Alter Everything Podcast

AlteryxMatt · ‎07-29-2025

Unlock the power of Alteryx for tax professionals in this insightful episode of Alter Everything! Join us in an interview with Adrian Steller, Director of Tax Technology at Ryan, to explore how Alteryx revolutionizes tax processes, automates data workflows, and enhances efficiency for tax teams. Discover real-world Alteryx use cases in VAT compliance, transfer pricing, and automation, and learn practical tips for transitioning from Excel to Alteryx. Whether you’re a tax analyst, data professional, or business leader, this episode provides actionable insights on leveraging Alteryx for tax data transformation, reporting, and analytics.

Panelists

Adrian Steller, Director @ International Tax Technology - @adriansteller, LinkedIn
Megan Bowers, Sr. Content Manager @ Alteryx - @MeganBowers, LinkedIn

Topics

Ep 190 (YT thumbnail).png

Transcript

Episode Transcription

Ep 190 final mixdown
===

[00:00:00] Megan Bowers: Welcome to Alter Everything, a podcast about data science and analytics culture.

I'm Megan Bowers, and today I am talking with Adrian Stellar, director of Tax Technology at Ryan. In this episode, we chat about the value and efficiency Alteryx delivers for tax processes, some interesting use cases Adrian's team has worked on, and tips for transitioning Excel work to Alteryx. Let's get started.

Hey, Adrian, it's great to have you on today. Could you give a quick introduction for our listeners?

[00:00:38] Adrian Steller: Sure. I'm Adrian Stellar. I'm a director in our tax technology practice based in London. I'm working with a firm called Ryan. Prior to joining Ryan, I held in-house roles in VAT Advisory and returns preparation. Prior to that, I was indirect tax advisory in the UK, and prior to that, back in Australia, all that, that was 20 odd years ago now. So, um, converted from pure tax to tax technology and now definitely more on the technology side of tax technology.

[00:01:14] Megan Bowers: Awesome. I'm excited to chat about all things tax technology with you today. I think a good place to start is just that many tax teams are still living in the Excel world, so could you walk us through a typical Excel-based tax process and how you could see it evolving into something more efficient with Alteryx?

[00:01:37] Adrian Steller: Sure. So prior to us connecting, I did check in with a colleague of mine who is very much in the VAT advisory side to see if what I did nearly 20 years ago when I was in-house is still how things go today. Very little seems to have changed. It all starts with getting data. Usually you're extracting data out of one or more source systems, your ERP, your transaction reports. You might also be getting master data to join in on it to, you know, enrich that data so that you can use it for reporting or analysis. You are then looking to do a lot of reformatting. Some of the reports you get out of ERP systems were designed for back in the days of dot matrix printers, if anybody knows what they are anymore, rather than clean column and row. Uh, that's improved over the last few years, but there are still a lot of businesses using old systems or legacy systems. So yeah, a lot of the process starts with getting the data.

Because you need to use a lot of transactional information in VAT, and I assume it's similar for sort of sales and use tax, which is also one of the indirect taxes people have to cope with. Once you've got your data and you've done your reformatting, you're then often having to do some reconciliations. Sometimes you are looking at reports rather than extracts from the general ledger, so you wanna make sure that the reports are complete. It might be that you're also checking some of those postings against documentation, so to make sure that, for example, you might need to check that your transaction information is, is backed by documents. The process from then varies considerably from organizations. Some will do checks and controls and analysis to to look at how the tax has been booked on the transactions, especially in accounts payable, you have a lot of people in the business who are not tax professionals making tax decisions. They're looking at an invoice and they see tax and they think, okay, tax is tax, but it might be taxed from another country. It might be tax that's not deductible. So, you know, depending on the business and also depending on their appetite for risk, the level of analysis varies. Some will just say, I'm gonna take the data from my system and I'm gonna put it on my return after I've cleaned it up. Others will do a lot of checking on the data.

And some will sit somewhere in between. And the checks that we've seen in the past will vary from a simple sort of rate check, is the rate 20% to some complex VLOOKUPs and things like that. And I've seen formulas that look more like paragraphs. It, it can be pretty intense analysis. And again, it depends on, do they have tools in their environment that help with the data accuracy? So maybe they've got a tax engine, in which case they might be able to place more trust in the data, or are they just willing to roll the dice? We don't have time to check it. It's immaterial, those sorts of things. But yeah, there will be some analysis that goes on at some level to try and check the data before it goes onto the return. Obviously if you find things that are wrong, then you need to make some adjustments, and if you've gotta make adjustments, then you've gotta post it back in the source system so that your GL lines up with what you're reporting to the tax authorities. You'll also have businesses out there that need to make industry specific adjustments. You talk to a VAT professional about something called partial exemption, you'll hear them use things like floor space and headcount. This is not normal transactional information. They have to pull data in from external sources and make calculations often in a separate spreadsheet that then feeds back into their VAT return. So it, it's all about getting the data out, reconciling it, checking it, adjusting it. And then finally, you are looking to say which number goes in which box. On the return, you gotta report your sales in one box. You gotta report your purchases in another box, the VAT on the sales, the VAT on the purchases. Some of those VAT returns can run to many, many boxes depending on the country. Um, you may also have some consolidation that goes on, so combining data from a number of different source systems. It's often a process that's run on a monthly basis, although some businesses are running it more often, but either way, the return is typically monthly or quarterly. So you've got some sort of data aggregation going on before you can file a return. You've got things called VAT groups. So that's where you treat a number of different entities as a single entity. So you're combining data from multiple entities. So I mean, that, that's the process in a nutshell. Um, right.

