Daren Campbell, EY Americas tax tax technology and transformation leader, discusses the use of big data, robotic process automation, AI and other technology, in corporate tax strategy and compliance.
Transcription:
Michael Cohn (00:03):
Hi, welcome to On the Air with Accounting Today. This is Mike Cohn of Accounting Today. We're here today to discuss tax and technology. We're joined today by Daren Campbell, EY America's tax technology and transformation leader. Welcome Daren.
Daren Campbell (00:16):
Michael, thank you for having me. Excited to be here.
Michael Cohn (00:18):
Oh, it's great to have you with us. Well, I understand you worked with the emerging technology here at Ernst Young. Could you tell us about what you're seeing about the use of big data in tax technology?
Daren Campbell (00:28):
Yeah, definitely. This is an area that is really coming into its own. I think we've been talking about big data in tax for a while, but early on, 10 years ago or so, when we talk about big data, I would use a definition to say big data is just anytime you're having, it's big. Anytime you're having trouble understanding the details of the data. But most of what we were dealing with still would fit in an Excel spreadsheet. So wasn't really truly big data, but there's been a lot of drivers in the industry, both internally coming from the C-suite, looking to tax to be more strategic in a strategic part of business operations as well as coming from the government. So not as much in the US today, but around the world. We're seeing a lot of the tax administrations go digital that's requiring greater volumes of data than that, than has ever been required before. So when we talk about big data in tax, it is definitely something that is really coming into its time where tax professionals now need to handle data and volumes that they haven't had to handle before.
Michael Cohn (01:38):
Is it important for a company to have a digital tax strategy to deal with cross borderer taxes in the international? For example,
Daren Campbell (01:45):
Michael, it's interesting you talk about the digital tax strategies. I know on one of the prior podcasts you were talking overall about strategies and this is something that does carry over to digital tax as well. It is very important. We did a survey in 2022 where over 44% of the companies listed not having a sustainable plan as one of the largest barriers to tax, being able to deliver on their purpose and vision. And we did a similar survey at a client tax tech conference that we had about a year ago and only 23% of the responders indicated that they actually had a formal tax technology strategy or roadmap. And why this is so important is tax is one of the heaviest users of data in an organization. But one of the challenges, the big challenges we face is we don't own or control any of the source systems.
(02:44)
And so that leads us to a big challenge where we have data coming from a lot of different places, the data sources continue to change, but we don't have systems or even ecosystems that are really built specific for tax. So it's very, very challenging and we end up spending a lot of time just dealing with the management and manipulation of data. We've done a variety of surveys over the years and it's anywhere from 40 to 60% of a company's time. The tax department's time is spent just dealing with the data, so trying to collect data, validate it, manipulate it. Often the data that we receive is not in the format that we need it for tax based on the output of whatever team is already using it. So where we're getting financial data, we're getting data that is formatted specifically for financial reporting. We have to take that information and oftentimes we need to link that to data that's been omitted somewhere in the upstream process, meaning that the really the triggers that determine the tax stream, it is something that it existed at some point upstream in the system, but as the data moved through its life cycle and it got down to where it needed to be for say financial reporting, those important tax elements were lost.
(04:04)
And to be able to adapt and be able to adjust for the speed of tax, I think that's one of the other changes that we're seeing. So volumes of data is increasing, but also the speed of reporting is becoming much more critical. I mentioned some of the digital tax administrations we're seeing tax administrations that are requiring digital data, transactional level data to be submitted on a monthly basis in some countries like Spain for example, is one that that's actually looks at daily transmissions of data. And to be able to effectively manage that and be able make sure that you're reporting the right information, you have to have a handle upfront on what's in your data. And this is also true when we're looking at tax being a strategic part of business operations, we need to be able to have access to the data much more timely than we generally do today.
