Data analytics is transitioning from a “consideration” to an “expectation.” As your team explores data analytics tools, it is important to know how each tool is set up and what it can be used for. Watch data analytics experts explain common tools and their applications. Below is a transcript of the video.
This transcript was assembled by hand and may contain some errors.
It has been edited for readability.
Bradley Carroll (BC): Hello, and welcome back to Frazier and Deeter’s webcast series on Data Analytics. If you’ve watched our former Culture of Compliance video you’ve heard Tripp Stedham and I talk a little bit about getting started with Data Analytics and looking for the opportunities, today we’re going to delve in a little bit deeper not only talk about opportunities but look at an individual one.
I’ll be the moderator today. I’m Bradley Carroll a Principal with Frazier Deeter’s Process, Risk and Governance (PRG) team and Tripp Stedham is a director on the same team. You’re going to hear us talk a little bit about a very specific opportunity that we performed recently for a client and how that could work out for you as well.
So, you’ve heard a lot about data analytics, we all have and why?
The Institute of Internal Auditors (IIA) standards require that you consider data analytic techniques, but they don’t really say you have to use data analytics. But if we go to the IIA mandates that came out in the Global Knowledge Brief, you’ll see that there’s a part one, which indicates, there is a part two. So, it is strong enough for the IIA to issue two Global Knowledge Briefs on data analytics mandate so it’s a clear indication of what we should be doing and the direction we should be heading as internal auditors.
The question is why? Well, audit needs to stay relevant; we need to have a seat at the table, and we need to continue to do risk-based auditing and data analytics really help us do this in a world of big data. Companies have become data hogs because data storage is so cheap and in order to get through the data and be relevant in your audit comments, data analytic tools will help you to do that.
Tripp Stedham (TS): You mentioned tools Bradley and that’s part of what we’re going to talk about today. Most people think data analytics, “I don’t know how to do that,” “I don’t have the skillset,” I don’t have the tools at hand,” so one of the things we want to talk about today is just kind of demystify what some of these tools are, and trying to help you understand well what are some of the things I could use to do different types of work. But walking through just a few of the examples out there, the types of tools and generally break the tools down into several different categories.
I think of the most simplistic ones that you and I use every day, which we might call Guided User Interface, or point and click, or things like Excel and Word, you have all those types of normal Office features, so I don’t have to know how to code and there’s a whole slew of tools that go into that category.
Then you get a little more technical with things that require scripting and recording what your steps have been or making something more repeatable. Again, with analytics we love tools that can repeat the process, running a test once is great, but running a test month over month or quarter over quarter, we really love that and really getting that continuous analytic time.
And then we get to the most technical with the pure what we call coding, and you think of somebody really doing computer programming, really getting very technical in the details, these are some tools that are very specific to knowing certain computer languages. So those are the most technical ones to do that those are Bradley’s favorite ones.
BC: I’m stuck on Excel, it’s great to see that in there too just to break down some of these barriers and the mystique about data analytic tools.
Excel absolutely is a data analytic tool and I want to go back to something you said as well, “I don’t know how to start with data analytics,” well if you’re an auditor you’re analyzing data. Period. We’re just talking about using an automated tool, something to help you analyze more data, more quickly, and to become more efficient to see trends and anomalies.
It is great to see that you talked about a maturity of data analytic tools and internal audit, so you don’t have to start with learning R and Python you can start with doing a sort in Excel and that is a data analytic tool so it’s great to stair-step this up to help people get over that fear of the unknown and data analytics.
TS: It is great, the maturity of tools. We love Excel, most of the work we do we start with that, but then we want to get more mature, we want to get more abilities, we want to do different things, so we want to look at how does the data relate to its self, there are certain tools that are better to show how the relationships between data. There are better tools that show how to visualize data, and I’ll talk about some of these individually in just a minute, but part of doing data analytics is letting the data tell you a story.
A table on the screen can only tell you so much, but when you can turn it into a graph or a chart or then further into an interactive, where people can work through it, you begin to learn and you get to understand what the data tells you. You get your more robust analytics, you want to take that next step past what can Excel offer and you really want to get into some modeling or some other things, and there’s a whole level of analytics there.
