Skip to content
Home>All Insights>Podcast: Rashed Younas – Using generative AI to improve CX and meet regulatory requirements in insurance

Podcast: Rashed Younas – Using generative AI to improve CX and meet regulatory requirements in insurance

“It’s about understanding your customers, but also understanding the people that are not your customers today.”

~ Rashed Younas

Insurance is a challenging market to be operating in these days. A perfect storm of the squeeze on household budgets, rising costs when handling claims, and new regulation, mean many insurers are struggling to stay afloat and remain competitive.

Rashed Younas is the insurance practice lead at The Dot Collective, which helps organisations make smarter use of their data. During his chat with Zoe Cunningham, Rashed shares some of the ways he’s seeing forward-thinking insurers use their data intelligently, to navigate these choppy waters.

From generative AI, to single view of the customer, he paints a (perhaps surprisingly) positive outlook for insurance companies. Hear the full discussion on this page, or wherever you get your podcasts. You can also read the transcript below.

Digital Lighthouse is our industry expert mini-series on Softwire Techtalks; bringing you industry insights, opinions and news impacting the tech industry, from the people working within it. Follow us on SoundCloud to make sure you never miss an episode.

***

Transcript

[music]

Zoe: Hello and welcome to the Digital Lighthouse. I’m Zoe Cunningham. On the Digital Lighthouse, we get inspiration from tech leaders to help us shine a light through turbulent times. We believe that if you have a lighthouse, you can harness the power of the storm.

Today, I’m super-excited to welcome Rashed Younas, who is the head of insurance practice at The Dot Collective. Hello, Rashed, and welcome to the Digital Lighthouse.

Rashed: Hi, Zoe, thank you for having me.

Zoe: Maybe if you could start by telling us a bit about your current role, and also your business.

Rashed: So The Dot Collective is a fresh young data consultancy. We started just over two years ago, now. We had our second birthday last month, or the month before. I am currently the head of the insurance practice for The Dot Collective. And I’m also one of the consultants. And as you can imagine, from the title, I work a lot with insurance clients.

Zoe: So what was your route to the role that you have now? Have you had a generalised tech journey, or have you had to specialise?

Rashed: It was specialised, through to generalised, and then specialising again. So I actually started out in very traditional IT support and systems administration, way back in the mists of time. [I was involved] in some quite large system implementation projects for the company I was working for at the time, and started to do more project-based work, and was really interested in business analysis as a skillset, as a craft.

And from there, I did some very big software development projects, working with outsourcing software development houses. I found that project-based consultancy really interesting. So I flipped from working as the client, to going and joining the consultancy side, and doing more of this agile, project-based work. In that journey, I’d always been very interested in data. A lot of the stuff I was doing was around implementing databases, or doing SQL Server development. And when I moved to consultancy, I had the opportunity to work on some interesting data-based projects, and got to mix that product management skill set with a ‘let’s build a data platform’ skill set.

And that’s what I’ve now continued to do. And I get to flex the data specialism here at The Dot Collective – that’s what we’re about, we’re a data consultancy. And I get to merge that with my knowledge of the insurance industry as well.

Zoe: Let’s talk insurance then. So what would you say are the biggest data issues that are facing insurance companies at the moment?

Rashed: I think insurance companies being able to understand all of the data that they have about customers, or prospective customers, is probably one of the biggest challenges.

There are lots of, let’s call them ‘sources’, of information and data. So there’s the systems that I hold my policies and my contact. Maybe I have a CRM system. Maybe I’ve got a website, and I’ve got analytics from, say Google or Adobe Analytics. I’ve got call information. And this isn’t yet introducing all of the potential external sources that I have.

So maybe I’ve got data from Facebook, or other social media platforms that my customers can interact with me [through]. There’s data from maybe price comparison websites, where customers that I have today are interacting with those price comparison websites, because they’re looking for a new line of insurance, or they are looking to renew the current insurance that they have. And that ability to capture all of that data that I have and build a holistic picture of the customers that I have, and the associated data with them, is probably one of the big data-based challenges that insurers have.

Zoe: As soon as you start describing that, I’m like ‘Oh yeah, that’s actually going to be super-difficult,’ because it’s about knowing where all the information is, and then being able to collate it all and collect it all and then combine it into ‘What is your profile, which bits are more important?’

What are the solutions you use to help insurance companies?

Rashed: If I look at one of the projects that The Dot Collective has been working on recently, it’s to build a strategic view of a customer. So we recognise that a customer is a natural person. You are Zoe. You are a physical person. You live in a particular address. And there are sets of data that are associated with you.

So the first thing we would look at, and we’ve built for a client, is to start with the internal sets of data that we have, and build the rules and the logic for how we say, well, I’ve got Zoe in System A, B, and C. How do I match Zoe across these different systems? And now what’s the relevant information that I want to pull together from those systems?

And now, how do I model that in my data platform? And what are the use cases that my users have that makes that, what is now information, useful to them?