[00:06:51] Megan Bowers: Yeah. Sounds pretty complex. I mean, for a lot of data, ETL, but it seems like the transform part could be pretty intensive depending on where you are in the world or what, what kind of taxes you're dealing with.

[00:07:07] Adrian Steller: Yeah, exactly.

Exactly.

[00:07:08] Megan Bowers: And for me personally, I'm not sure I'd wanna be doing that all in Excel spreadsheets, so,

[00:07:15] Adrian Steller: Well, yeah, the process as well. It, it tends to be, you've either got a template that your predecessors predecessor built, or you're taking what you did last month, deleting the data and putting the new month's data in, or the new periods data in. So sort of a, a rinse and repeat. Back when I had to do my VAT compliance, we did move out of Excel because our reports were blowing the 65,000 line limit and we were using access, which was cutting edge back then. There will be some more complex processes, but they tend to follow this Excel model in our experience. Um, you might see people using things like linked worksheets or power query, but the basic process hasn't really changed. This is your typical Excel based process. Get the data, reconcile it, analyze it, adjust it, file.

[00:08:09] Megan Bowers: So you're a big Alteryx advocate for these kind of tax processes. What makes you interested in using Alteryx for it? Where have you seen efficiency gains?

[00:08:21] Adrian Steller: Yeah, so I guess it was one of those things. I looked at the process and when I'm talking to clients, a lot of the time I'm asking the question, well, do you guys have Alteryx in your landscape? Now, a few years ago, the answer was no. More recently, it seems to be, yeah, I'm aware that the firm's got it, but I dunno how to use it, or I don't have access to it, or I've never seen it before. And it, it's that mental leap I suppose, of how to learn something that is not Excel, even though it, it can do everything you can do in Excel and more. And I, I sort of look at it as well, that ETL process, that is what Alteryx has in spades. You know, you don't even really need to touch the source data to reformat it because your workflow. It's importing a version of it, and then you can have your transformations all set up to do what you need to do. You can feed in multiple sources of data. You could link to other sources of data, so you know you can build an API to pull your foreign exchange rates. For example, if you need to use an external source of foreign exchange rates, you don't need to log into the website, download a file, copy, paste, reformat, you just pull it straight in. Yeah. And.

Reconciliation. You've got a lot of, a lot of nice tools that you can use in Alteryx, and there's still very basic tools like a, a simple join or a sum tool or something like that, depending on what level you need to aggregate at. Writing that as an IF equation can be pretty hairy. Whereas Alteryx, you literally just stick a tool on the canvas, connect it to the two data sources you want to compare. Set up what you want to join it on. And then you can see where have I got commonality? Where do I have differences in source one and source two? And that, that's before you even get to the fun stuff from an analysis point of view. You know, the to do some of those tests that you do in Excel today. So yeah, a basic rate check is the rate 20%.

Yeah. You know, that's not too hard. Um, can I do things that are more advanced? Like I wanna look at the invoice reference text to see if the word tax appears, but not the word taxi. Now I can do that pretty quickly in Alteryx. Can I do it in Excel? I, I probably could, but it would be a lot more effort and.

It probably wouldn't be as flexible. You're talking lots of different sort of lookup formulas, and you know, can I easily combine those different tests together? Once you've written a formula that's a paragraph long, it's very hard to change it.

[00:11:04] Megan Bowers: And it's hard too, if you're passing that to the next person, or if you go on vacation and, and then they're stuck with a paragraph versus if you're using Alteryx, you've got that flow, the process flow, you can click into each tool, and just is a lot more traceable and understandable.

[00:11:23] Adrian Steller: Yeah. Also, if, if you got hit by bus and somebody have to pick up your spreadsheet, not that we want anyone to be hit by a bus, but it's the analogy we always use. If you need to troubleshoot what's going on in the Excel sheet, it's very difficult. Whereas the way that Alteryx lays things out in a flow, you're able to see what's the data like going into this particular tool. And I can see, well, if I've got three anchors where data is leaving the tool, so say that simple join, I can see what the effect is at each stage, and if people document their workflows properly, then you can even see what is the intention behind that tool. Now documentation is the make or break of any process, but what I typically found was when I was preparing my documentation of my Excel process, you know, if I roll forward like three or four months, I've already changed the process quite considerably. So my documentation's out of date. Because it's in a separate place in a an Alteryx workflow. If I decide I'm gonna insert six tools between tools one and two to do some slightly different variation of my analysis, as long as I put the comments or the annotation on the tool, or I can drag a comment onto the canvas, the documentation is encapsulated within the workflow as well.