(05:00)
In your previous podcast with Jim Thompson, I love the title that he had that he was named as a strategic business partner to the company he was working in. And I often talk about this as we look at this shift that we're seeing in the industry that tax for a long time I like to say has really been archeologists we're digging through transactions that occurred sometime in the past to try to determine the appropriate tax treatment and we are being pushed where we need to shift to be strategists. And in order to do that we have to reduce the amount of time that is taking us to gather and collect and make sense of the data. And again, that's being pushed from both that the internal agenda and how we be more strategic to business operations as well as externally with the tax administra administrations. And the only way to be able to do this effectively is to able to have a plan have a strategy about how all the different pieces of information across the life cycle for tax understand how that data comes together from the systems that we may be pulling it from, whether it's the enterprise resource planning systems for finance, whether it's coming from treasury or business operations, is really having a plan in place that allows us to be able to collect and report and analyze the information on more of a near-time, real-time basis.
Michael Cohn (06:28):
How effective do you think the big data is when you're dealing with tax compliance and planning out a tax strategy for a client or a company?
Daren Campbell (06:39):
But well, being able to understand the underlying data is really critical to any of planning understanding that underlying data. And so being able to use to have skills to be able to handle big data and to have some of the analytic tools to be able to help make sense of the data is increasingly important. We see, as you mentioned on the compliance side, we have seen a beginning of a shift of companies that are looking towards big data. They're looking to be able to bring data to allow them to do year over year analysis that be able to use analytics to be able to do drill downs, to be able to better understand the details of the data, what's driving the transactions. As I had mentioned before where tax has really been archeologists looking at the past, most of the analytics that we've used in tax understand data to date have really been descriptive analytics. It's kind a describing what happened and in order to shift to be more strategic, we need to move to where we're more predictive or prescriptive in the use of those analytics. And big data is really critical in being able to do that to understand really the drivers of business, the drivers and the levers that can be used when you're beginning to do planning for tax purposes and as well as making sure that you're doing compliance in the right way. Oh
Michael Cohn (08:10):
Let's talk about some of the specific technologies like robotic process automation. Is that useful in tax preparation?
Daren Campbell (08:17):
Yeah, absolutely. And to start with, robotic process automation definitely has its place in tax in this move to being more strategic in what we're doing as tax, as well as dealing with some of the challenges around big data, going back about six years is really where I think we saw the peak. Sometimes we refer to it as the boom in RPA in tax and there was a lot of interest ear early on because of this challenge of having data in so many different systems. So what robotic process automation does very, very well is it takes care of what we to as the swivel chair problem where you have data in different systems that you need to be able to access to be able to bring that together and consolidate it in a way that you could then use so you have all the information and triggers that you need for tax purposes.
(09:11)
We've also still seen a trend in technology where companies are beginning to open up systems and create a API layers application protocol interface layers that allow systems to communicate directly with one another A and that that's kind of solving some of the same problems that R P A was initially solving. And so as we look at the use of rpa, there are definitely some still some great use cases for it but it also can be very fragile because it's working on the front end, the interface side of applications. So it accesses applications just as a user would. And that's one of the benefits of it is you don't have to spend a lot of time working with IT to get backend and access to different systems. It's able to work in those systems as any end user is so it logs in through a user interface, it can navigate around that interface and it construct in information out of it.
(10:09)
And so that strength though can also be a challenge with it because if there are changes to the interface that can cause the RPA tool to be a little bit brittle or that it needs to be updated or changed to be able to account for some of those changes versus what we're seeing with APIs where APIs because they do direct connect at the data layer, generally the data layer of technology and systems do not change. So those tend to be a little bit more robust. Now, just to give you an example, one example of where we see RPA being used in one of the biggest use cases across tax is around extracting invoices out of client systems that typically those invoices are PDF P D F versions of or images of the documents that have been posted and logged into the system. In many of those systems, there's not another mechanism to easily pull that have those discreet in invoices down in batch.
