Ep. 4: Daniel Smith - Opportunities Created by Data and Technology

Daniel Smith, Head of Innovation and Founder of Theory Lane Integration Solutions, puts all accounting and finance professionals at ease as he simplifies AI, machine learning, RPA, and other emerging technologies and explains the various opportunities created by data and technology. He spoke to us at length about all things affecting the evolving role of the accounting and finance professional. From disruptive technologies, to quantitative vs. qualitative data, to data science and data engineering, and all related concepts and tools, Dan answers a wide range of questions to help management accountants understand these important competencies. Stay tuned for the conclusion of the episode! [ Now available! ]

Hey everybody! Welcome to Count Me In, IMA's podcast about all things affecting the accounting and finance world. I'm Adam Larson here with Mitch Roshong. And this week we covered the topic of data analytics and emerging technologies in accounting and finance. Mitch, can you tell us more about it? 
 
Mitch: (00:16)
Thanks Adam. Yes. So earlier this week I had a great conversation with Daniel Smith. He is the head of innovation and founder of theory lean integration solutions. I reached out to Dan because I wanted to ask him a few questions about data analytics as it pertains to accounting, but our conversation got so in depth, we ended up talking for well over an hour. Dan offers a great perspective on the impact technology and data have on the accounting and finance world. I really enjoy talking with him about these topics and I'm excited for you to listen to the first part of our conversation.
 
Mitch: (00:54)
Being that technology seems to be very disruptive in today's world between the jobs that it's overtaking and the economy. Many people seem to be a little scared about these emerging technologies. So my question for you, in your opinion, how do we as humans prepare for a future that's kind of dominated by this artificial intelligence or machine learning, blockchain, all these buzz words like, that.
 
Daniel: (01:18)
I get this question a lot, Mitch. It's an excellent question and a valid concern for anybody who's worried about that or who is kind of on the periphery of this space. I will tell you, anybody who's deep into this space is not that worried about it. You'll hear a lot of doom and gloom and fear, uncertainty and doubt and doubt being spread. This world that's dominated by artificial intelligence, machine learning, robotics, blockchain, whatever word you want to use for it isn't as different as people foresee it to be. AI machine learning. Those solutions aren't giant monolithic replacements for humanity. If you think about it as any time that you repeat an action, something that you do over and over again, if you try to speed it up and you feel like, Oh, I'm gonna get so fast at doing this. I know all these keyboard shortcuts and I know all these things and I'm going to optimize this. That's a optimization problem. That's a locally optimizable solution. Those are the types of things that AI can do well. It can make a little bit of problem solving for itself. It can make slight judgment calls, but it can't deviate from the path that's been set in front of it. Look at it as a intern if you will. You get an intern to do a single thing really well and make a little bit of judgment for itself. But the second that there's a deviation, it's going to spill coffee all over itself or do something that's a mess. AI is exactly the same way. So we can think of ourselves as we learn programming and as we learn basic technical competencies, we'll figure out how we can act as managers for AI and how we can have the business itself be focused on deployment of these small applications and these small things to help the business. It doesn't hurt us. We just have a different skillset that we're using to benefit the business or benefit or even do business for us. So it's not scary. It's just a little different, which also can be scary until different is normal and now it's not scary anymore. 
 
Mitch: (03:54)
That's right. So I kind of see this from what you're saying is we need to look at this as more of an opportunity rather than something that's scary.
 
Daniel: (04:03)
Absolutely. We should always be learning and developing our skills. This is just another opportunity to learn and when there's an opportunity to learn, there's an opportunity to apply it and grow our career. 
 
Mitch: (04:18)
That's a great perspective. I really think you know, a lot of publications coming out now, you know, there's a lot of articles and a lot of research that's going into this and the more tools that are becoming available to us you know, it's more stuff to learn. But again, it in my opinion, looks like it could potentially increase jobs and, and create new ones. So I know the whole field of data you know, we talk about financial data more or less in accounting. But I think there's other data that goes into, you know, a successful business. So my question to kind of follow up on this at what point do you kind of see maybe an organization suffering from what people are saying is like a data overload? How do you figure out what's really meaningful for your business and what should be used with all of these new opportunities? 
 
