Episode 27: The Future of Technology in Financial Services
Download MP300:00:00 Dr Genevieve Hayes
Hello and welcome to value driven data science brought to you by Genevieve Hayes Consulting. I'm doctor Genevieve Hayes. And today I'm joined by Ben Shapira to discuss the future of technology in financial services. Ben is a digital strategist and UX specialist turn tech entrepreneur.
00:00:20 Dr Genevieve Hayes
He is the founder and chief product officer of Australian fintech startup De Niro.
00:00:27 Dr Genevieve Hayes
As well as being a lecturer in the Master of Media and Communications programme at Swinburne University, Ben, welcome to the show.
00:00:36 Ben Shapira
For having me.
00:00:37 Dr Genevieve Hayes
Despite its conservative reputation, the financial services industry has always been a bigger doctor of cutting edge technology.
00:00:46 Dr Genevieve Hayes
I started my own career working in financial services because it was one of the biggest employers of people with technology and data related skin.
00:00:55 Dr Genevieve Hayes
And prior to DJ Patel declaring data science to be the sexiest job of the 21st century, becoming a quantitative analyst for a big investment bank was what every ambitious data nerd seemingly aspired to be. But what does the future look like with regard to the use of tech?
00:01:15 Dr Genevieve Hayes
In the financial services industry.
00:01:17 Ben Shapira
I mean it's a fairly open question, which I think is is really nice because it if you look historically, banks have been quick to adopt technologies.
00:01:25 Ben Shapira
That would streamline their their services and increase their bottom line. So that would be something like ATM's, reduced number of bank branches, automation as much as possible of digital lending. The the interesting part of technology implementation now is open banking, which has been a global initiative.
00:01:46 Ben Shapira
For the last.
00:01:47 Ben Shapira
Say 7 to.
00:01:48 Ben Shapira
Eight years in large part Australia's been a little bit late to the party, but not in so far as a lack of skill set.
00:01:56 Ben Shapira
It's largely been due to sort of a pragmatism of seeing everyone else attempt at 1st and seeing where the problems are and seeing not sit that we haven't had our own, but to to sort of.
00:02:07 Ben Shapira
Assess how open banking is being implemented and how it can best serve the Australian public.
00:02:14 Ben Shapira
For us, it's an opportunity for consumers to have significantly more control over their financial well being who has access to their financial data to be able to share their financial information either on an ongoing basis or on a single one.
00:02:34 Ben Shapira
Time use scenario like applying for a credit card and allows them to really be the gatekeepers of their own technology, their own, their own data.
00:02:42 Ben Shapira
It's in large part democratising the financial data that a consumer has.
00:02:48 Ben Shapira
And allowing allowing them.
00:02:49 Ben Shapira
To leverage it to the maximum capability.
00:02:52 Ben Shapira
For their own benefit in the long term.
00:02:54 Dr Genevieve Hayes
That's the principle that De Niro's built on, isn't it?
00:02:58 Ben Shapira
Yes. So dinero came about in large part because I was applying for a home loan. I've been living in Australia for at the time, approximately 14 years and I wanted to buy.
00:03:09 Ben Shapira
Film and I'd saved up my deposit. I didn't really know who to talk to in Canada, which is where I'm from and actually where I'm speaking.
00:03:16 Ben Shapira
To you from now.
00:03:17 Ben Shapira
The broker market is is very, very small, but in in Australia the broker market represents about 70% of all loan originations for home loans. So I.
00:03:28 Ben Shapira
Didn't really know.
00:03:29 Ben Shapira
Who to talk to? I didn't know what I could borrow. What kind of rules I had to follow. So I found a broker in my neighbourhood and I went to go and talk to them.
00:03:39 Ben Shapira
And I went through the process of going through the pre approval like.
00:03:42 Ben Shapira
Most Australians go.
00:03:43 Ben Shapira
Through at the same time, I was going to try and find a home.
00:03:48 Ben Shapira
Well, every 30 days you have to resupply your information because your pre approval only only lasts for 30.
00:03:54 Ben Shapira
Days. So it's a very, very frustrating experience having to manually download all of my data from a variety of different sources, collate it, send it to my broker, and then, you know, every 30 days having to go to that same process again until I potentially.
00:04:08 Ben Shapira
Buying the property.
00:04:10 Ben Shapira
I've been in the digital industry for 30 years doing building products for other businesses, so I thought there's got to be a better way to do this and open banking really is really at the forefront of of of our capabilities.
00:04:22 Ben Shapira
To do that.
00:04:23 Ben Shapira
We can aggregate all financial data into a single source, allow you to find a broker.
00:04:30 Ben Shapira
Or a representative, either financial advisor within the app to share your financial data on an ongoing basis, set financial goals for yourself.
00:04:38 Ben Shapira
You got a really good solid understanding.
00:04:40 Ben Shapira
Over what your?
00:04:41 Ben Shapira
Net worth is.
00:04:42 Ben Shapira
And you can share that data with a financial professional where we've chosen to be fairly pragmatic though.
00:04:48 Ben Shapira
Is that we don't allow any transactions in our app and everything is fully encrypted. So we we wanted to find a really interesting balance between facilitating communication and creating that engagement between consumers and professionals, but also being very, very mindful of consumer data and consumer privacy so.
00:05:08 Ben Shapira
We we we are doing our best to sort of straddle that line of maintaining security and privacy while enabling the consumer to choose to whom that they are sharing their financial data.
00:05:20 Dr Genevieve Hayes
Like to keep going and talking about taking the financial industry a bit more later, but one thing I'm interested in is your own background because.
00:05:30 Dr Genevieve Hayes
I've read your LinkedIn profile. You have an incredibly impressive background in the advertising and marketing industry, with decades of work experience teaching that industry.
00:05:41 Dr Genevieve Hayes
Was that experience with applying for that loan the first time when you thought, hey, it might be a good idea to work in the financial industry?