(11:11)
And that's one of the biggest use cases that we see towards that. I had a company we worked with a while back that part of their exemption certificate process actually still used an old I B M mainframe system, so green screen system. So there weren't APIs or other ways to connect in with the data on the backend and we were able to use RPA to go in and read information out of that older system, bring that together with some other data and do some additional processing. So you RPA and kind of similar tools are something that we're seeing increasingly useful in helping deal with the volume of data that we're experiencing. A number of automation tools. R RRP is one class of those. Data integration is kind of another suite of tools. So tools like Alteryx that often we see used in combination with the R P A tools that are very good at the transformation of data as well as we've seen a big emergence in those type of tools within the Microsoft platform.
(12:20)
And so with many of our clients, I think many of those in the industry aren't even aware of some of these changes. There's been a lot of training and focus that we've done with companies around some of the tools like Query that that's actually part of built into Microsoft Excel that also built into Power BI that can be used to help transform data in a automated fashion. And these tools have gained a lot of popularity in tax because they are end user tools for the most part. So they're tools that the tax professionals can use to transform and automate the data. And I think that's one of the key elements because one of the challenges that tax faces and especially in this as we move into this world of big data is often when it comes to C'S priorities for their IT department, those priorities are driven by business operations and they're driven by finance because there's bigger initiatives that come in from those sides of the business where taxes is often a river of nickels.
(13:32)
We have a bunch of smaller things that but huge needs and huge impacts on the outputs but they're kind of smaller needs when they're building a business case that might be taken to it. And so we have seen a rise in popularity around some of these citizen developer type tools and I think that kind of goes back to our roots of digital within accounting that I like to say that the number one accounting software, tax software today is still Microsoft Excel. So I think we can trace the roots back to 1979 with the very first spreadsheet application. But if you look at why spreadsheets have been so sticky in our industry, it's largely because it allows the end user to make the changes and update the models without having to go to or rely on it or a vendor to make those updates. And so we see things like provision software as an example that although there are commercially available provision tools that are out there, there's still a lot of companies that still use spreadsheets to do their provision because it gives them that kind of flexibility and control.
(14:46)
And then this gets back to your roadmap question as we start looking at the volumes of data that are coming and some of the challenges with some of the new regulations. So looking at global minimum tax under the base erosion profit shifting 2.0 that there are hundreds of new data points that tax needs to collect and gather and those data points aren't completely finalized as the regulation still gets worked through. But again, another example of where taxes really relying on technologies that A allow them to be able to have access to a lot of different data but also needs to be very nimble in the way that it's transformed and ultimately used for modeling, modeling, reporting purposes.
Michael Cohn (15:32):
Thanks. We're going to take a short break. We'll be right back with more from Darren Campbell of ey. Hi. Well we're back with Darren Campbell of EY America's tax Technology and Transformation Leader. Welcome Darren. We're talking about some of the different types of technology that tech professionals are using and we were talking about big data and also robotic process automation. There's been a lot of talk lately about artificial intelligence, these AI chat bots like chat G B T, you've been getting a lot of publicity. Do you see role for AI in tax technology there at Y?
Daren Campbell (16:21):
Absolutely a at EY and everywhere else across our industry I think it's very, very ripe for that. As we talk about ai, I like to say it's a 75 year old technology, it's just it's day has finally come. A lot of the low hanging fruit for the use of AI in tax comes from an algorithm SP that have been around for a long time. If you think about a lot of things that we do in tax, we are trying to group items into some type of bucket and then that bucket has a specific tax treatment. So again, it could be something on the sales and use side of the house where we're looking at whether a certain item is exempt or not exempt or sales tax purposes. It could be looking at fixed assets to determine which asset class particular item goes into. There's a lot of things that we're doing where we're looking at bucketing, how do we get things kind of bucketed into the right way and then once it's in the bucket then some of the rules are then a little bit more automatic.