[00:12:47] Megan Bowers: That's super nice. We'll definitely link some resources as show notes too. We got a lot of good blogs about documentation, and I know like a lot of our users are pretty, pretty passionate about it, which makes sense. If you're handed a workflow, I would be passionate about it, having documentation that comes with it, and especially for these processes.

[00:13:07] Adrian Steller: Even if the workflow's not documented, if you hit the run button, and maybe you wanna switch off the output tools first, but if you hit the run button, you can see at each step what's happening. You can open the tool up and you can say, oh, that's the IF formula. If I click on the left half of the tool, I can see what the data was going in. If I click on the right half of the tool, I can see what the data was going out. So even if I did put some insanely complex IF formula inside my formula tool, for example, I could unpick it. Whereas that's a lot harder in Excel.

[00:13:42] Megan Bowers: A hundred percent. Yeah. So then I wanna hear a little bit about the value you've seen from Alteryx around data ETL. Do you have any examples of the kinds of systems you can pull from now where it was challenging to do so before?

[00:13:57] Adrian Steller: A lot of the data that we're dealing with tends to come in flat file. Some of it has some really hokey formatting on it, which yeah, you've got the challenges in in Excel to deal with that, but in Alteryx you can deal with it fairly quickly. Hmm. Sometimes if you're having to deal with data that is more like a, a three line journal, if I can call it that, versus a tax report. So I, you know, from a tax perspective or a VAT perspective in particular, I wanna see the sales amount and I wanna see the tax charged. Now if that's presented on two separate lines. That can be harder to work with. To deal with that in Alteryx, I can effectively pivot that so I get the net amount and the VAT amount on the same line attached to the same document number. It just makes those transformations much easier to work with. In terms of some of the things that we've used it for with clients, so we had one client that was implementing a piece of VAT compliance software. So this software is designed to take the transaction data in and effectively map it to the VAT return boxes, and prepare your working papers. Now, they had six different source systems. They had, I think it was one of their source systems, did not have tax codes, which are vital for preparing VAT returns because you need to know if something is a good or a service, for example, whether it's deductible or not deductible. So for this particular client, we managed to introduce some logic that would actually build out custom tax codes on the fly that we could then import into the VAT compliance software. And we were also running all six data streams in to produce a, a data that would be loaded into the VAT compliance software in a unified format because the other way to do it would be to build the, the import specification in the software, but we would have to build six of them. So this way we could handle all of the logical operations and produce the unified format. That was incredibly valuable. It was also a process that it was repeatable. Because we weren't working on the data as such. We were processing the data and that that's one of the key benefits there.

[00:16:20] Megan Bowers: And then have there been any other ones around like automation reporting, scheduling, email, things like that?

[00:16:28] Adrian Steller: So probably our most complex one to date has actually been in transfer pricing. Whereby we needed to query their sort of reporting data warehouse. So our workflow, we were building a SQL query on the fly within the workflow to go and query the SQL database. The query was pulling a subset of data, so we were able to use Alteryx's in-database tools to, rather than pull out hundreds and thousands of lines and then try and consolidate it, we were doing all of the processing in the SQL server and pulling out the summary that we needed rather than all of the detail. Part of that process, we then needed to reformat the data, which is common to a, a VAT process, but it was feeding some transfer pricing software. So it had to be presented in a particular format, but it also had to be output as zip files, and we needed to combine multiple files within the same zip file. Then we would have to drop that onto an SFTP server. Oh my gosh. Or another process to go and pick it up and deal with it. But we did all of that sort of ETL process inside Alteryx. It was a workflow that was designed to go on the client's Alteryx gallery. And it was scheduled. So we had to have the process run so that the China data was ready for the start of their day, that the European data was ready for the start of the European Day. The, the rest of Asia Pacific data was ready for the start of their day. So the parameters for every run of the workflow changed. They were running this process multiple times a month. We also introduced some logging so that if there was an issue with the process, that an email was sent out to the team that was monitoring it to say, oh, we had a, something went wrong with the process. Either there was something in the data we didn't expect, or the schedule didn't run as expected, you know, whatever it was. But at least people knew because they got an email. The email included the log file, so they could go through and say, oh, okay. There was an issue with one of the, one of the data sources was offline so we couldn't run the script, or things like that. That's probably been the most complex one that we've done to date on an ETL point of view. We've done some interesting things as well with a lot of sort of string functions, text mining. So we had to analyze a product list to identify products that are subject to what they call domestic reverse charge. Ba basically identify products that are subject to a different tax treatment. And it was things like phones and laptops and, you know, computer chips and stuff like that. Their product list would be like you would say, okay, iPhone, great iPhone. That's one that we need to use. But if it's an iPhone case, it's like, no, we don't want the iPhone case to have that treatment, just the iPhone. So you would have to, in the text, mining, go through and say, well, this is my list of words that I need to include, but then this is my list of things that, if that appears with that word, I'm not interested. That was a pretty complex one that, that, you know, this was going through thousands of products because you got a blue iPhone case, a red iPhone case. Yeah. Different manufacturers the whole bit. It was thousands of lines to do that manually, even if you're wanting to search it in Excel. Very painful. Right. So yeah, that was, that was an interesting project. We really enjoyed that one.

