In this episode, Dancerace CEO Elliot Avison and Codat COO Alex Cardona discuss the future of open finance, unpick the foundations of great customer experience and explore how data extraction and similar technologies can give lenders the focus they need to level up their operations.
Please note that this interview was recorded shortly before the Coronavirus pandemic.
We hope you enjoy it!
Podcast, July 2020
Elliot Avison, Dancerace & Alex Cardona, Codat
Elliot Avison, Dancerace: Hi Alex. Can you introduce yourself and what you do at Codat?
Alex Cardona, Codat: Sure, I'm Alex Cardona, the COO here at Codat. I'm broadly responsible for everything non-technical within the company.
We're a universal API for small business financial data, and so our core product is an API for people to access data on their small business customers. My role is to ensure that the software developers who make up more than half of our team have everything they need to do what they need to do and then I’ll go to market and put this technology out there.
EA: Great. For those of you listening, my name is Elliot Avison, CEO of Dancerace and I have been leading the company now for coming up to five years.
My goal is to re-imagine and refocus our business to working on partnerships with the likes of Codat to expand our products and create a new world for the invoice finance and factoring industry. We’ll do that through both the mature products we already have but will also come up with new, interesting ways to do that and hunt out new markets and new opportunities in the exciting fintech market.
Alex, what's Codat's story? Where have you come from and how did you got to this point?
AC: Sure, so as you might imagine, you don'tof come out of school and say, yes what I really want to spend my life doing is standardising integrations with accounting software. There’s a bit of a story about how we got into this.
The three of us set the business up just over three years ago now, at the beginning at 2017, having all worked together for a number of years at Market Invoice – now Market Finance – here in London. For those of you who don't know who that is, it's an online invoice finance player, digital first, that started off with selective invoice finance and then moved to whole turnover.
At Market Finance, we were thinking about the best ways to make the application flow as simple as possible, and a key stumbling block was the process of actually getting ledger information from new customers coming onboard – from the systems that the small businesses were using to manage their ledger. How could we see what payments are made using these systems and generally get an understanding of what their book looks like?
We therefore decided to hook directly into those systems that the small businesses were using. In doing so, we build a few integrations: Xero first, which gave us about 25% coverage of the market, then Sage 50 which probably gave us another 20-25% coverage, then KashFlow, that probably gave us 5% coverage of the market.
Needless to say, every time it's the same amount of technical time to build each one of these integrations. It's hard, complicated work, to build up an integration, so Market Finance turned around and said hold on, this isn't what we do; we don't build loads of backend data applications. Wouldn’t it be great if there was a piece of kit where you could hook in once, all the pain was abstracted away and you could integrate with all of the underlying packages – no matter what the small business was using on the other side?
So, that’s what we did, just over three years ago. Now we've got about 60 clients in a bunch of different sectors, but a large proportion still work in capital finance, where we help them pull in data to make different credit decisions, while reducing the pain on both sides in terms of administrative costs.
EA: Yes, it's how we met, wasn't it? This problem has existed for the invoice finance market; which is the more that you know, the better you can fund or the less problems can occur throughout funding.
It's such a key part of what we do. If our backoffice systems don't have data then we're blind. Back in the noughties, we actually tried to build a sage integrator and it was like pulling teeth, so I can see the issue at Market Finance being that your focus was supposed to be on lending money – not spending time building integrators.
Now that you've got all of these different accounting packages integrated to your system, what other use cases are there for your extractors, other than just in the ABL and invoice finance space?
AC: So it's a broad collection of different use cases. We generally split them into two different buckets. One is pulling data from the small business to make some kind of decision – we work a lot within credit insurance, for example – as well in more straight-forward term lending.
Then, on the other side, we push a lot of data into accounting packages. For any financial services product, as a small business, I’m providing my data over to you, so it would be really helpful then if I get my data pushed back in return. For example, we work with some of the challenger banks now, whereby we take the data from the bank and push it into the small businesses' accounting package, so they can do their end of month reconciliation, for example.
This is exactly where we imagined ourselves to be. What we really believe is that if we’re just the pipes ensuring data can flow from one place to another as easily as possible then we're providing the most value, because as we well know, a huge part of the problem is just making sure that the right data gets to the right people.
EA: Yes, it really is, and I suppose how people view data extraction or data transparency is changing as well. People are probably requesting more and more from you now that the market has suddenly realised that open banking is there, which means open accounting is a thing.
People have become more comfortable because of those larger scale technology shifts, so that the adoption rate of products like yours and being comfortable about you pushing data into systems must be an easier sell than before.