Zoe: So you’re looking really almost on a per-company basis for your specific set of customers? What data do you have? And what data do you need?

Rashed: Even where you might have common systems, what you’re storing about your customers is likely to be different. And actually, depending on business goals, what it is you want to achieve, what are the problems you’re trying to solve for your business today? How you approach building and viewing that data is likely to be different as well.

There doesn’t have to be a cookie-cutter, one-size-fits-all solution. You can take ideas, concepts. ‘We’ve applied a concept here. Actually, that

concept would work over here as well. And maybe you can take some of the code that you’ve built to support that solution.’

But it’s not always going to be the same answer for every organisation. What it is you want to glean from that customer data, is potentially going to be different, depending on what insurance vertical you’re in.

Zoe: We’re living in this era of new technologies springing up every few months. Generative AI is a very hot topic, what’s your opinion on how that’s going to impact on the insurance sector?

Rashed: So I find it really interesting, because, as you mentioned, I think it’s a topic that’s everywhere. Everyone’s talking about it. And I think if we use generative AI as the example, it’s also very easily accessible at the moment. Anybody can go to ChatGPT, for argument’s sake, tap in a few sentences, and they’re getting answers to something. And in some cases, people are using it as a replacement for Google.

The thing I find really interesting with a service like that is how can you change customer experience? Or how can you improve customer experience, potentially?

There was a very interesting scenario with a client that I’m working with, a couple of months ago, where we were in a call centre. We were listening to call interactions with customers to understand what are the types of calls that are coming in? Are there efficiencies we can potentially be getting? Or is there better customer service that we can be offering?

In the space of, I think it was 50 minutes, we had something like six or seven calls, all of which were a customer calling in to say, ‘I have tried to do this on the website, I can’t find the place I need to go. Can you help me do this?’ Or ‘Can you show me where I should?’

That’s a very short, easy phone call to deal with. But the thought that occurred to me was, well, actually, if I had a service on the website, where I can type in a question, the way I think to ask it, and that service is able to understand the intent behind my question, and direct me to the place that I need to go… That’s a better experience for me as a customer. It’s not the traditional chatbot that expects you to use specific words, [and] if you don’t use the specific words the bot eventually gives up, or it hits a dead-end in its decision tree and doesn’t know where to go.

And if we take that to the other side, to look at the call centre. Well, the calls that are coming in, I now have more time for that call to be, let’s say, valuable. I can actually dedicate some time to that customer. I’m not looking at a list of calls that are coming in, people that are on hold… I’m sure we’ve all experienced the post-COVID ‘Your call is important to us. Please hold the line. We’re very busy.’

It means if I’m using these generative AI services, the customer experience I can offer is now more about ‘I want to fulfil your need as a customer.’ Not ‘I’m trying to reduce the load in my call centre.’

Zoe: Yeah, I think that’s a really good point. And I think that’s a really good example of where generative AI can be really valuable. Because it’s not about replacing humans per se. It’s about freeing up humans to do the things that humans do best, and like you say, provide the most value, rather than having humans do something that could be done by a machine.

Rashed: Yeah. And if you look at regulations [that] came in through the FCA for financial services organisations called Consumer Duty, a lot of that is about trying to be fairer to customers, but also how you treat vulnerable customers now.

And there’s a big focus on identifying vulnerable customers at the outset. And that shouldn’t have to be a customer explicitly coming to you and saying, ‘I am vulnerable’. That might be you’re talking to a customer, and maybe they’re asking you the same question over and over again, but they don’t realise that they are. Or perhaps somebody mentions to you they’re feeling quite anxious about the phone calls. It has to be explicit, you have to now be more aware of the kind of customers that you have coming to you. And you need to explicitly be fairer to those people that have vulnerabilities.

A good example that I read last week was, if somebody comes to you and says, ‘I’m blind, I would like all of my communications in Braille,’ you have to do that now. This is no longer just good customer service, there is now a regulation that says you have to.

And to bring that back to what we were talking about with generative AI, well, if my call centre is under pressure, and they are trying to deal with customers as best they can, but they’ve also got call turnaround times in the back of their mind, if there’s something I can do to alleviate that pressure into the call centre, it means the time my call handlers can spend speaking to customers that do genuinely need to speak to another person, they can then dedicate that time. They’re not under any kind of stress, [and can] spend that time with the customer and actually do needs-based assessment.

Zoe: Fantastic. And I really like that idea of it’s incumbent on the companies to recognise which customers are vulnerable. So it’s the companies taking the burden of it, rather than the vulnerable customers, who have enough to deal with already. So I think that’s an example of quite smart legislation. And I think it’s awesome if we can use technology to help implement that into these win-wins – I love win-win solutions. It’s one of my favourite things.

What would you say is the climate for the insurance industry at the moment?

Rashed: So I think there’s a number of pressures. We all know about the cost of living crisis, increased inflation. I think what’s interesting from an insurer’s perspective is, if we look at what insurers do, you are provided a policy, which covers some services. And if you make a claim, those claims will have costs. And this is a pretty consistent story across a lot of the insurance market.