[00:20:17] Megan Bowers: Yeah, that sounds super interesting. And all of those that you shared, really interesting use cases, getting into some really complex Alteryx workflows, I'm sure. But yeah, I'm just thinking about, for anybody listening who maybe feels a little bit intimidated by Alteryx, by learning it, or maybe it doesn't have the time. I know that when we talked before, you had mentioned that there were some new grads that joined your team and were able to create value with Alteryx quickly and ramp up. So can you just share a little bit about that?

[00:20:49] Adrian Steller: Um, we had a new starter that's never touched Alteryx before. I mean, he, he does have a computing background, has a little bit of an IT bent to start with, which may be an advantage, maybe not, I'm not sure, but with a little bit of training and just practice, really practice and coaching. We were able to get him to the point where we could hand him a task and say, okay, this is the input that we have. These are the operations that we need to do, and this is the desired output. And he is been quickly able to build out this workflow within a couple of days of just tinkering around and trial and error. Even the really complex workflows that we've been dealing with, a lot of them are using maybe 10 or 15 at the most different types of Alteryx tools. So you can start very basic. Achieve quite a lot with a limited tool set before you even get into the really complex stuff. Importing data, drag a tool onto the canvas.

Okay, job done. Um,

[00:21:51] Megan Bowers: mm-hmm.

[00:21:51] Adrian Steller: But things like filters, formulas, you know, in, in an Excel landscape, you're thinking pivot tables. That's just the sum tool. Your, uh, your transposes, your cross tabs, tho those sorts of tools, they, they're pretty basic tools at the end of the day, but they can be very powerful and you can build very complex workflows, even with a limited tool set. And that's before you even get into things like your machine learning and all of the other stuff that is at your fingertips.

[00:22:21] Megan Bowers: Definitely. And then you mentioned before that you have some clients who, they're aware they have Alteryx at their organization, but maybe they haven't used it yet. They don't know where to start. What advice would you give for starting to transition some of these Excel-based processes into Alteryx?

I guess

[00:22:40] Adrian Steller: it's similar to how we train our people that are using Alteryx now. It's a case. Find something. Doesn't necessarily have to be your process. You wanna learn the basics before you, you jump into the deep end. So you'll usually, in your day job, come across something where it's like, oh, I need this. I need to do a quick calculation. Fire up Excel is a knee-jerk reaction. If you've got the time, rather than firing up Excel, fire up Alteryx, bring the data in, and then try and think about how do I do what I wanted to do in Excel in Alteryx? Because you'll probably find that it'll be something like, oh, I need to do a pivot table, or I need to filter my data, or I need to transform my data. And there'll be tools that you can use trial and error wise to achieve those sorts of functions and don't try and boil the ocean. Start with a small job and learn a couple of tools there. There are lots of online resources with how to use those tools as well. The Alteryx community as well. I mean, if you've got somebody that has experience in Alteryx, you can lean on. That's a big help too. My colleagues and I. When we were learning Alteryx, we'd be working on independent things, and I'd jump on teams and send somebody a chat. It's like, I'm trying to do something. I really need to pick your brains. So we would screen share, go through it together, and they'd say, oh, have you tried this or tried that? Sometimes you'll see that there are a number of different ways to do the same thing within Alteryx. It might be that the approach that you were trying was too complex. You know, you've used the wrong tool. Or it might be that the person you're asking for help goes, oh wow, I didn't think of that. And then you can troubleshoot the approach together. We have had clients where they've known about Alteryx and they said, look, can you train us. We've got a process, we've got something we want to do. So we would do a session whereby we sit down with them and say, okay, let's start with defining the process that you're trying to do. And then we'd set them little homework assignments. So this is how this tool works. Why don't you go off? Go off and play with that in your data. Then we'll check in tomorrow and see where you got. So, you know, that way they were able to lean on us because they didn't have anybody that, that they could lean on within their organization or anybody that was willing to help them. So, uh, um, start small. Use the experience of people around you that have used it before. I think. One of the big barriers that people have is that they've heard of it, but maybe they've gotta have a business case to to get it installed or something like that. And it's like they don't know what they don't know, so it's like, well, maybe just ask, play with it, figure out whether it's something that's for you or not. And you'll probably find that it is at the end of the day. Because if you've been doing all of that work in Excel, once you've got your feet under you, it's easy to transition that stuff into Alteryx.

[00:25:42] Megan Bowers: Definitely, and I like what you said about you can't know. What you don't know. And I think that like another piece of that is sometimes you don't know the use cases that you don't know. So hoping that this episode, all the interesting use cases that you mentioned before, can get people's brain thinking about what could be possible with Alteryx. But yeah, it's been super nice to have you on the show today. Thanks so much for coming on and sharing your experiences.

[00:26:09] Adrian Steller: Thanks very much for having me.