BC: Right Tripp and I think you brought up a good point about all the different tools and this can be paralyzing to some people, you know, “what tool do I need,” and I think that the first thing you really have to start with is, “what question about data do you have,” “what question are you trying to answer,” because depending on the type of data you have and the question you have about that data is going to help you select one of these tools and that is very critical in selecting the right tool for the data you have, and the question you have.
TS: Well and even knowing what some of the tools mean, you see a lot of acronyms and tools, you get kind of scared away, some of the things you hear about a lot is RPA, and what does that even mean?
Robotic Process Automation (RPA) has been a big buzzword in the industry for years, it’s a fancy way to say “let the tool do the work for me,” without trying to take a lot of people’s time and effort, and automate what somebody would do pointing and clicking on the screen. You have tools that can automate workflows, automate business processes, somebody just entering the payables automatically. Some call them robots or other things along that line that automate so, is it analytic or is it just automation? What are we trying to do with it are we trying to automate our fraud test?
BC: Tripp, you know you talk about RPA and that sounds like a big fancy thing, but really all it is, when you were talking about the maturity earlier and writing scripts that are repeatable instead of an auditor kicking it off, the system kicks it off, and it’s set up to run at a certain time. So it’s a big fancy term for just something moving up the maturity scale, and maybe not everybody wants to get to that point, but we’ve always talked about continuous monitoring and what better way to do continuous monitoring than automating a data analytic technique that is going to give you your anomalies and your trends, so it goes back to risk-based auditing.
TS: If you take the next step, we’ve heard about it for a long time and it sounds like a science fiction item; Artificial Intelligence (AI) and Machine Learning.
What we’re really trying to say is “can we predict where the risk is going to be based on past circumstances and predict fraud,” if we think of the pure auditor mindset, “can we reduce the exposure or can we just learn from the past,” and that’s some tools like R and Python where you can get really complicated but the theory is the same; you’re trying to answer a question. Now you’re trying to predict what the question is going to be, or you’re trying to predict what the environment will look like in six months or a year from now and using all your past experience.
So, we moved from the most complex like R and Python to do AI Machine Learning, we can step down into RPA, stuff like SAS, Microsoft Power Automate and there’s several more out there. Again, you can see SAS is more of a coding-based item, whereas Power Automate is a more Guided UI so even within the level of tools you can find a tool that fits your skillset. One may be more powerful than the other, and one may be easier to use, but the tools can get you a lot of the same things.
BC: Tripp you talked about coding, and it reminds me of when I first got into this back in the mid-1990s working for a rather large bank, we were using a focus reporting tool off the mainframe and we actually had to code, we needed to know how many characters were in each field in order to read that field. It was a lot more work and I can understand how thinking of data analytics that way is intimidating but with these tools today, it is a much better user interface, much more point and click and it’s much more friendly to get started, so people shouldn’t have to worry about that we’re not trying to say jump into the deep end of the pool, but you can start with Excel and wade into the very shallow.
TS: Even in the advanced analytics space we have players like ACL, IDEA, Alteryx, Arbutus, they’re all more scripting to use interface, to try to break that barrier doing analytics.
You move back towards a tool on this end, some companies have really taken that more towards the Guided UI folks, like TeamMate Analytics. The different tools have different pros and cons and look for what fits your abilities and what fits your people at their skill set, and that’s really the main thing to think about when doing analytics, is not having the full suite of tools, but having a few here and there to develop your toolbox. Start with what you know, most of us all have Office so maybe start with Excel. Can you do an add-in to Excel? Maybe there’s something simple like starting to graph things. You start to get more mature and start to learn, get a little bit better tool where you can do more analysis.
BC: What I found is that using data analytics with staff is infectious, once somebody starts and everybody starts seeing the visualizations, the fact that we can eliminate sampling risk and we can do this testing so much more efficiently, everybody will start using data analytics. You get that one champion to get started and it’ll spread throughout your staff. So, we talked about some theory here, now let’s focus and get down into talking about an actual use case that we’ve had for some of these analytics.
TS: Let me walk you through a little bit of what I want to talk about, and this use case is a real example of what we’ve done in some areas but it’s not using just one tool. In this case, we’ve run an accounts payable script set that we have. We used ACL to do that work, we’ve built it out, it’s repeatable, we can run it from client to client, it’s a suite of 75 more tests that we’ve developed over time and we can run that in a repeatable fashion.