00:05:51 Ben Shapira
I'll answer your question I've.
00:05:53 Ben Shapira
I've worked with financial financial businesses in the past, but when you're working for an agency, you are largely taking what I would call a superficial understanding of your customer in in large part your focus primarily on understanding their product or service insofar as you will need.
00:06:12 Ben Shapira
To sell that part or services on their behalf across social media, other digital platforms, Google Adwords or even through traditional sources like billboards in a product development perspective, it's a little bit different because.
00:06:25 Ben Shapira
You have to take.
00:06:26 Ben Shapira
A more in depth understanding over their business. When I stopped working for the agencies my last full time role was head of digital for Clemenger BBDO in Melbourne and I started consulting for for Repco and it was a five year stint working with them to help build their consumer facing website.
00:06:47 Ben Shapira
So I in large part did a lot of the master design work and usability work worked with their third party providers, their their coders, their SEO and SCM providers and worked with the team to sort of keep them, keep them on track and keep them applying best.
00:07:04 Ben Shapira
This, I think it was that experience in going from an ad agency side to a A A product side that really gave me an appreciation for how much in-depth you really have to understand an industry before you can really start building something for yourself insofar as dinero is concerned. It took me a year and a.
00:07:23 Ben Shapira
Half of research.
00:07:24 Ben Shapira
To really assess the viability of the idea that I.
00:07:28 Ben Shapira
And that included speaking to over 350 consumers speaking to dozens of brokers and asking them, you know, question after question after question to really understand whether what I was experiencing was just for me or whether it was a an institutional problem, something that would be ubiquitous for everybody across the country that I could then.
00:07:48 Ben Shapira
Build the business around.
00:07:49 Dr Genevieve Hayes
With regard to actually getting the technology capabilities, was that challenging?
00:07:54
Well, you.
00:07:55 Ben Shapira
Very yeah.
00:07:56 Ben Shapira
There's there's multiple vendors who can supply financial data. There's also limitations on what they can provide. So it's it's not really relying on a single data source, but multiple data sources to give the consumer and the broker everything they need. Currently we have four different API's that we connect with. We're about to implement additional ones.
00:08:17 Ben Shapira
Or version two of our.
00:08:18 Ben Shapira
Software, but it's also understanding what kind of data you're going to get, how in depth and how correct the data is going to be, how many different vendors you're going to be able to access.
00:08:29 Ben Shapira
We've chosen to partner with a data provider called Yodlee. They're the largest and most respected data provider in the.
00:08:36 Ben Shapira
World, we've done that.
00:08:38 Ben Shapira
In in part because they are the largest.
00:08:40 Ben Shapira
They give us the opportunity to expand globally should we want to in the future, but they also offer us the the largest number of connections so we we can connect to over 330 different financial institutions.
00:08:52 Dr Genevieve Hayes
In understanding your customer base, you obviously went out and spoke to customers and brokers. How did you gain that understanding that you should go and subscribe to some sort of API for data?
00:09:05 Ben Shapira
As part of my role in the UX space, we do a lot of research, so typically when we're building or designing something, we build in what are called wireframes.
00:09:16 Ben Shapira
They're designed agnostic box set diagrams and we we build prototypes of the functionality that we're trying to create.
00:09:25 Ben Shapira
We leave it as design agnostic so that people can focus on the specific task and and we invite people into rooms.
00:09:32 Ben Shapira
We record them both the audio and the video, so we can see what they're engaging with. We can also see their faces and we listen to their audio and typically what we do is we just assign them a task and we say find a product of this type, add it to your shopping cart and check out and then we ask them to describe their experience.
00:09:52 Ben Shapira
As they're going through that process. So that's how UX typically works from, say, an E commerce website for what?
00:10:00 Ben Shapira
We had to do for web.
00:10:00 Ben Shapira
So in my case I was asking them those same sort of questions but asked them to describe their experiences in.
00:10:09 Ben Shapira
Applying for home loan.
00:10:10 Ben Shapira
Did you know who to like? How did you find someone to speak to? What was the process in your application? How many different pieces of software did you have to connect with? How many different bank accounts?
00:10:21 Ben Shapira
Do you have how many different apps on your phone do?
00:10:23 Ben Shapira
You have for financial services.
00:10:25 Ben Shapira
How long did it take for you?
00:10:27 Ben Shapira
To find your home.
00:10:28 Ben Shapira
How many times did you have to reapply for your pre approval? Questions like that that really helped me form a broader understanding over whether or not these people have.
00:10:38 Ben Shapira
The same experience that I had.
00:10:39 Ben Shapira
And what I noticed was that there were really three different types of customers that I was dealing with first time home buyers who in large part were dealing with what I was dealing with, a lack of education, a lack of knowledge, a lack of familiarity with the systems and services that they were engaging with. And then they were existing homeowners who were either looking to.
00:11:00 Ben Shapira
Upsize or downsize or to refinance their loans. Who had their own issues around having to reengage with the broker and then the third were were investors, so people who already owned property who were constantly engaging with their brokers.
00:11:18 Ben Shapira
And needed a better way to transact. So all three of those provided a different perspective on the different features and capabilities that we wanted to use the app.
00:11:27 Ben Shapira
And on the flip side, asking those same sorts of questions to the brokers about how they engage with consumers, how they generate leads, what happens during the nurturing.
00:11:38 Ben Shapira
Process of acquiring customer to converting them.
00:11:42 Ben Shapira
But also, once a customer becomes a customer, how often do you engage with them? How do you leverage the portfolio that they've developed to help them expand and create more value, what we call in the in the market in, in the marketing side of things, the customer lifetime value, how do you maximise the value of a customer?
00:12:02 Ben Shapira
Not just from your own commissions, but also helping them create a more diverse portfolio.
00:12:08 Ben Shapira
All of this contributes to how we've taken a view on building our product, and I've been really lucky in the fact that I have.