[00:26:12] Megan Bowers: Thanks for listening. To learn more about Ryan and their own Tax Lab podcast series, head over to our show notes on alteryx.com/podcast. And if you like this episode, leave us a review. See you next time.

Ep 190 final mixdown=== [00:00:00] Megan Bowers: Welcome to Alter Everything, a podcast about data science and analytics culture. I'm Megan Bowers, and today I am talking with Adrian Stellar, director of Tax Technology at Ryan. In this episode, we chat about the value and efficiency Alteryx delivers for tax processes, some interesting use cases Adrian's team has worked on, and tips for transitioning Excel work to Alteryx. Let's get started. Hey, Adrian, it's great to have you on today. Could you give a quick introduction for our listeners? [00:00:38] Adrian Steller: Sure. I'm Adrian Stellar. I'm a director in our tax technology practice based in London. I'm working with a firm called Ryan. Prior to joining Ryan, I held in-house roles in VAT Advisory and returns preparation. Prior to that, I was indirect tax advisory in the UK, and prior to that, back in Australia, all that, that was 20 odd years ago now. So, um, converted from pure tax to tax technology and now definitely more on the technology side of tax technology. [00:01:14] Megan Bowers: Awesome. I'm excited to chat about all things tax technology with you today. I think a good place to start is just that many tax teams are still living in the Excel world, so could you walk us through a typical Excel-based tax process and how you could see it evolving into something more efficient with Alteryx? [00:01:37] Adrian Steller: Sure. So prior to us connecting, I did check in with a colleague of mine who is very much in the VAT advisory side to see if what I did nearly 20 years ago when I was in-house is still how things go today. Very little seems to have changed. It all starts with getting data. Usually you're extracting data out of one or more source systems, your ERP, your transaction reports. You might also be getting master data to join in on it to, you know, enrich that data so that you can use it for reporting or analysis. You are then looking to do a lot of reformatting. Some of the reports you get out of ERP systems were designed for back in the days of dot matrix printers, if anybody knows what they are anymore, rather than clean column and row. Uh, that's improved over the last few years, but there are still a lot of businesses using old systems or legacy systems. So yeah, a lot of the process starts with getting the data. Because you need to use a lot of transactional information in VAT, and I assume it's similar for sort of sales and use tax, which is also one of the indirect taxes people have to cope with. Once you've got your data and you've done your reformatting, you're then often having to do some reconciliations. Sometimes you are looking at reports rather than extracts from the general ledger, so you wanna make sure that the reports are complete. It might be that you're also checking some of those postings against documentation, so to make sure that, for example, you might need to check that your transaction information is, is backed by documents. The process from then varies considerably from organizations. Some will do checks and controls and analysis to to look at how the tax has been booked on the transactions, especially in accounts payable, you have a lot of people in the business who are not tax professionals making tax decisions. They're looking at an invoice and they see tax and they think, okay, tax is tax, but it might be taxed from another country. It might be tax that's not deductible. So, you know, depending on the business and also depending on their appetite for risk, the level of analysis varies. Some will just say, I'm gonna take the data from my system and I'm gonna put it on my return after I've cleaned it up. Others will do a lot of checking on the data. And some will sit somewhere in between. And the checks that we've seen in the past will vary from a simple sort of rate check, is the rate 20% to some complex VLOOKUPs and things like that. And I've seen formulas that look more like paragraphs. It, it can be pretty intense analysis. And again, it depends on, do they have tools in their environment that help with the data accuracy? So maybe they've got a tax engine, in which case they might be able to place more trust in the data, or are they just willing to roll the dice? We don't have time to check it. It's immaterial, those sorts of things. But yeah, there will be some analysis that goes on at some level to try and check the data before it goes onto the return. Obviously if you find things that are wrong, then you need to make some adjustments, and if you've gotta make adjustments, then you've gotta post it back in the source system so that your GL lines up with what you're reporting to the tax authorities. You'll also have businesses out there that need to make industry specific adjustments. You talk to a VAT professional about something called partial exemption, you'll hear them use things like floor space and headcount. This is not normal transactional information. They have to pull data in from external sources and make calculations often in a separate spreadsheet that then feeds back into their VAT return. So it, it's all about getting the data out, reconciling it, checking it, adjusting it. And then finally, you are looking to say which number goes in which box. On the return, you gotta report your sales in one box. You gotta report your purchases in another box, the VAT on the sales, the VAT on the purchases. Some of those VAT returns can run to many, many boxes depending on the country. Um, you may also have some consolidation that goes on, so combining data from a number of different source systems. It's often a process that's run on a monthly basis, although some businesses are running it more often, but either way, the return is typically monthly or quarterly. So you've got some sort of data aggregation going on before you can file a return. You've got things called VAT groups. So that's where you treat a number of different entities as a single entity. So you're combining data from multiple entities. So I mean, that, that's the process in a nutshell. Um, right. [00:06:51] Megan Bowers: Yeah. Sounds pretty complex. I mean, for a lot of data, ETL, but it