BC: Tripp I’m sorry to interrupt, but I want to make sure everybody heard what you just said, you ran 75 tests against an accounts payable file that we got from a client. And those 75 tests that you did not have to recreate you were able to repeat it from what you had built up in the past, you just had to bring in a different data set.
TS: Yeah, so you have an investment on the front end to have a scriptable tool, spend the time to build that out, but then running for us from Client A to Client B, we just do some slight modifications, and running in-house from first quarter to second quarter, you just keep repeating. That’s the repeatable maturity process we talked about if you want to continue to look at data over time.
A lot of times we do this we export results to Excel and we go back to the best way to kind of play with stuff or maybe even Power BI to visualize a little bit. You spend time working with a client, you’re going to have false positives, you’re going to have things that pop out and say, “what does that mean?” There’s going to be more questions, what the data is going to start to tell you, what it wants you to know, you have to use the tools.
Again, it’s a simple dashboard, a trend over time, are we getting better, are we getting worse. It’s an iterative process with these tools and that’s why we like to use repeatable tools. Because doing something once in Excel that’s a great one-time test, but you start to want to look at things over time and that’s where you start to get into more wanting a little bit better tool. And it doesn’t have to be ACL, it can be a multitude of things, but the idea is the tools allow you to make this a repeatable process. The tools allow you to diagnose the data, the tools allow you to visualize it, I love a good dashboard that shows the trends over time, how are we last year versus this year, are we tracking improvement.
From a pure auditor’s standpoint, back to Bradley’s mandate, using analytics to look at audit, if I have an accounts payable auditor and I have somebody want to walk through this at the first year we run this as internal audit. We run it every quarter, we track if they are closing the deficiencies we noted, are they closing the audit issues?
Maybe we operationalize this and hand it off to the business, once you get the tools in-house you start to embed those amongst people. Their accounts payable department starts to do their own analysis to say, “what’s the second line versus first-line versus third-line at events, but we’re trying to elevate the risk knowledge throughout the organization using analytics whether its internal auditor or it’s someone at [FDNA?] or someone in accounts payable doing the analysis, the tools themselves are all pretty much the same.
BC: Tripp and what this really helps to do is it builds a partnership. And there’s been many times, where management has asked us as auditors where did you get that information well that’s your data, we’re just looking at it differently we’re analyzing it differently, and then the question comes well can you do this more than just once a year. Absolutely how about we hand it off to you, and it becomes a part of their internal control framework, so now you’re testing as an auditor is, are they running it, are they escalating it, are they responding to what the data is telling them?
Because what this did for our client was it helps them focus their audit, so it goes back to a risk-based internal audit instead of selecting a random sample of 25, 30, 50 or whatever it is. You can look at the entire population and say here are your anomalies investigate those and that provides more efficiency to the audit department and more impact to the audit client, so it follows the standards for using a risk-based internal audit approach and that’s huge and it builds a great partnership, so this is a very powerful technique.
Again, you start with asking the question, what do I want to know about this data and then how do I get this data and what format is the data to help you select what tools and it may not, and I say tools, because it may not be a one solution answer for the data analytic you want to run the visualization is going to be much more powerful than the raw data that we’re working with, but you might not have a data analytic tool that does both. So you may be crunching the numbers in one and presenting the results from another data analytic tool, so there’s a combination between tools and talent that you need to have on your staff.
TS: There are some tools, we mentioned Excel, and you can do a lot of this with that tool, but the more advanced tools allow you to be more efficient. We go back to our list of tools, ones that are repeatable like Tableau is a visualization or Power BI, not only make it easier to do, but it makes your time much more well-spent because you can do it quicker and then how to reinvent this. You can make charts in Excel, so don’t let not having some of these tools hold you back from starting. I know there are more tools out there, but I can get started today with the tools I probably have on my laptop.
BC: Alright, well, this has been Bradley Carrol and Tripp Stedham with Frazier & Deeter’s Process, Risk and Governance team. We hope that you found this helpful and we look forward to helping you find opportunities to implement data analytics, helping you ask the questions, select the tools to run the data analytics with you and we just thank you for attending.