00:12:15 Ben Shapira
I belong to an entrepreneurial group called Cub, and I've had some fantastic connections with members there, and they've in large part given me hours and hours and hours of time to help me qualify.
00:12:29 Ben Shapira
That what I went through was was correct in the end took to quantify the number of their customers that would want this particular piece of software and how best to sell it. So not to consumers but through the brokers and through the aggregators, the software software companies that.
00:12:45 Ben Shapira
Brokers used to lodge their applications taking a top down approach. All of that really helped me in understanding how to build this particular product.
00:12:54 Dr Genevieve Hayes
With all those, I think you said 350 consumers that you interviewed, plus I forget how many brokers it was, were all those people who were connected to your network in.
00:13:05 Dr Genevieve Hayes
Someway. Or did you have to go out and find people who weren't connected to your network through? I had no advertising for people who'd be willing to be interviewed.
00:13:16 Ben Shapira
Yeah. So about a third of them were my own network. The other 2/3 there's, you know, platforms on the Internet where UX research is conducted and you pay people for those interviews.
00:13:26 Ben Shapira
And depending on.
00:13:27 Ben Shapira
The the complexity of the interviews you.
00:13:28 Ben Shapira
Pay them different amounts.
00:13:30 Ben Shapira
So I conducted these interviews and mostly over type form.
00:13:35 Ben Shapira
So you know the.
00:13:36 Ben Shapira
It's like a questionnaire software and once I got the feedback back then I would take samples.
00:13:42 Ben Shapira
Out of those, so I would.
00:13:44 Ben Shapira
Take, you know, random samples of different age groups, different ethnicities, different genders, different geographic locations, different income levels.
00:13:53 Ben Shapira
So I I try to get a really good snapshot of the people who are answering the.
00:13:57 Ben Shapira
Questionnaire and then I.
00:13:58 Ben Shapira
Would do zoom interviews with them.
00:14:00 Dr Genevieve Hayes
That's really good. So it sounds like.
00:14:03 Dr Genevieve Hayes
You were able to take a lot of the skills that you'd learned from your background in advertising and marketing and apply them to the tech startup industry.
00:14:13 Ben Shapira
Yes, in large part this is and this is actually no different than being a social media analyst, which is part of what I teach at Swinburne.
00:14:21 Ben Shapira
It's really being able to understand who your core audience is being able to create personas so stories about who these people really are going beyond traditional demographic breakdowns.
00:14:34 Ben Shapira
Different. Different. So the historical side of things of of purely gender, age, income.
00:14:40 Ben Shapira
And really understand what makes them tick. Giving them a story, giving them a name, putting a picture to that profile, understanding their feelings about topics that don't necessarily relate to your particular agenda, but can help you formulate a better understanding of who these people are.
00:15:00 Ben Shapira
One of their politics how do they?
00:15:01 Ben Shapira
Feel about health and.
00:15:02 Ben Shapira
Well being, do they buy new or do they DIY?
00:15:06 Ben Shapira
Are they more into technology or are they more analogue like understanding who these people are can help you understand the likelihood the propensity for them to use your software down the.
00:15:17 Dr Genevieve Hayes
Are there any other skills that you learned in your career in advertising and marketing that you're now putting to work in your new career as a Fintech startup founder?
00:15:27 Ben Shapira
Yes, absolutely. So being a UX designer and a graphic designer which I have been since 1996, I am the the chief creative for everything. That is De Niro. So I've designed and built the website.
00:15:42 Ben Shapira
I've designed the app interface, the portal interface. I've been lucky to have a really good set of people to do the build for me.
00:15:51 Ben Shapira
We now have our own team, the our first build was largely offshored, but I designed it myself and even with all of my experience, it took 43 iterations of the of the design to get it right.
00:16:03 Ben Shapira
For the first version of the app and the key learnings that I learned from that process has now helped me design and build for version two, which was currently in in development as we speak.
00:16:14 Dr Genevieve Hayes
Do you write computer code?
00:16:16 Ben Shapira
I write some but very little. I write HTML and CSS, so basically I can design and build a website for you.
00:16:22 Ben Shapira
I could, you know, build you. I could integrate, you know, a static website that I could build with, say, WordPress.
00:16:28 Ben Shapira
But I don't write JavaScript, I don't write any code. Our platform is built on react and react native. I certainly don't write any of that.
00:16:36 Ben Shapira
I would very much prefer to hire specialists, especially for you know for privacy and security.
00:16:42 Ben Shapira
Versus then to rely on my own skill set.
00:16:45 Ben Shapira
For that I.
00:16:46 Ben Shapira
Prefer focusing on the creative when it comes to to anything in the digital.
00:16:50 Dr Genevieve Hayes
Space from the point of view of someone who has worked as extensively in advertising and marketing as you have, is there any advice that you would give to data scientists who are looking to market their work?
00:17:02 Dr Genevieve Hayes
Two clients or senior managers or just to build their personal brand?
00:17:07 Dr Genevieve Hayes
In general.
00:17:08 Ben Shapira
Yeah, they're look the data analysts in the advertising space are in very high demand in a number of different roles.
00:17:14 Ben Shapira
So in the social media space, specifically, if you're going to be talking about behavioural analytics that helps you when it comes to targeting lead generation, conversion rate optimization, all of those skills are really quite in handy if you're going to work for a media agency.
00:17:30 Ben Shapira
Media buying agency, for example, like, say for example like hatched media or non media.
00:17:35 Ben Shapira
The they're really looking for data scientists who once again can help them better understand who their audiences are so they can use that from an Omni channel perspective to help develop the purchase criteria.
00:17:47 Ben Shapira
The purchase plan for media spend that might be a mix of online, offline media spend, you know tram wraps.
00:17:55 Ben Shapira
Billboards, digital displays, as well as online media.
00:17:59 Ben Shapira
And then I.
00:17:59 Ben Shapira
Think you also?
00:18:00 Ben Shapira
Have huge number of corporations who are also looking for data specialists in their marketing teams in their customer service team.