seems like the transform part could be pretty intensive depending on where you are in the world or what, what kind of taxes you're dealing with. [00:07:07] Adrian Steller: Yeah, exactly. Exactly. [00:07:08] Megan Bowers: And for me personally, I'm not sure I'd wanna be doing that all in Excel spreadsheets, so, [00:07:15] Adrian Steller: Well, yeah, the process as well. It, it tends to be, you've either got a template that your predecessors predecessor built, or you're taking what you did last month, deleting the data and putting the new month's data in, or the new periods data in. So sort of a, a rinse and repeat. Back when I had to do my VAT compliance, we did move out of Excel because our reports were blowing the 65,000 line limit and we were using access, which was cutting edge back then. There will be some more complex processes, but they tend to follow this Excel model in our experience. Um, you might see people using things like linked worksheets or power query, but the basic process hasn't really changed. This is your typical Excel based process. Get the data, reconcile it, analyze it, adjust it, file. [00:08:09] Megan Bowers: So you're a big Alteryx advocate for these kind of tax processes. What makes you interested in using Alteryx for it? Where have you seen efficiency gains? [00:08:21] Adrian Steller: Yeah, so I guess it was one of those things. I looked at the process and when I'm talking to clients, a lot of the time I'm asking the question, well, do you guys have Alteryx in your landscape? Now, a few years ago, the answer was no. More recently, it seems to be, yeah, I'm aware that the firm's got it, but I dunno how to use it, or I don't have access to it, or I've never seen it before. And it, it's that mental leap I suppose, of how to learn something that is not Excel, even though it, it can do everything you can do in Excel and more. And I, I sort of look at it as well, that ETL process, that is what Alteryx has in spades. You know, you don't even really need to touch the source data to reformat it because your workflow. It's importing a version of it, and then you can have your transformations all set up to do what you need to do. You can feed in multiple sources of data. You could link to other sources of data, so you know you can build an API to pull your foreign exchange rates. For example, if you need to use an external source of foreign exchange rates, you don't need to log into the website, download a file, copy, paste, reformat, you just pull it straight in. Yeah. And. Reconciliation. You've got a lot of, a lot of nice tools that you can use in Alteryx, and there's still very basic tools like a, a simple join or a sum tool or something like that, depending on what level you need to aggregate at. Writing that as an IF equation can be pretty hairy. Whereas Alteryx, you literally just stick a tool on the canvas, connect it to the two data sources you want to compare. Set up what you want to join it on. And then you can see where have I got commonality? Where do I have differences in source one and source two? And that, that's before you even get to the fun stuff from an analysis point of view. You know, the to do some of those tests that you do in Excel today. So yeah, a basic rate check is the rate 20%. Yeah. You know, that's not too hard. Um, can I do things that are more advanced? Like I wanna look at the invoice reference text to see if the word tax appears, but not the word taxi. Now I can do that pretty quickly in Alteryx. Can I do it in Excel? I, I probably could, but it would be a lot more effort and. It probably wouldn't be as flexible. You're talking lots of different sort of lookup formulas, and you know, can I easily combine those different tests together? Once you've written a formula that's a paragraph long, it's very hard to change it. [00:11:04] Megan Bowers: And it's hard too, if you're passing that to the next person, or if you go on vacation and, and then they're stuck with a paragraph versus if you're using Alteryx, you've got that flow, the process flow, you can click into each tool, and just is a lot more traceable and understandable. [00:11:23] Adrian Steller: Yeah. Also, if, if you got hit by bus and somebody have to pick up your spreadsheet, not that we want anyone to be hit by a bus, but it's the analogy we always use. If you need to troubleshoot what's going on in the Excel sheet, it's very difficult. Whereas the way that Alteryx lays things out in a flow, you're able to see what's the data like going into this particular tool. And I can see, well, if I've got three anchors where data is leaving the tool, so say that simple join, I can see what the effect is at each stage, and if people document their workflows properly, then you can even see what is the intention behind that tool. Now documentation is the make or break of any process, but what I typically found was when I was preparing my documentation of my Excel process, you know, if I roll forward like three or four months, I've already changed the process quite considerably. So my documentation's out of date. Because it's in a separate place in a an Alteryx workflow. If I decide I'm gonna insert six tools between tools one and two to do some slightly different variation of my analysis, as long as I put the comments or the annotation on the tool, or I can drag a comment onto the canvas, the documentation is encapsulated within the workflow as well. [00:12:47] Megan Bowers: That's super nice. We'll definitely link some resources as show notes too. We got a lot of good blogs about documentation, and I know like a lot of our users are pretty, pretty passionate about it, which makes sense. If you're handed a workflow, I would be passionate about it, having documentation that comes with it, and especially for these processes. [00:13:07] Adrian Steller: Even if the workflow's not documented, if you hit the run button, and maybe you wanna switch off the output tools first, but if you hit the run button, you can see at each step what's happening. You can open the tool up and you can say, oh, that's the IF formula. If I click on the left half of the tool, I can see what the data was going in. If I click on the right half of the tool, I can see what the data was going out. So even if I did put some insanely complex IF formula inside my formula tool, for example, I could unpick it. Whereas