(17:32)
That is one area that I see is really low hanging fruit is something that we are using at EY AI in a number of places to help with the classification problems and helping our clients solve those challenges as well. And this gets back into maybe the big data discussion as well. We had one client a while back that they had over 48 million different types of products that, and they were looking to try to classify those into kind of particular buckets again for sales tax purposes. And we were able to code those 48 million items into the appropriate buckets within about a 10 day period. So something that if trying to go after that manually we never would've been able to do and we're close to that timeframe and we were able to do very quickly using ai. And as you talk about chat G P T and I, I'm really excited how this is bringing AI into some of the mainstream conversations that I do see that there's a lot of hype around it which is exciting and it's creating kind of a wave of interest that I think is going to drive a lot of really good outcomes.
(18:47)
Chat G P T also shows just the pace of change and how that's evolving that looking at some of the products around natural language generation just a year ago what was commercially available, there were a lot of good products that were able to take certain information and be able to put together a summary and we saw that going back outside of the tax area to looking at some of the sporting events and AI's long been used to write up some of the recaps of the sporting games, taking the box scores and putting that into dialogue and we, we've seen application of that and the finance and tax world taking financial statements, another information, we're creating a summary of where there's variants year over year, but being able to really do some freeform writing haven't seen a lot of technologies that have done that very well.
(19:45)
And we do see that with the potential of that with Chad G P T and similar technologies that now really have feel like they've kind of come onto the scene overnight. When we look at AI use in business versus our personal lives, we still have a ways to go. And I think one of the reasons for that is in our personal lives AI was really just infused into the products that we were already using, helping I improve our experience. So as we were streaming different shows, we were getting recommendations about what to watch next. When we're shopping online, we get recommendations about other things that we might want to buy if we're driving somewhere. AI is now infused into the navigation tools that we used and so we didn't have to seek out a separate AI app. And in business we still see that a lot of times the a AI is still a little bit separate.
(20:46)
It's not fully integrated into the tools that our professionals are using. And I see that as one of the big opportunities for our profession right now is truly to infuse those infuse AI into those products so that our people they don't know need to know all the detail and math behind how all those tools work, but that they do get embedded directly into the tools that are being used. And one of the challenges that I see is with the embedding, there still needs to be transparency. So we still need to tax purposes the accounting purposes, we need to have explainability and that's also one of the challenges that we're addressing at EY is how do we build explainability into the AI driven applications to try to give some insight into how the AI is driving the decisions that it's making and really infusing that as part of the process because I do view AI very much as a human in the loop sometimes when we talk about robotic process automation and then we talk about ai, there's a concern that's raised about is does this mean that this is going to take my job?
(22:07)
Is this going to displace workers? And what we've seen through R P A over the last eight, nine years now is what it has done is it's actually freed up tax professionals to spend more time on the things that they went to school for analyzing the law and be able to do modeling and analysis to really make sure that the best tax positions were taken as opposed to spending 40 to 60% of their time just churning through data. And AI I view will do the same thing. I view this really as an augmentation of the human that is helping enhance our capabilities and if we look at some of the challenges that we have around the number of resources that we do have coming into the profession coupled with the increase, the rapid increase in the demands and reporting or and regulatory requirements that are coming in we need to be able to do things in better ways.
(23:13)
I've been with EY this year will be 26 years and when I first started I had heard the phrase about doing more with less and I think I hear that every year. I've heard that probably every year of my career that we need to do more with less or even just do more with the same is I think a lot of where we are right now. A lot of companies have gone very lean in their tax departments and tax functions. Again, the complexity of the rules that we're dealing with and the volume of data we're dealing with is just rapidly accelerating. So we need to have tools that can help us do our jobs more efficiently and be able to spend more time on the analysis and the application, the kind of strategic thinking than just the data preparation and compliance. That's been a big part of the industry historically.
Michael Cohn (24:06):
Well, we were talking a little bit about the O E C D rules and the digital service taxes. How do you think tech tax technology is going to be able to adapt in the future to some of these upcoming rules? There's also the ones from the Biden administration on the minimum taxes on corporations that make at least a billion dollars. Is technology going to help with compliance by companies with these new rules?