00:18:11 Ben Shapira
To really help them build.
00:18:12 Ben Shapira
A better mousetrap, you know, they're they're really trying to understand who their customers really are.
00:18:18 Ben Shapira
And Australia is really great at building some fantastic products and services, but we do take our time when it comes to really, truly understanding who our customers are. So we are slow to.
00:18:31 Ben Shapira
To launch, but what we do launch tends to be really high quality products and services.
00:18:37 Dr Genevieve Hayes
Do you incorporate data analytics or data science into dinner?
00:18:42 Ben Shapira
We do so it will scale up overtime. What we're doing right now is looking at consumer purchase, sorry, consumer interaction with our app to understand the various feature sets that we have and understanding their their relative value. We are also looking at the same from a brokers perspective.
00:19:02 Ben Shapira
And we do.
00:19:03 Ben Shapira
Interviews with both consumers and brokers on a regular basis to ask them, you know what they like, what they would like to see.
00:19:10 Ben Shapira
We are quite transparent with our existing customers over what our future backlog is, so they can help inform what comes next and how we build that out.
00:19:20 Ben Shapira
Some of them we actually do UX interviews with some of our existing customers and we actually engage with them and say you're gonna see what's gonna come next.
00:19:29 Ben Shapira
We want you to help us make it better or even to help us break it so.
00:19:33 Ben Shapira
We can try and.
00:19:34 Ben Shapira
Make it better.
00:19:35 Ben Shapira
So that's that's sort of where we are right now. But as we progress over the next.
00:19:39 Ben Shapira
Couple of years.
00:19:40 Ben Shapira
The financial data that we're collecting will help us really understand consumer purchase behaviour and that has any number of applications.
00:19:49 Ben Shapira
It has to be anonymized, so by law we have to remove any unique consumer identifying information. But we can take financial.
00:19:57 Ben Shapira
Data in aggregate.
00:19:58 Ben Shapira
It and help us better understand consumer purchase behaviour. The likelihood that someone is at a specific stage in their life and are likely to need various products and services.
00:20:08 Ben Shapira
So I'll I'll give you an example. Facebook has determined that newlyweds have an 80% chance of having a child within the first 12 months after.
00:20:18 Ben Shapira
They've been married.
00:20:19 Ben Shapira
If we can determine based on purchase history, the likelihood that that particular cop.
00:20:26 Ben Shapira
Is actively trying for a child. We can prepare that data in advance.
00:20:31
What that means?
00:20:32 Ben Shapira
Is that the constellation of businesses that would be interested in that financial data or in that persona would be insurance companies, companies who sell vehicles because the family is going to want to get a bigger, safer?
00:20:47 Ben Shapira
Vehicle like you're gonna want to upsize their home. They're gonna need not just, you know, more insurance for their home, but also life insurance for the.
00:20:55 Ben Shapira
Their family, their companies that have very specific and very well thought out, well considered marketing plans like baby bunting, who has an 8 year marketing plan for conception to about 8 years old where they will change their marketing behaviour based on the data they've already collected and they know about their own customer base.
00:21:15 Ben Shapira
If they can have the data to know when someone is actively trying or has recently announced they're pregnant, that makes the data quite relevant to their marketing strategy. So in large part what our future looks like is anonymized financial data and.
00:21:33 Ben Shapira
Better get so we can extract consumer personas. Let marketers let brands.
00:21:40 Ben Shapira
Use that data.
00:21:41 Ben Shapira
To find what are called look alike audiences.
00:21:44 Ben Shapira
So in the social media space and in the digital marketing space, we can upload data sets and those companies will then say, well, now that we have a framework of the type of customer you're looking for.
00:21:55 Ben Shapira
We will then leverage that framework to find other consumers that fit that same criteria, or large chunks of that criteria and help to drive more lead generation, more visits, more eyeballs to your advertising.
00:22:09 Dr Genevieve Hayes
And I take it the more detailed that profile is, the more useful it is.
00:22:14 Ben Shapira
It is and and that really comes down to sample size. So we know we need to make sure that we have a large enough sample sample size that.
00:22:22 Ben Shapira
The data becomes statistically relevant.
00:22:24 Ben Shapira
Otherwise the data is just that, it's just data.
00:22:27 Dr Genevieve Hayes
At what point would you say that the sample size is big enough for it to become statistically relevant?
00:22:34 Ben Shapira
For financial data, it's probably a lot smaller than in other types of data. Financial data and medical data statistically have require fewer, smaller samples.
00:22:44 Ben Shapira
Size, I would say, depending on who you're talking to, who you're selling that data to the the relevant size will differ and you know you have to look at that also is do you partner with other with other providers to to provide a more complete.
00:22:59
View of that.
00:23:00 Ben Shapira
Data, for example you.
00:23:02 Ben Shapira
Know our data talks about.
00:23:04 Ben Shapira
Total spend individual transactions, but what we don't see is what's inside your shopping.
00:23:08 Ben Shapira
Basket while someone like flybys, for example, will see what's in your basket but has a limited view over individual retailers.
00:23:17 Ben Shapira
So you know combining our data might be an interesting prospect down the road as a sort of combined effort to provide a more comprehensive.
00:23:26 Ben Shapira
View of of the.
00:23:27 Ben Shapira
Of the customer, a more holistic.
00:23:29 Ben Shapira
Understanding over you know what they're spending in total versus what's in their individual shopping baskets and specific and.
00:23:37 Dr Genevieve Hayes
This combining data cause any conflicts with regard to data anonymisation.
00:23:43 Ben Shapira
Well, what we're looking for is patterns. So all the data that we would end up putting together would have to be anonymized in order to be leveraged. It's legally required that we have to pull all information that could identify a specific consul.
00:23:59 Ben Shapira
That information has to be removed so you know name, phone number, email address, physical address, bank account numbers, all that sort of information has to.
00:24:08 Ben Shapira
Be pulled from the.