Daniel: (05:14)
Hmm so this is a great question and another one that I get a lot. Okay. Every question is a great question to me because if somebody's asking it, it means that they're interested. So I love, questions, period. A lot of people don't even bother to ask. I digress, though. I don't actually look at it as it being anything different in the emphasis on data. Any accountant knows that data is the lifeblood of an organization. If anybody me present, I always talk about the data creation to value process. How any stakeholder or business activity creates data in some form or fashion. What we used to do and nobody thought anything of it was we would have our general ledger, our general ledger would take all the purchases and sales, the activity, payroll, et cetera, all the activity of the business, put it into a transactional system, a general ledger. We would then take that information and we would clean and shape it. We would aggregate it into our various financial statements and financial reports that would then subsequently be used by business specialist to make business decisions based on that information. Say what volume of inventory should I order in support and correspondingly, what should my sales budget be in order to sell that inventory? That would result in our sales and marketing mix. That sales and marketing mix would then result in stakeholder activity as they purchased products, resulting in more information that would lead to other business decisions. It's the same thing we've always done. It's just a little faster now and we're learning. If we can generate more data, we can create more opportunities for making business decisions. Sometimes that needs to be accelerated and the problems that are being created that need to be solved through having all this data no longer fit within the traditional bounds of an organization. Remember, all the organizational specialties that were created were as a result of what is effectively the accounting information process. We only were able to put paper on a spreadsheet or a t-table or a clay tablet, all of which are proxies for the same thing. Now that we're in a digital environment, data has a whole new way of moving around. So we have to reorganize. We have to have these cross functional specialties or even completely new functional specialties. Data science is the mechanism by which we break down the barriers, traditional barriers of business information and create an entire new organization structure. Companies are not actually being that overloaded with data. It's the same amount of data they always had. They're just able to do new stuff with it. So they're having to change their entire organizational behavior and structure to adjust to it. It's that was a very long answer to the question and I think I answered a few other questions with it. 
 
Mitch: (09:13)
I was gonna say that actually that kind of leads into my next question and covers a little bit of it, but because so much of this data has historically been rather financial I know there is a big focus on operational data and more of the qualitative aspects of the business. So I'm just curious if you have any experience in improving business because the organization was able to see something based on analyzing the qualitative data that they have in front of them now. 
 
Daniel: (09:48)
Yeah, qualitative data in my world, it's funny. qualitative data doesn't, I hate to say anything that's definitive because it doesn't really exist, but I think, I think that there's definitional difference cause they're saying that it's qualitative as opposed to raw financial data. And if I understand the question correctly, is that what they're considering qualitative? 
 
Mitch: (10:26)
Yeah, I believe so. Yeah. More and more of the the operations as opposed to the dollars. 
 
Daniel: (10:31)
Yeah. Okay. Okay. That makes sense then. So like a HR survey or maybe even something really qualitative, like the sentiment of emails or tweets. Got you. Okay. Yeah. The trick then, if I can pivot the answer to this a little bit the trick with any qualitative data is you stop making it. So qualitative you try to put as much of a sense of objectivity as you can so that there's some quantitative value to it. It's half of analytics and data science isn't knowing the right tool for the job. It's being able to shape the problems so that it fits the tools that you have. What you'll see a lot when it comes to that type of qualitative data are a, what are called natural language processing tools. I've used them a lot in the past in marketing engagements where you'll use sentiment analysis, natural language process called sentiment analysis, which will look up either a dictionary words or similar words. You can, you can apply math to terms and sets of terms to see if they're similar to things that are generally positive or other blocks of words that have been labeled as positive, to assign a general sentiment. Are they angry? Are they happy, are they supporting this activity, et cetera. With that value. Now we can use those in traditional models to say that, well, when emails to HR have a general positive sentiment, we feel that our productivity increases over the next few weeks. So we should identify what patterns of behavior resulted in those emails so that we can enhance productivity or when this, when we use this particular marketing, we see the sentiment of tweets increase in this region. Correspondingly, we saw an increase in sales. There's a bit of correlation and causation there that you also have to tease out. But the story itself is that when we've been able to influence and when able, we've been able to see a market increase in that subjective or that qualitative behavior of our audience. We're stakeholders. We've been able to also demonstrate a quantitative effect of that qualitative information. 
 
Announcer: (13:28)
This has been Count Me In, IMA's podcast, providing you with the latest perspectives of thought leaders from the accounting and finance profession. If you liked what you heard and you'd like to be counted in for more relevant accounting and finance education, visit IMA's website at www.imanet.org.
©2019, Institute of Management Accountants, Inc.