that's a lot harder in Excel. [00:13:42] Megan Bowers: A hundred percent. Yeah. So then I wanna hear a little bit about the value you've seen from Alteryx around data ETL. Do you have any examples of the kinds of systems you can pull from now where it was challenging to do so before? [00:13:57] Adrian Steller: A lot of the data that we're dealing with tends to come in flat file. Some of it has some really hokey formatting on it, which yeah, you've got the challenges in in Excel to deal with that, but in Alteryx you can deal with it fairly quickly. Hmm. Sometimes if you're having to deal with data that is more like a, a three line journal, if I can call it that, versus a tax report. So I, you know, from a tax perspective or a VAT perspective in particular, I wanna see the sales amount and I wanna see the tax charged. Now if that's presented on two separate lines. That can be harder to work with. To deal with that in Alteryx, I can effectively pivot that so I get the net amount and the VAT amount on the same line attached to the same document number. It just makes those transformations much easier to work with. In terms of some of the things that we've used it for with clients, so we had one client that was implementing a piece of VAT compliance software. So this software is designed to take the transaction data in and effectively map it to the VAT return boxes, and prepare your working papers. Now, they had six different source systems. They had, I think it was one of their source systems, did not have tax codes, which are vital for preparing VAT returns because you need to know if something is a good or a service, for example, whether it's deductible or not deductible. So for this particular client, we managed to introduce some logic that would actually build out custom tax codes on the fly that we could then import into the VAT compliance software. And we were also running all six data streams in to produce a, a data that would be loaded into the VAT compliance software in a unified format because the other way to do it would be to build the, the import specification in the software, but we would have to build six of them. So this way we could handle all of the logical operations and produce the unified format. That was incredibly valuable. It was also a process that it was repeatable. Because we weren't working on the data as such. We were processing the data and that that's one of the key benefits there. [00:16:20] Megan Bowers: And then have there been any other ones around like automation reporting, scheduling, email, things like that? [00:16:28] Adrian Steller: So probably our most complex one to date has actually been in transfer pricing. Whereby we needed to query their sort of reporting data warehouse. So our workflow, we were building a SQL query on the fly within the workflow to go and query the SQL database. The query was pulling a subset of data, so we were able to use Alteryx's in-database tools to, rather than pull out hundreds and thousands of lines and then try and consolidate it, we were doing all of the processing in the SQL server and pulling out the summary that we needed rather than all of the detail. Part of that process, we then needed to reformat the data, which is common to a, a VAT process, but it was feeding some transfer pricing software. So it had to be presented in a particular format, but it also had to be output as zip files, and we needed to combine multiple files within the same zip file. Then we would have to drop that onto an SFTP server. Oh my gosh. Or another process to go and pick it up and deal with it. But we did all of that sort of ETL process inside Alteryx. It was a workflow that was designed to go on the client's Alteryx gallery. And it was scheduled. So we had to have the process run so that the China data was ready for the start of their day, that the European data was ready for the start of the European Day. The, the rest of Asia Pacific data was ready for the start of their day. So the parameters for every run of the workflow changed. They were running this process multiple times a month. We also introduced some logging so that if there was an issue with the process, that an email was sent out to the team that was monitoring it to say, oh, we had a, something went wrong with the process. Either there was something in the data we didn't expect, or the schedule didn't run as expected, you know, whatever it was. But at least people knew because they got an email. The email included the log file, so they could go through and say, oh, okay. There was an issue with one of the, one of the data sources was offline so we couldn't run the script, or things like that. That's probably been the most complex one that we've done to date on an ETL point of view. We've done some interesting things as well with a lot of sort of string functions, text mining. So we had to analyze a product list to identify products that are subject to what they call domestic reverse charge. Ba basically identify products that are subject to a different tax treatment. And it was things like phones and laptops and, you know, computer chips and stuff like that. Their product list would be like you would say, okay, iPhone, great iPhone. That's one that we need to use. But if it's an iPhone case, it's like, no, we don't want the iPhone case to have that treatment, just the iPhone. So you would have to, in the text, mining, go through and say, well, this is my list of words that I need to include, but then this is my list of things that, if that appears with that word, I'm not interested. That was a pretty complex one that, that, you know, this was going through thousands of products because you got a blue iPhone case, a red iPhone case. Yeah. Different manufacturers the whole bit. It was thousands of lines to do that manually, even if you're wanting to search it in Excel. Very painful. Right. So yeah, that was, that was an interesting project. We really enjoyed that one. [00:20:17] Megan Bowers: Yeah, that sounds super interesting. And all of those that you shared, really interesting use cases, getting into some really complex Alteryx workflows, I'm sure. But yeah, I'm just thinking about, for anybody listening who maybe feels a little bit intimidated by Alteryx, by learning it, or maybe it doesn't have the time. I know that when we talked before, you had mentioned that