00:24:09 Ben Shapira
Data completely.
00:24:10 Ben Shapira
And that's not just from us, but also prior to engaging with any third party partnership or or resale.
00:24:17 Ben Shapira
We would have to.
00:24:18 Ben Shapira
Do that as well.
00:24:19 Dr Genevieve Hayes
There are other ways you can reidentify data. So for example, if you had someone's banking transaction.
00:24:28 Dr Genevieve Hayes
Things. And you knew that they'd purchase made a purchase for $10, the local supermarket, and you had their flybys data.
00:24:38 Dr Genevieve Hayes
You could theoretically match those two data sets based on that transaction at that date at that location. Do you have to remove stuff like that which would allow you to connect the data sets?
00:24:49 Ben Shapira
Well, so flybys would not include the flybys number.
00:24:52 Ben Shapira
So that would.
00:24:53 Ben Shapira
Have to be.
00:24:53 Ben Shapira
Removed but yes, I.
00:24:55 Ben Shapira
Mean you're what you're probably talking about is a lecture that actually give my students every year to some really to scare quite a bit of them about using systems like Alexa and Siri and those types of.
00:25:05 Ben Shapira
Is a lecture that I give on something called single source attribution data. Yeah, so single source attribution data is the ability to attribute multiple sources of data to a single source.
00:25:16 Ben Shapira
And that single?
00:25:17 Ben Shapira
Source is your mobile phone.
00:25:19 Ben Shapira
If you consider the fact that your mobile phone has your SMS communication, your photos, your GPS tracking location.
00:25:26 Ben Shapira
Your search history, your bank information, your health information, your photos, along with you know, all the other different app information you have, health data, social data, social triggers. You know, we can get a really good understanding.
00:25:40 Ben Shapira
If we take that data.
00:25:42
And you would.
00:25:42 Ben Shapira
You attribute a.
00:25:43 Ben Shapira
100,000 people you know to your data set, we can make a really good statistical correlation over who your friends are and your political leanings and your purchase history and your, you know, your even like your holiday habits, for example. So yes, you you, you know, the more data you have the more.
00:26:03 Ben Shapira
Data sets you have from more the more sources you have, the data actually becomes can become more and more specific.
00:26:11 Ben Shapira
Only enough. In our case, we have to be really pragmatic about delivering our data in a way where we limit.
00:26:19 Ben Shapira
Possibility it.
00:26:20 Ben Shapira
Would be more of an issue for companies like OMD, who is a media buyer or media seller, or big corporations that have purchased multiple different data sets.
00:26:31 Ben Shapira
It would be more of an issue for them to be to be concerned about that eventuality. We're.
00:26:36 Ben Shapira
Obviously, selling our data mean we have it yet.
00:26:39 Ben Shapira
But that data will be sold to a very limited number of partners and obviously our data is one of those single sources.
00:26:47 Ben Shapira
It's a bit of a minefield, but this is, I mean this is once again the reality of technology when it comes to the finance space.
00:26:54 Ben Shapira
Is also the reality of technology when it comes to the healthcare space. It's also no different than talking about our artificial intelligence as it relates to finance or even just the general human equation.
00:27:05 Ben Shapira
The reality is that the technology is itself quite benign. It's how we choose to use it. Dinero is very mindful.
00:27:14 Ben Shapira
About consumer privacy and consumer protection, we do not have an FSL. As in we don't have a financial licence.
00:27:21 Ben Shapira
We don't have a credit licence so we will never provide financial advice. We will never compare your credit cards and offer you a better, cheaper rate, right. We're not one of those comparison services.
00:27:33 Ben Shapira
Our purpose for being is specifically to allow consumers to engage with financial professionals and get the absolute best advice they can get for their financial well being.
00:27:44 Ben Shapira
So we are a facilitator more than anything else. If we can serve that purpose to the best of our ability for the long term.
00:27:53 Ben Shapira
And that won't just necessarily be in the lending space, but in multiple different financial verticals.
00:27:59 Ben Shapira
Then I think it as a business. We've done our part. We want to make life better for the consumer, make life better for the financial professional, so that all ships rise with the tide.
00:28:10 Dr Genevieve Hayes
When you were describing the use of data in dinero and how other companies make use of data, I'm guessing that the sorts of techniques that would have been used in order to do that analysis would be just your standard data analytics. Some machine learning. Are there any other key technologies that you.
00:28:30 Dr Genevieve Hayes
Imagine yourself incorporating into something like genera.
00:28:34 Dr Genevieve Hayes
Or some other financial services fintech companies incorporating into their products in the next three to five years.
00:28:44 Ben Shapira
I would definitely say that artificial intelligence and the development of curated algorithms will will definitely be the future. The weighting of different aspects of purchase intent and purchase history can, you know, often give you different results and machine learning can help you to a certain degree. AI will help you to a certain.
00:29:04 Ben Shapira
Degree machine learning is in large part basically large data set decision engine. While artificial intelligence is sort of taking really trying to to take a contextual perspective over a consumer's purchase behaviour. The problem with all of that is the lack of the human equation with any of all.
00:29:25 Ben Shapira
With any of this technology, I don't personally see.
00:29:28 Ben Shapira
The any point at which the technology will replace the human contact, there will always be those who want just the cheapest product and cheapest price and.
00:29:38 Ben Shapira
There's a market for that.
00:29:39 Ben Shapira
Absolutely. And I have.
00:29:40 Ben Shapira
No problem with.
00:29:41
That but I.
00:29:42 Ben Shapira
Think the larger proportion of the market?
00:29:44 Ben Shapira
Want to find a way to get the most?
00:29:48 Ben Shapira
Out of their.
00:29:48 Ben Shapira
Dollar that they can.
00:29:50 Ben Shapira
To have the largest, more most diverse portfolio they can have when they retire and that having technology as a backbone.
00:30:00 Ben Shapira
To facilitate that and to allow them to more efficiently engage with financial professionals, people who will understand their context, how when do you want to retire?