AC: Absolutely, and I think that in invoice finance in particular, people have always known there's a possibility of getting this done because the information is in the machine-readable format within the small business.
You have a system like Dancerace that allows you to look at all of this internally; it's just matching the two up, and I think the really key shift here has actually been the rise of the cloud and the opening up of APIs on the accounting packages. There’s been a massive impetus into making it not just possible but worthwhile, and what we hear a lot from people is that they're surprised by how quickly they can get up and running with accessing this data, simply because customers are very used to sharing their data because we live in a world where authorising and consenting to other people seeing your data is something that we probably all do multiple times per day now.
EA: Yes; only a few days ago I was connecting all of my credit cards – that sounds like I've got loads – up to my main Lloyds bank account and yes, it makes a massive difference being able to see it in a list of how much money I have here or there. Previously, I had to have five different logins, but now it's all there in one screen.
From a personal perspective that’s been significant but in the world of business, where you might have multiple bank accounts or different accounting systems for different departments, all of this data sharing and open and transparency is going to save time and money and headaches.
On that, what do you think the government is going to do next following open banking? It was quite a shock that the adoption rate was so quick, there, so where next?
AC: It's a great question and I think people are looking around for precisely this. The FCA recently put out a call for input actually, on this. They said that open banking has been well received by the market; customers really like being able to share their data with other people; we can build more valuable products for consumers off the back of it. They've called this Open Finance, and suggested that pensions data, insurance data and utilities data would similarly be available to be shared through an open model.
We're big advocates of this being applied to small businesses as well, and we are very strongly of the opinion that when it comes to small businesses, you can't just have the bank statements as that’s only half the ledger. The fundamental reason that small businesses have accounts is to provide context to their bank transactions. Whereas as a consumer, I might be happy with a bank statement, I understand most of what that looks like, small businesses are very complicated and therefore, you must have a lot more rich data around everything that the small business does. Wee’re therefore going to be talking to the FCA about that.
The Bank of England interestingly also put out a paper in Autumn last year, proposing what they called a Commercial Credit Passport, which is essentially going to be one standardised method to share all the information that's going to be needed to underwrite a vanilla small business loan. The speed of regulation is such that it will probably take quite some time but we're pretty confident that good things are going to happen from all of this over the next few years.
EA: Yes, I think it's really encouraging, and I don't think it’s necessarily about playing to the fintech model. Actually, it’s just that engagement with technology in general is just going up.
Millennials are going to be the next business owners, and they have a very different perspective on transparency and sharing, sometimes obviously over-sharing on the likes of Facebook and what-not, but at the same time this will have an impact on their business decisions. It's ingrained in the way people think now.
From a personal perspective, I would happily put Dancerace's information into a passport, so that, should I need to, it becomes easy for us to go to an underwriter should we want to borrow money or go through invoice finance. This would make life so much quicker, and that speed and access is really where the fintechs have an edge right now.
This is one of the main reasons why we were looking at data extraction with yourselves: technology like Codat, bridges the gap between the fintechs and the more classic players in invoice finance.
How do you see data extraction making life easier for invoice financiers and their customers?
AC: As you said, speed is really the number one thing. Someone put it well to me the other day; they said that applying for finance shouldn’t feel like a second job, and I think that's entirely right.
We work with a couple of the Tier One banks that, having used Codat have managed to take their average time from initial contact with an interested client through to funding from 90 days down to 42, just by automating this piece.
Where previously they would have had to send someone out on the road to meet the finance team to do an audit on site, they can now do a desktop audit. That means they don't have to get someone booked in; they don't have to take up the finance team on the business' side by going through all of that. They can just share the information and the bank can the survey from their own office.
That seems like a prosaic, minor thing, but in practical terms the time gained is and obviously reduces costs. Those people who were going out to do all the audits can now spend their time, thinking about those second-order analysis questions, as opposed to the simple collection of ledger information.
EA: You’re spot-on. At Dancerace, we’ve launched f3 to do just – an onboarding tool that uses Codat data extraction to offer a desktop audit for lenders.
I suppose the power of your extractor technology is that every time you get an extraction it's that same audit again and again and again. It's not a different process that you run through; it's as if you've got into your car and turned up at the customer’s offices every half an hour. That's extremely powerful, and is made possible for none of the cost that would normally be associated with that level of oversight that you’d have to perform manually, so you're absolutely right.