Daren Campbell (24:32):
Technology I think is going to be a must as we look at it, at the requirements as they're coming out and then again continues to evolve, but looks like they'll be around 150 to 200 new data points that are required. And as I mentioned, this spans all different parts of the organization, so it's reliant heavily on financial data and other tax data and HR data and corporate data. It's coming in from a lot of different areas. And then there's also all the local variations in elections. So trying to manage this manually or even through the use of spreadsheets, it's going to be extremely challenging for companies. And so in the discussions that we've been having with companies, a lot of the multinational companies here in the US to date, one of the streams that we have been talking to them about is around their data strategy about how are they going to get their arms around all the different data points, how they're going, are they going to do that in a timely way where again, they have visibility and view in into that data as well as being able to just pull the data together in order to do complete and timely filings.
Michael Cohn (25:52):
And there were a lot of changes coming up too. I know EY is considering a split as well between its auditing, assaulting sides and there's also these new rules were coming out. Do you think tax technologies can be able to evolve in the future and to embrace all the kinds of changes we see coming up out there?
Daren Campbell (26:17):
Absolutely. I think technology actually I think is the evolution of tax. If we look at what the future of tax looks like several years back, we had had a meeting with the head of tax in Russia that, and his background is actually as a data scientist and he had made the comment to our team that he didn't view tax as legal issue. He viewed it as a data problem to solve. And again, the tax regime they have in Russia is different, more simplified than what we have in the us but we are seeing globally, we are seeing tax being handled more as a data problem. We saw this down in Mexico as well as they were getting ready to digitize their tax administration, they hired a team of data scientists and in a lot of the way that the taxes is moving globally, it is adopting more of a data driven a approach.
(27:22)
And I think there still will be the legal interpretation, legal elements of it, tax as is used to drive a lot of different social agendas. It it's used in times of crisis is one of the first levers that governments are able to pull to provide relief. And so because of that they'll still, and they continue to be a lot of changes and need to be nimble, but increasingly the volume of data that's going to be required for tax compliance is going to increase and the role of the tax professional is going to move from a compliance focus to more of a planning and controversy focus. Again, this is something we've seen outside the US as companies have moved to a digital format that the governments themselves become the ones that are doing a lot of the compliance and the role of the tax function is to help make sure that the information being submitted is both complete and accurate, but also that if there's any planning that needs to be done or other things that need to be done, that is done at more of a near-time, real-time basis.
(28:38)
And we're also is seeing an increase in controversy work because there is a bit of a shift in the way that it works today as companies are filing tax returns or one traditional format company prepares a tax returns, sends it to the government and they be liability responsibility is on the government to push back on anything that they think wasn't filed appropriately. With a lot of the governments now taking on that compliance side of the activity, the controversy side is being shifted back to the taxpayer but the window to be able to respond to any notifications of adjustments, those are shrinking and is requiring that company's are become much more aware of what's in the details from of that transactional level detail that they're submitting to government. So I think that in increasingly one thing I did you want to clarify? I think sometimes I get the commenter question.
(29:42)
We see this in some of the surveys around again in our 2022 survey, 95% of the respondents indicated that the tax technical competencies of their tax professionals were being augmented with process data and technology skills. And I don't think that that means that our tax people have to become computer programmers. What I do think that that means is that our tax professionals in the future need to be more analytically, have an analytical mindset and be able to have an understanding of the art of the possible with the various technologies that are being used because that being able to utilize those technologies is going to be really critical in being able to meet those two in the internal external objectives of being a strategic business partner within an organization and meeting the external tax obligations coming from the various tax authorities.
Michael Cohn (30:48):
Oh, that's great. Well, I wanted to thank Daren Campbell, EY America's Tax Technology and Transformation Leader for joining us on today's podcast. This episode of On the Air was produced by Accounting Today with audio production by Kevin Parise. Please rate or review us on your favorite podcast platform and see the rest of our content on accountingtoday.com. Thanks again to our guest, Daren Campbell of EY, and thank you for listening.