00:30:11 Ben Shapira
How many kids are you going to have? Where do you want them to go to uni? Has something changed like do you want to change your plans? That context requires human interaction.
00:30:21 Ben Shapira
So I think there's.
00:30:22 Ben Shapira
Always gonna be a balance to be played, but technology will create more efficiencies, more economies of scale, which I think will only benefit the the market overall.
00:30:31 Dr Genevieve Hayes
Few minutes ago you used the term curated algorithms. Could you explain what you mean by that?
00:30:37 Ben Shapira
So the creation of algorithms where you are creating it specifically to look for a specific outcome and also that the weighting of the different aspects of your algorithm will need to change over time to adjust and fine tune it. I'm using the term curate. It may not necessarily be a proper data science term.
00:30:58 Ben Shapira
So you'll have to forgive me if that's the if that's the case.
00:31:01 Ben Shapira
But from our perspective, from a marketing perspective, that's a terminology.
00:31:04 Ben Shapira
That we use.
00:31:04 Ben Shapira
We we curate algorithms to allow us to extract.
00:31:07 Ben Shapira
What we're looking for?
00:31:09 Dr Genevieve Hayes
Sorry, I'm still trying to get my head around because I haven't come across the term curated algorithm before, so what's what sort of outcome would you be looking for in a marketing context?
00:31:18 Ben Shapira
OK, so if we are a company, let's say a.
00:31:20 Ben Shapira
A new a brand new company that is looking to understand market.
00:31:25 Ben Shapira
Right. There are a number of different.
00:31:27 Ben Shapira
Algorithms that can help you do.
00:31:28 Ben Shapira
That but if.
00:31:29 Ben Shapira
There are specific niches that you're trying to find. You will need to adapt that.
00:31:34 Ben Shapira
Those sort of stereotypical algorithms to help you rebalance the different aspects of your algorithm which are giving you your output. In other words, the weighting of your.
00:31:46 Ben Shapira
Algorithm will need to change so.
00:31:48 Dr Genevieve Hayes
You have fine tuning the assumptions.
00:31:50 Ben Shapira
Absolutely. Sometimes we have to do that earlier than than what we would necessarily like. So that's what we call curation.
00:31:58 Ben Shapira
Eventually, that sort of mitigates itself as.
00:32:01 Ben Shapira
You have more and more data.
00:32:02 Ben Shapira
That tends to balance out with the fine tuning itself.
00:32:06 Dr Genevieve Hayes
How do you see technologies like generative AI impacting the financial services industry?
00:32:13 Ben Shapira
I mean, I was saying for the context of your customer is an ever changing ever moving equation part of the equation and I don't think that artificial intelligence will solve for that no matter how many questions that you teach it to ask.
00:32:29 Ben Shapira
That being said, there are specific parts of the financial community that can really benefit from AI, so something like there's another part of the market called Sherlock.
00:32:39 Ben Shapira
They're relatively new to the market and what they do is they look for businesses who are, it's are individuals who are more likely to.
00:32:49 Ben Shapira
To leave a broker because they're looking for a cheaper price, right? They're looking they're they're, they're refinances ready or they're paying too much and they're right for picking another company called Stay or Go, which is a competitor of, of, of Sherlock, which does the same thing.
00:33:03 Ben Shapira
So it's businesses like that that can leverage artificial intelligence to identify customers who are more prone to churn in a in a brokers book and search for offers that are potentially better for them if their only concern is specifically the bottom line in the financial equation.
00:33:23 Ben Shapira
And offer them those opportunities.
00:33:25 Dr Genevieve Hayes
I'm just imagining the how I would build something like this, so yeah, it's basically a churn model then.
00:33:31 Dr Genevieve Hayes
In some sort of search model and then.
00:33:34 Ben Shapira
And then you would want to reengage the broker if they.
00:33:37 Ben Shapira
Would be willing to pay for.
00:33:38 Ben Shapira
It or potentially you go direct to consumer.
00:33:41 Dr Genevieve Hayes
To go back to data security, we're talking about ANONYMISING data in order to perform analysis before. But there's also the matter of data breaches. The more data financial Services company collects, the greater the chance there is that there's going to be.
00:34:02 Dr Genevieve Hayes
Data breach. You know Allah, the recent Optus breach? The medibank.
00:34:08 Dr Genevieve Hayes
Breach Australia has had some really bad luck this year, I think.
00:34:12 Ben Shapira
Yeah, I just got.
00:34:13 Ben Shapira
Caught up in the most recent latitude 1.
00:34:15 Ben Shapira
The companies, who are largely being targeted, are ones that that have transactional capability.
00:34:21 Ben Shapira
So they are the primary target. I'll tell you what we do to try.
00:34:25 Ben Shapira
And mitigate as much risk as possible.
00:34:27 Ben Shapira
And we are, as far as our insurance company is concerned, the lowest cyber security threat. It's not to say that we are in a target.
00:34:33 Ben Shapira
Potentially, but what we do is we take as much steps as we can in terms of security as a matter of fact, we've just done our gonna pen test.
00:34:42 Ben Shapira
So it's a what's called penetration test. It's a company that we've engaged to essentially trying to hack our app and they provide us with results we remediate those results, get another test done to ensure that we have addressed those particular considerations.
00:34:56 Ben Shapira
And as an output we get certain different types of certifications. So the one that we're going for is called SoC 2 which is bank global security.
00:35:05 Ben Shapira
So for us.
00:35:06 Ben Shapira
We take a few, a few different steps. One we've chosen. We do not allow transactions in our House.
00:35:12 Ben Shapira
So right then and there, many hackers will look at our data and say, well, it's not really worth as much to us because.
00:35:18 Ben Shapira
It's non transactional.
00:35:19 Ben Shapira
2nd, we never Store bank login details. When you connect your bank account to our app for the first time, you will use your bank supplied login.