If we take cost of living as an example, the cost of providing goods and services has been going up. And it’s gone up in a way where it would have been difficult to predict this 12 months ago. And that’s interesting, because 12 months ago, I sold a policy based on some assumptions about what it was going to cost me to provide those services should they come up. And now 12 months down the line, well, let’s say eight months down the line, I’m eight months into that policy. I’m earning against that policy and claims are coming in. The costs of those claims do not match what my expectations were eight-to-12 months ago.

And I can be adapting now, and the new policies that I’m selling can be adapting to the new set of conditions that we’re experiencing in the marketplace. But I’ve still got that back book that I’ve got to cover. And for some insurers, that back book is operating at a loss. For others, where perhaps that ability to predict, ‘What do I think is going to be that future-case scenario?’ have potentially done quite well.

And now other insurers are a little bit under duress, they’re having to react. They’re realising the profit margins on their insurance book are not what they anticipated them being.

Those insurers [that did predict more effectively], they’re able to explore, ‘What are the next strategic initiatives I want to do? I have some breathing room. I’m not under duress here.’

Whereas I look at some of the market and there are insurers out there who are having to make fast tactical decisions to bring their market expectations back into line, to bring their margins back into line, and to bring their in-force policy counts back to what their budget-setting expectations were.

Zoe: So in short, times are tough. I suppose you’ve got limited options for how to deal with this. Because like you say, technology can make you more efficient right now. And improving your data practices can improve your business right now. But where you’ve got this lag factor from policies you’ve already sold, that’s just kind of a commitment you’ve made that you have to follow through on…

Rashed: Yeah. So you can use your data to understand, ‘What is my position now? And if I want to cover the position that I’m holding, what does my three/six month forecast actually need to look like?’

And now this is not just to meet my financial year budget – it’s potentially actually my prior financial year budget: ‘What do I need to do to cover the in-force policies that I have today? And what data can I use to support that? Is there data that I can draw in to understand, how are the costs of materials changing? What has that change looked like over six months, one year, three years, five years?

Is there predictability that I can build into this? How are salaries changing in the different industries that my pool of workforce is coming from? Whether that’s internal, or if it’s motor insurance, repairers? What are those costs looking like? And what are the factors that affect those? And are there models that I can build that will help me to predict what the cost of those claims might be into the future? And what’s the likelihood? What are people’s driving patterns like today?’

Whilst we are coming back to what we might call ‘normal’, we still call it ‘post-COVID’. You know, there was that big period where lots of things suddenly changed, and our ability to predict what the world looked like, wasn’t there. We were in a new world. And now we’re gradually coming back. But we also recognise that things are still slightly different. We now need to refactor this into our predictive models.

Zoe: Right, exactly. We live in exciting times, I think that’s the only way to sum it up. Just finally, sticking with this theme of how rapidly things are changing and this need to predict the future and look to the future and get as much of a grip as possible on what is coming up, rather than just dealing with what’s happening now… What do you think are the key changes in terms of insurance tech, or data in insurance, that people should be worrying about?

Rashed: So I think for data, it’s interesting to look at price comparison websites, where they want to leverage open finance, as a platform. They want to understand more about, ‘What is it customers want? What is it they’re looking at? And what are the cycles that I already know about? Are they buying contents insurance, buildings insurance, medical insurance, motor insurance? What do I know about them just from the things that they’ve interacted with?’

It’s about understanding your customers, but also understanding the people that are not your customers today. And to bring it back to a theme we talked about, the cost of living crisis. People have cars. They have to have insurance policies, but people are struggling with their finances today.

There are people making decisions about do I buy food or do I buy heating? I have a car to get to work, and I can’t give it up, I have to buy motor insurance. So as an insurer, how can I offer a product that meets the needs of this customer, is affordable, but still allows me to operate my business, [and] that’s still profitable for me? These are the kinds of challenges that insurers in every sphere are going to be facing today.

The data that helps me understand the people that I interact with today, but also, what’s the market that I’m not interacting with? And how can I attract myself to that market? I don’t think you can always just rely on brand. I think price comparison websites have shown that if you’re if you’re an insurer of any kind, you want to be in that top three in that list in the price comparison website. Nobody’s going to scroll down saying, ‘Oh, I want that particular brand. And that’s the only one I’m going to buy.’ At the moment, in the climate we’re in, affordability is key. If you’re not affordable for people, they won’t come to you.

Zoe: The market’s more challenging and more competitive.

Rashed: Yeah, very much.

Zoe: And hopefully, on the flip side, there are opportunities to be more efficient and to deliver a better service by keeping up-to-date with what’s happening, and making the most of it.

Rashed: Absolutely.

Zoe: Thank you so much Rashed, for coming on and helping us to shine a light for others.

Rashed: Thank you so much, Zoe.

[music]

ENDS

Digital Engineering

Get expert help with your digital challenges and unlock modern digital engineering solutions.

Get started now