AC: And then one of the nicest things about this regular automated is the response that we see from relationship managers, auditors and underwriters, who say they can spend time thinking more seriously about the client's business, as opposed to getting in the car to go over there. That makes their job more enjoyable, which is, I think is a very real piece of this.
EA: Yes, and I think it also leads to a better customer experience, and that customer experience will be shared to other businesses and potentially the word of mouth growth of this industry will benefit enormously from it.
What I know from our customers, having spoken to a good many of them about this, is that there's been a unanimous concern about automating decision-making in this sector because so much of the decision making that is done is still very manual. Our customers don't want to change that, because there’s something to be said about experience and about people who can look someone in the eye and sit across the table and read a new customer’s body language. What you don’t want to be doing, however, is spending weeks and weeks capturing and trawling through data and then getting to a point where you decide you won’t proceed.
By doing this and increasing the speed up front on the data capture, you can get to those face-to-face conversations quicker. You can make those more human decisions without any of the huge operational overheads that are currently mandatory.
AC: Yes, and I think that's spot on. Something that you said to me a few months ago, I think, was that the Dancerace model is really for pushing a return to relationship-driven lending. That’s very real and actually a relationship comes from being properly able to understand a customer.
Partially that’s made up of being able to see all their data easily, but mostly it comes from having a clear understanding of them as people and where their business wants to go and that can't all be made up from data, but you need to make sure that you've got all the data so that you can spend that time, that's the piece I really like.
EA: This industry has actually got a lot of good new stories about helping businesses and I wonder how many businesses have slipped the net in terms of accessing finance because too much time had to be taken up.
I think what we're discussing here is good thing for the marketplace and it actually does return back to making the human factor more impactful and less about admin and hard work.
My next question for you, then, is slightly more complicated and tricky. One of the biggest challenges for the invoice finance market is tackling fraud and making sure that business continuity or lending continuity is not interrupted by fraud; people playing around with accounts, raising fresh air invoices, etc.
How can data extraction help towards reducing fraud and other issues that can arise during invoice finance?
AC: Yes, it's a very good question, and the way it plays out is slightly dependent on the way people approach the underwriting process.
The simplest way, I think, to think about it is about getting closer to the actual internal workings of the business. Rather than waiting until the end of the month – when people could be up to all kinds of stuff – to get a new set of data across, you can get it on a daily basis.
We have people who pull their customers' data every minute, which I think might be overkill sometimes, but it's definitely there and can raise interesting points. People have been looking at, for instance, the time that people make changes to invoices and whether this is an indicator of fraud. If you can see that someone's put in an invoice on a Sunday night at 3 in the morning, what does that mean in terms of the validity of the invoice?
It might go either way, but having more data allows you to do is ask those questions. The other element is simply about the volume of data. When you're looking at the full set of information contained within someone's accounting package, you might have 20 years of invoicing information, as opposed to the one snapshot that you currently get. Instead of a distilled view of how that book has changed over time, you get line item-level information, going all the way back historically. This means you can see patterns within this.
Again, you want to make sure you're understanding those patterns – which takes a human eye – but at least you have that information, and the customer can't be brushing up the data in order to send it over to you because you're getting it in its direct format. It's a huge amount of work to just do accounts normally, let alone fabricate a full set of fake historical accounting data.
EA: Yes, it's a complicated thing isn't it? You don't want this to seem like a surveillance product, but indirectly it can provide a level of transparency that does allow a financier to have quicker or more deeper insights.
Human behaviour is not random, so when people do commit fraud, the most acute fraud detection systems are checking for patterns in data. Being able to see what happened yesterday versus today, checking what changed, what hasn't changed, what changed outside of the norm; comparing this to many, many years before to spot seasonal trends; all of this is exactly what you need for a fraud detection system.
The most comprehensive banking systems have used these sort of algorithms to detect fraud and the only way they can do that is transparency. If you don't have the data, you don't have the mechanism to detect when something is not going the way you'd hope.
I think compared to where we were and where we have been for many, many years, this sort of information will be infinitely better than it has been before. The businesses we work with have got very comprehensive mechanisms and processes internally to capture this information but they're working with what they've got. If we triple or quadruple what data they've got access to, it’s only going to become easier, obviously as long as it doesn't increase noise.
I think the proof is in the pudding with this. While I can imagine it being very difficult to commit fraud, I can imagine it would be much more difficult to manipulate my entire accounting system rather than a CSV upload which is sent via email.
Moving on; how do you see data extraction evolving in the future, and do you feel it will become the de facto way of transferring data between a financier and a customer?