00:35:29 Ben Shapira
Details we do not store them. What we get from our data partner is an alphanumeric key that identifies our app, your particular account within our app and access to your financial data.
00:35:42 Ben Shapira
And from there it allows us to update our.
00:35:44 Ben Shapira
Data every 24 hours.
00:35:46 Ben Shapira
But it ensures that we are at no point.
00:35:49 Ben Shapira
Reusing your bank login details. In fact, if you change your bank login details, you will need to re log in to your bank account through our app to continue being engagement.
00:35:59 Ben Shapira
We also encrypt our data in a different way. Our financial data is encrypted in a separate database to our customer like to our consumer.
00:36:08 Ben Shapira
Data. So anything that's identifiable to you as a consumer is in a separate database to your financial data, and it's only through the decryption within the core software of the app that the output comes to the API.
00:36:23 Ben Shapira
If you try and steal that information, you're left with two databases that are separately encrypted without a key to decrypt.
00:36:28 Ben Shapira
So we try and add these sort of multi layer approaches to things to ensure that at all times we're taking the most pragmatic approach to to data security.
00:36:38 Dr Genevieve Hayes
Do you think this sort of approach will be adopted by a lot of other financial services firms or tech companies going forward?
00:36:46 Ben Shapira
I certainly hope so. I will say the government is very stringent on their cybersecurity requirements, especially for open banking and our data partners are also very stringent on their requirements.
00:36:57 Ben Shapira
So we have to go through these pen tests to prove that our code is up to standard. It's not just our code, but our hosting infrastructure.
00:37:06 Ben Shapira
So, for example, no financial data ever leaves Australia. Our apps can only be downloaded in the Australian App Store, so everything is geofenced to Australia to ensure that you know someone outside the country isn't trying to hack in and get out gain access.
00:37:22 Ben Shapira
But the the government, the industry as well as the fintech partners themselves are all you know, very well, motivated to ensure that we are taking all of this, you know, in stride that you know, we understand that breaches are going to happen.
00:37:38 Ben Shapira
They tend to happen to businesses that have.
00:37:41 Ben Shapira
Have spent a lot of time in business and are taking a little bit more of a lacks perspective around security, potentially in favour of features or benefits.
00:37:51 Ben Shapira
We will not.
00:37:52 Ben Shapira
Be one of those companies for us.
00:37:54 Ben Shapira
It's really clear that trust is a primary key to any of this fintech future. If that trust is breached, we all are are struck down by it.
00:38:05 Ben Shapira
So we're we all work together. I mean, all of all, most fintechs know each other. We all typically work out of a similar hub like we were. We've been working out of stone and chalk.
00:38:14 Ben Shapira
For the last of a while.
00:38:15 Ben Shapira
And it's nothing but fine.
00:38:18 Ben Shapira
And you know, we all.
00:38:19 Ben Shapira
Have to deal with these sorts of very similar problems, so we all learn from each other. We all take cues from each other and in some cases we actually work together.
00:38:27 Dr Genevieve Hayes
Well, I mean the fact is if one fintech gets breached, it causes, it casts a shadow over all the others. So it's in everyone's best interest for you all to be safe.
00:38:37 Ben Shapira
Absolutely. And. And look, we're dealing with, we're dealing with an industry where.
00:38:40 Ben Shapira
There more than 95% of our consumers don't understand the technology that we're implementing. So it's in large part a trust issue.
00:38:50 Ben Shapira
We need to provide as much assurance as a much proof to consumers that we are following, if not exceeding, best practise for them to want to use our software.
00:39:01 Ben Shapira
And to continue to use our soft.
00:39:02 Ben Shapira
To where? So it's incumbent on us to be as fully transparent as possible. Our pen test results we will publish to the to the general.
00:39:10 Dr Genevieve Hayes
And what's next for you? And for Deniro?
00:39:14 Ben Shapira
There's two aspects to our growth plans. One is geographic growth and these once again will be geofenced by region.
00:39:21 Ben Shapira
Obviously New Zealand would.
00:39:22 Ben Shapira
Be our first sort of cap off the ranks in large part a lot of our partners are here are also there as.
00:39:28 Ben Shapira
Well, Canada UK have very similar banking structures. So for us those, those countries are very appealing to us. And then there's also the opportunity to work in multiple financial verticals.
00:39:41 Ben Shapira
Currently, we're we're working primarily with mortgage brokers and lenders, but we also had lots of interest from financial advisors.
00:39:49 Ben Shapira
We've had interest from business lend.
00:39:52 Ben Shapira
Being we've had interest from you from insurance companies, we've had interest also from other financial markets who really just they want that sort of that, that that means of communication with their consumers that they currently are not being offered by their by the software vendors in their in their.
00:40:09 Ben Shapira
Industry right now.
00:40:10 Dr Genevieve Hayes
So is there anything on your radar in the AI data and analytics space that you think is going to become important in the next three to five years?
00:40:19 Ben Shapira
I mean, I think we've seen the growth of AI insofar as what we've been allowed to see. I mean, the reality is what we see is probably 5% of what's actually been designed and built.
00:40:28 Ben Shapira
I think what we're going to see is a large scale global race towards AI and you know that can have really positive influence on, you know, any sorts of industries by, you know, allowing for rural communities to gain access to, you know, high quality medical care, for example, to fully level the playing field when it comes to access to education, for example.
00:40:50 Ben Shapira
Where we can also see problems is, you know in our media industries, in our political political spheres, by seeing a lot of AI being used for nefarious purposes.
00:41:03 Ben Shapira
When it comes to specifically to the financial side of things, there are once again positive and negatives by having AI driven services, we can see for efficient ability to say for example, find a cheaper, better price to find a product or service that meets your specific needs. If you are so inclined and educated.
00:41:23 Ben Shapira
To be able to provide those same services to people regardless of geography, regardless of income and regardless of education.
00:41:31 Ben Shapira
Where I see a problem in it.
