understood that if we actually looked at this as a credible data source, then we would have the ability to build all different types of tools by using this data, whether those are predictive tools, preventative tools, or even visualisation automation tools so that you could have easier access to seeing and understanding what your customers are saying.
I wanted to take all of this and actually build technology around it, fuelling that technology with data and then build the tools to demonstrate all of these capabilities. And that was the beginning of PositivityTech.
You launched the Bias Index on the PositivityTech platform earlier this year, could you tell us more about that?
The Bias Index is an AI predictive model that identifies prejudice within customer and employee complaints and makes it possible for financial institutions to repair products or unjust practices.
I wanted to empower people by using their voices, and now with the intersection of our health crisis, economic crisis and social crisis, it’s become more commonplace for people to use their voice. Everyone should be given a chance to express their opinion and everyone deserves dignity and respect. We recognise that if we listen to what customers are telling us in their complaints then we can identify the biases and work to remove them.
How do these personal and societal biases reveal themselves within these complaints and how does your software pick up on them?
If you can understand the driving forces behind the complaint, then you’ll be able to identify and extract the broader implications from it. We understood that it wasn’t just the words but the contextual reference of the words and the framing that indicates the suspected bias. We also understood that bias in customer complaints isn’t hugely prevalent, in fact, there is a relatively small amount of bias in these complaints making it harder to identify. However, it is important to be able to identify it and use it to grow.
We look at bias in three ways: explicit, implicit or suggested. Explicit means that the customer has actually told you that they have experienced discrimination. Implicit is where what someone is telling you suggests there is bias or potential bias there in the context. And then suggested bias is that the situation can lead to or is suggestive of the potential of bias. It might not always seem apparent that there is bias withing these complaints, but when looking at them with the contextual references around it with the AI tools we have to help then you can identify these kinds of biases.
Where in the industry is the most bias found?
Generally speaking, the most bias takes place face to face. So, in theory, all of the digital environments that we’re spending time on is actually better for bias, because you’re taking out that direct contact where the bias can happen, whether intentional or not.
The index has been live now for quite a few months, have you had much feedback from your clients?
We released the index back in June and one of the exciting things that has been happening is that we’ve now broadened across different industries as well. The reason for that is because bias is universal, it’s not as if it’s only going to be present in one in one kind of industry versus another. We’re finding the opportunity to expand beyond financial services where PositivityTech was first focused and now are moving into health care and the public sector.
What steps can banks and financial institutions take to prevent bias from happening in the first place?
The first step is recognition and awareness, which has to be rooted in data. You have to recognise what your customers are telling you as a credible data source and analyse it in a way that it can be utilised. At PositivityTech we have a four-step process: identify, understand, predict, prevent. First, identify where your risks are and understand them and the bias that is present. Then prediction is using tools like the bias index, where we can find a correlation between the severity of a complaint and the presence of bias within that complaint. And then finally there is prevent, which obviously means stopping it before it can happen.
From these steps we can find the root cause of the bias, perhaps connecting it to certain policies you need to alter or maybe identify that more training needs to take place for your customer-facing employees. You could also look at the diversity of your employee base and see whether that aligns with the diversity of your customer base. I’m an expert in driving data-driven cultures but I think having respect, dignity, patience, understanding and listening right as part of your culture is always going to make an environment people want to be in.
What are your ambitions for the future of PositivityTech?
The first is to get more and more clients to understand the value of what it is that we are bringing to their business. Sometimes it’s not that easy because you’re really having to change the mindset around customer complaints. To us, a complaint is an opportunity, not a problem, and needs to be resolved, and that is the work of many parts of your organisation. We’re not replacing that what we’re doing is trying to raise the discussion and to understand a complaint can be an impetus for success. I have a lot of excitement in seeing clients begin to really embrace what it is that I’m doing.
The second is that PositivityTech has many components to it, so a client doesn’t need to licence the whole thing. We want to help clients understand the four aspects, (identify, understand, predict, prevent) and pick out what they want. Whether that’s just visualisation capabilities, the predictive tools or reports that benchmark how one company is going compared to another organisation. We want to raise awareness and embracement of the integration of your customer complaints as a data source to grow businesses. We also have a lot of ambitions in the data itself, and what broader ways that we can use this data in the future.