AC: So, I would say, not just absolutely, but I believe for the newer businesses that are coming out, this is the way in which they demand their products to be served to them.
What we see increasingly in the micro-lending market – loans of up to £25,000 or £50,000 – is an expectation that I should be able to share my data as part of the online application process, and I will actively choose a more expensive funder if I can do it by sharing the data automatically, as opposed to manually collecting it, sending it, off and waiting in the dark to receive an idea of what the pricing would be. If I know that I've done my piece, and that's the full set of information that's going to be asked for, then I'm much happier to wait for several days a week to get that information back and that's definitely true in other areas as well.
Small businesses are now very likely to have multiple integrations with their accounting package for multiple different use cases. That might mean integrating with their point of sales system or with their inventory management system. There are a bunch of different systems that they will use internally that rely on integrations now, and we believe that if people don't offer this as a standard part of their product, they will increasingly lose out on customers because it is just so much simpler.
EA: Yes, we've seen situations where customers are driving the demand for data extraction, simply because of the operational overhead it resolves. Having to create a CSV containing your ledger details or customers takes time and as you said earlier, it shouldn’t be a second job to receive finance, because sometimes the reason why you require financing is you simply don't have the time to chase new business or things are getting a bit tricky. There's no point tying yourself up in an operational overhead. It's a catch 22, it has to be an easy process.
AC: Yes, and the other side to this is there are certainly some people who aren't super comfortable with it, who haven't built their business around this type of online system yet, but what we've seen there is that if you explain the benefits to the customer and explain why you're taking the data, what you're going to do with it, actually they are more than happy to do it.
It's just an education process; it's not that people have some fundamental objection to sharing this data because it's the same data that they would have been sharing historically anyway, but people do need to get comfortable with the mechanism, sometimes.
EA: Yes, and in the press there have been worrying stories about data sharing and about who can see information. I think there are some larger businesses out there that have been irresponsible with the way they’ve used data, but as providers of this software, we've got a particular use case with a particular requirement on the information, and the value therefore for the customer should be quite transparent.
The faster you get your information to a financier, the faster they can give you the money you require, and that can be transformational to your business. I know that's a reasonably vague overview of the process, but it does always fall back down to that eventual requirement, and I think that we as providers of this sort of software need to prove ourselves as being responsible with this information, because it is key to these businesses.
I certainly know that it’s very important to both Dancerace and Codat that we get this absolutely right and that customers' information is protected and secure.
In terms of shifting customers’ expectations, though; these things take years. It's social change and we shouldn't pressure businesses who are going through this transition but we should provide a solution for those who are ready, and as we both know, it does work and it does speed things up. It's key for us to be patient with this transition process.
I'll give a use case myself. At Dancerace we used Sage 50 for many, many years and last summer we finally moved over to Xero. It's opened up enormous doors for us. We’re a small business with quite a simple proposition in terms of our accounts, but we're using apps now; we've integrated Codat to our systems, and yes, it's a completely different world. My accountants and bookkeeper took a little while to transition, so we're not unrealistic about how the world that we're trying to create takes a bit of a leap. The positives definitely outweigh the negatives, however.
AC: That's precisely it. We have to embrace change but not expect the world to move at the drumbeat of constant change, because actually people have lots of very good reasons as to why they move at the pace that they do. All we believe is that you should make both channels available to people.
EA: Yes, and at the end of the day, these businesses are trying to make money, they've got to prioritise products. Going through these changes can be challenging, so it's just about supporting both scenarios, That’s why we’re offering Codat extraction to our customers when they're ready. It’s a simple integration that minutes and is just there and waiting in our customers’ systems.
Moving on; I'm personally really interested in how we can create and leverage some of the work you're doing in terms of pushing data to customers. That's really interesting because financiers currently spend a lot of time doing duplicated work. Take pushing allocations, for example: if the financier is actually doing some of the work for their clients on their system, or they're the first to find out and then they have to send information to the client, who then have to perform allocations on their system, that can be really time consuming. It would be such a massive win for this industry if we can leverage data extraction in the ‘push’ direction.
Can you give me a bit of information on how this works then, so how we might be able to use that?
AC: So, effectively the client account is being held on behalf of a customer. The way that this historically worked is with effectively double manual reconciliation, both at the financiers' end and the small businesses' end. What we can do is take that bank statement or list of transactions, and we can push it directly over into the accounting package as well as breaking down fees, which doesn’t sound like a huge amount but actually it's just another piece to take away in terms of administrative burden for the small business
I've always been a huge believer in invoice finance as one of the most powerful methods of financing, but the reason why invoice financing isn't a much larger market is because a lot of this operational cost is put on small businesses. If you can provide data back to them and reduce that internal burden, I think actually there's a big opportunity to increase the reach of the market.