00:41:33 Ben Shapira
Is stuff like.
00:41:35 Ben Shapira
When we see over the last few.
00:41:36 Ben Shapira
Years, for example, like robo training.
00:41:38 Ben Shapira
So using AI to understand and preempt training practises for stocks by consumers, because now we see consumer trading of stocks being ever increasing.
00:41:51 Ben Shapira
Volume in the mark.
00:41:52 Ben Shapira
Get, but they can't.
00:41:53 Ben Shapira
Train at the same speed as say for example you know a Morgan Chase for example.
00:41:58 Ben Shapira
So AI playing a large part in.
00:42:00 Ben Shapira
That AI may also be able to give us.
00:42:03 Ben Shapira
Better super results for our retirement savings.
00:42:07
So there's a.
00:42:07 Ben Shapira
There's a variety of different ways that can be.
00:42:09 Ben Shapira
Really positive and really, really, you know, problematic.
00:42:12 Ben Shapira
We have to take a step back and look at AI to see what it can do and what it might do. I think we need to engage those who are highly educated in the space.
00:42:24 Ben Shapira
We need to have.
00:42:25 Ben Shapira
Of much broader discussions, both within industry with the general public and through the government to understand exactly where our opportunities lie and where our potential risks.
00:42:40 Ben Shapira
Might might be.
00:42:42 Ben Shapira
And to create a framework to create some some fencing.
00:42:45 Ben Shapira
Around the use of AI in specific circumstances.
00:42:49 Ben Shapira
We do have.
00:42:50 Ben Shapira
To be mindful, though, that there are going to be other countries that don't have those fences that don't have.
00:42:55 Ben Shapira
Of those barriers, countries that will say we don't care what the cost is, we always.
00:43:01 Ben Shapira
Want to be #1?
00:43:02 Ben Shapira
We will have to deal with that as well, but I think if we can look at, say for example how Australia navigated the financial crisis.
00:43:12 Ben Shapira
In 2008.
00:43:13 Ben Shapira
In large part that regulatory framework.
00:43:15 Ben Shapira
That regulation of the banking industry is what saved Australia from the worst of the outcomes. I think we need to take the same.
00:43:23 Ben Shapira
Sort of look at.
00:43:24 Ben Shapira
At AI and say, how do we use it to our benefit as a community to get the best results for all of us?
00:43:30 Ben Shapira
But also, where are risks and where do we need to protect ourselves so that we can be ready for what may come?
00:43:39 Dr Genevieve Hayes
Do you see any countries leading the way in regulating AI?
00:43:44 Ben Shapira
I don't think there's any one particular country at the moment because it's so new. I know the US has had discussions recently in open Senate hearings with industry special.
00:43:55 Ben Shapira
Lists I would say from a privacy perspective, Europe is far ahead of the curve of anybody else. The privacy regulatory systems they have in place are quite stringent, but I would say that that that this whole AI thing is still so new to the public. The governments of the world have not had to deal with this in.
00:44:16 Ben Shapira
An open way so far.
00:44:17 Ben Shapira
And now they're kind of being forced to. It's actually a good thing that that the, the, you know, the covers are being lifted quick.
00:44:24 Ben Shapira
I said I suspect over.
00:44:25 Ben Shapira
The next 24 to 36 months, there's going to be a lot of discussions, a lot of open discussions, and they'll be really interesting to see if the world can create an agreed framework like we have for, you know, for nuclear weapons or more. Biotechnology like, you know, like human cloning for example, or genetic research.
00:44:45 Ben Shapira
We can take those same sorts of cues as a global community and recognise its potential and its potential harm.
00:44:53 Ben Shapira
And create an agreed framework that we.
00:44:55 Ben Shapira
Have with those types of technologies, I think we could actually look at AI as a positive.
00:45:00 Ben Shapira
Overall influence, then, then, then a negative.
00:45:04 Dr Genevieve Hayes
And what final advice would you give to data scientists looking to create business value from data?
00:45:10 Ben Shapira
School will only teach you so much and I don't mean that in a way that is disparaging in schools.
00:45:15 Ben Shapira
Obviously I teach at Swinburn. Universities will lend large part, teach you the theory, get your hands dirty, do a lot of independent learning study on your own as well as in the.
00:45:26 Ben Shapira
Classroom, but also try and think outside the box, right?
00:45:30 Ben Shapira
I mean put.
00:45:30 Ben Shapira
Yourself in the shoes of the very.
00:45:32 Ben Shapira
People that you are trying to understand, it's through that process that you will be more effective, more efficient because data analytics is all about understanding context.
00:45:46 Ben Shapira
That's what you're trying to understand. So put yourself in the shoes of those that you're trying to to understand.
00:45:52 Dr Genevieve Hayes
Well, listeners who want to learn more about you and to learn more about De Niro, where can they go to get in contact?
00:45:59 Ben Shapira
So I'm happily, we'll speak to anybody over LinkedIn. You know, my profile is open. Happily, if you want to come and join my class at Swinburn, I'm sure the university would.
00:46:09 Ben Shapira
Love that dinero it.
00:46:12 Ben Shapira
You know you can find us online dinero dot app and you know, in, in general, I do lots of speaking engagements. I I'm more than happy to talk to people.
00:46:21 Ben Shapira
A variety of different issues. I'm kind of at the point in my life where I'm really keen on giving back so I do guest lecture spots from time to time.
00:46:29 Ben Shapira
As well.
00:46:30 Ben Shapira
Happy to speak.
00:46:31 Ben Shapira
To anybody on any topics that they like.
00:46:34 Dr Genevieve Hayes
Fantastic. Well, thank you very much for joining me today.
00:46:37 Ben Shapira
My pleasure.
00:46:39 Dr Genevieve Hayes
And for those in the audience, thank you for listening. I'm doctor Genevieve Hayes, and this has been value driven data science brought to you by Genevieve Hayes Consulting.
Creators and Guests