there were some new grads that joined your team and were able to create value with Alteryx quickly and ramp up. So can you just share a little bit about that? [00:20:49] Adrian Steller: Um, we had a new starter that's never touched Alteryx before. I mean, he, he does have a computing background, has a little bit of an IT bent to start with, which may be an advantage, maybe not, I'm not sure, but with a little bit of training and just practice, really practice and coaching. We were able to get him to the point where we could hand him a task and say, okay, this is the input that we have. These are the operations that we need to do, and this is the desired output. And he is been quickly able to build out this workflow within a couple of days of just tinkering around and trial and error. Even the really complex workflows that we've been dealing with, a lot of them are using maybe 10 or 15 at the most different types of Alteryx tools. So you can start very basic. Achieve quite a lot with a limited tool set before you even get into the really complex stuff. Importing data, drag a tool onto the canvas. Okay, job done. Um, [00:21:51] Megan Bowers: mm-hmm. [00:21:51] Adrian Steller: But things like filters, formulas, you know, in, in an Excel landscape, you're thinking pivot tables. That's just the sum tool. Your, uh, your transposes, your cross tabs, tho those sorts of tools, they, they're pretty basic tools at the end of the day, but they can be very powerful and you can build very complex workflows, even with a limited tool set. And that's before you even get into things like your machine learning and all of the other stuff that is at your fingertips. [00:22:21] Megan Bowers: Definitely. And then you mentioned before that you have some clients who, they're aware they have Alteryx at their organization, but maybe they haven't used it yet. They don't know where to start. What advice would you give for starting to transition some of these Excel-based processes into Alteryx? I guess [00:22:40] Adrian Steller: it's similar to how we train our people that are using Alteryx now. It's a case. Find something. Doesn't necessarily have to be your process. You wanna learn the basics before you, you jump into the deep end. So you'll usually, in your day job, come across something where it's like, oh, I need this. I need to do a quick calculation. Fire up Excel is a knee-jerk reaction. If you've got the time, rather than firing up Excel, fire up Alteryx, bring the data in, and then try and think about how do I do what I wanted to do in Excel in Alteryx? Because you'll probably find that it'll be something like, oh, I need to do a pivot table, or I need to filter my data, or I need to transform my data. And there'll be tools that you can use trial and error wise to achieve those sorts of functions and don't try and boil the ocean. Start with a small job and learn a couple of tools there. There are lots of online resources with how to use those tools as well. The Alteryx community as well. I mean, if you've got somebody that has experience in Alteryx, you can lean on. That's a big help too. My colleagues and I. When we were learning Alteryx, we'd be working on independent things, and I'd jump on teams and send somebody a chat. It's like, I'm trying to do something. I really need to pick your brains. So we would screen share, go through it together, and they'd say, oh, have you tried this or tried that? Sometimes you'll see that there are a number of different ways to do the same thing within Alteryx. It might be that the approach that you were trying was too complex. You know, you've used the wrong tool. Or it might be that the person you're asking for help goes, oh wow, I didn't think of that. And then you can troubleshoot the approach together. We have had clients where they've known about Alteryx and they said, look, can you train us. We've got a process, we've got something we want to do. So we would do a session whereby we sit down with them and say, okay, let's start with defining the process that you're trying to do. And then we'd set them little homework assignments. So this is how this tool works. Why don't you go off? Go off and play with that in your data. Then we'll check in tomorrow and see where you got. So, you know, that way they were able to lean on us because they didn't have anybody that, that they could lean on within their organization or anybody that was willing to help them. So, uh, um, start small. Use the experience of people around you that have used it before. I think. One of the big barriers that people have is that they've heard of it, but maybe they've gotta have a business case to to get it installed or something like that. And it's like they don't know what they don't know, so it's like, well, maybe just ask, play with it, figure out whether it's something that's for you or not. And you'll probably find that it is at the end of the day. Because if you've been doing all of that work in Excel, once you've got your feet under you, it's easy to transition that stuff into Alteryx. [00:25:42] Megan Bowers: Definitely, and I like what you said about you can't know. What you don't know. And I think that like another piece of that is sometimes you don't know the use cases that you don't know. So hoping that this episode, all the interesting use cases that you mentioned before, can get people's brain thinking about what could be possible with Alteryx. But yeah, it's been super nice to have you on the show today. Thanks so much for coming on and sharing your experiences. [00:26:09] Adrian Steller: Thanks very much for having me. [00:26:12] Megan Bowers: Thanks for listening. To learn more about Ryan and their own Tax Lab podcast series, head over to our show notes on alteryx.com/podcast. And if you like this episode, leave us a review. See you next time.

This episode was produced by Megan Bowers (@MeganBowers), Mike Cusic (@mikecusic), and Matt Rotundo (@AlteryxMatt). Special thanks to @andyuttley for the theme music track, and @mikecusic for our album artwork.

Alter Everything Podcast

Podcast Guide

190: Alteryx Use Cases in the Tax Industry

Panelists

Topics

Transcript

This episode was produced by Megan Bowers (@MeganBowers), Mike Cusic (@mikecusic), and Matt Rotundo (@AlteryxMatt). Special thanks to @andyuttley for the theme music track, and @mikecusic for our album artwork.