We've seen big growth in this market over the last ten years, and I think that we'll see only increased growth as more and more of these tools become available precisely to alleviate this burden.
EA: That's music to my ears, because what we're doing is constantly innovating the technology to reduce the operational burden on the financier.
If the financier can start spending time on providing additional benefits to their customers rather than just performing administration, then invoice finance becomes a service and not just a mechanism to provide finance. You can actually provide a service element to the customers and really deliver value back to them, which in turn can generate a new revenue source for a financial organisation because they can start saying ‘when you sign up to invoice finance, not only do you get the benefit of this, but you're also signing up to a technology platform that can really start speeding things up for you in your business.
That will lead to repeat business and the news will get round that invoice finance is no longer a process that’s time-consuming, costly and complex. It makes for a much more compelling sales pitch from the lender who, dare I say it, looks more like a fntech. Almost any one of our customers using our tools today can start really thinking about how they might do that. It opens a door.
AC: Absolutely, and I think that's going to be one of the most interesting things to play out over the next few years.
There are two pieces: Codat, which is representing the small business and providing a way to get that data in, and Dancerace, which is providing the engine within that financial service's organization. Once those two are tied together with a full, live, real-time understanding of the customer and their data, financiers can ask themselves how to expand the number of things they can finance.
With additional data you can move further up the receivables chain. Fundamentally, that might mean traditional or purchase-order finance can now be applied on a much broader scale. Increasingly, we see people looking at ecommerce data, for example; live data being pulled from the place that the transaction is being generated even before it's booked into the accounting package. There's a real wealth of information here, and invoice finance will be in a really interesting position to take advantage of those additional points of which people can be financed.
EA: Yes, you're right, and it's fundamental that you must have that visibility to the accounting information. As long as businesses are clear and they start using their accounting systems as that single source of truth for their business and their books, then yes, you're absolutely right, you can see their balance sheet and start identifying other forms of debt and collateral that could be financed. In turn, financiers can provide a real-time overdraft; a business finance solution that can be tailored to the specific business you're dealing with.
This will be an industry agnostic, shift, and yes, it's really exciting. I think our industry has got massive potential and I think we've got to work get rid of some of the PR issues of the past. We should be done with that and actually look at the future as a good news story for this country and for other countries supporting small to medium size businesses. Ours really is the only sector that lives outside of the borrowing concerns of the main banks and loans, and is therefore the solution to those sorts of pressures.
What are you working on next at Codat?
AC: Sure, so, we've got a long list of accounting packages. There are lots and lots of them and that's what we're really here to solve; to make it a no-brainer and ensure that everyone, no matter if they're using some kind of super sector-specific accounting package, is covered though the Codat system.
We're also looking at additional types of data; for example, making open banking data available through the platform. That's live today and you can see the different bank accounts from different institutions through the same authorisation process.
We want to do more of that and we want to start bringing in more card data, for example, along the same lines, and then its' about expanding into those other systems that businesses use, so that we can give people a better understanding of what that business looks like. That might mean their point of sale, their ecommerce shop, their inventory management; anything that allows us a better understanding of the whole of the small business, not just the core invoices piece.
EA: Have you considered going down the KYC route and anything like that where you could provide information or go through those channels as well?
AC: Yes, we have. It's an interesting one; the thing that we always come back to is that we call ourselves the data plumbers. We know that we're good at building these kinds of pipes and that's the bit we really want to make sure we nail.
EA: I think the take home message for me is that this is a forward-thinking product that improves the customer experience, and the financiers' experience. It should lead to better funding decisions and reduced risk, hopefully, and is a bit of a no-brainer.
I think between us we've considered it a no-brainer for a long time, and you know, it's really good to keep pushing and waiting for customers to become more comfortable with the concept because there is a lot of opportunity sat at the end of that and I'm excited to see it reach its full potential.
AC: Absolutely. I think we're at perhaps one of those moments, where we as an organization have only been around for 3 years, but we've been growing very fast.
We're now up to about 60 people and that's all driven by customer demand, so the speed at which people are taking this up and putting it in as a standard part of their products is really kind of wonderful. I think that we're going to move to automated data sharing being the norm within just a few years, so as you say, super excited by it all.