Request a demo

Get the latest posts from CrowdReason delivered straight to your inbox.

The Future Of Tax: Man + Machine

Posted by Carl Hoemke on Aug 13, 2020 9:49:19 AM

Future of Tax: Man + Machine

These days there’s much talk about artificial intelligence (AI) and machine learning. Will robots take over our jobs, or won’t they? What can we expect when these technologies ramp up? AI and machine learning technology will have a profound impact on the workplace. In many ways, they are simply advancements in a long line of improvements helping people do their jobs better. Computers changed the way people worked; the internet changed the way people accessed information, yet a succession of new computer applications will need mastering before they are useful—and this can take time.

No matter what industry you’re in, AI and machine learning are likely to touch it somehow. As a tax professional and the head of a technology company, I’m excited to see more and more teams embracing the latest tools now, rather than waiting for some “right moment” to arrive. But the future of tax is still to come for most organizations. What will tax professionals of the future be doing, and what role will technology play? Most tax teams will see the most significant benefit in the area of data analysis.

Wonder if your property tax team has the right mix of skills for the years ahead? Download our white paper, The 21st-Century Property Tax Analyst Hiring Guide, to find out.

The Future Of Tax: Data-Based Decisions

The volume of tax data is immense, and every year it gets broader, deeper, and more abundant. Historically we’ve only been able to skim the surface of this sea of information. As a result, tax professionals have limitations in their ability to:

  • Make accurate predictions—Professionals in all areas of tax need to determine expected outcomes based on historical data. A lack of big data tools meant they had to make decisions based on a cursory review of a small number of data samples, hoping the examples were right. Ultimately, the conclusions would apply to historical experiences.
  • Be strategic about tax planning—Whether it’s the level of detail on a tax return or the range of variables on property tax payments, there’s always a way to optimize your practice to achieve better outcomes. But much of the granular data needed to formulate advanced strategies have been difficult to isolate and quantify, limiting our ability to design the most advantageous tax strategies.

Technology will advance and become more powerful in the years ahead, giving us tools to do our job better. Think of it as a comparison between an old-school television screen’s pixelated image and poor color reproduction, and today’s modern video monitors with minimal pixelation and colors that pop. Similarly, advanced tax technology tools will allow us to “see” what was previously hard to identify, giving us multiple dimensions of data harnessing that benefits the organization.

For tax professionals of the future, sample data sizes will get more abundant, and technology to review and understand them will continue to improve. Some of the tools will be more traditional (like databases); others will be new applications that use advanced algorithms to extract value from vast granular data within those databases, giving us more predictability and more significant insights. As technology gives us more tools, we can consume more data and delve deeper into the data sets.

Will artificial intelligence displace tax professionals in the future?

There’s no doubt that artificial intelligence and machine learning applications will continue to expand in future years. However, humans in specific tasks are far superior to machines (for the time being). Tesla’s problems with “over-automation” of the Model 3 line clearly showed that robots aren’t as good as humans at handling complexity, inconsistencies, and process variations. For example, robots can apply torque consistently, but can’t account for bolts that don’t fit or fasteners that don’t align. When there’s an unknown aspect or something that’s not highly repeatable, it requires a higher intelligence form to solve it.

The same is valid for tax. Machines are powerful data-crunching tools, but even machine processing benefits from some human insight. For example, computers can apply deep learning algorithms to terabytes of data and make predictions by extracting insights from relevant data sets. While this process is efficient in some ways (it doesn’t require a person to cull through that data and curate/format it), it’s inefficient because deep learning is computationally burdensome and expensive and requires gobs of data. Machine learning, another form of artificial intelligence, works best when humans provide some degree of foundational truth to interpret the data. Training a computer with this truth allows for less burdensome automation, and provides a more efficient way to get the job done when terabytes of data are not available.

As we move into the future, computers will increase capability and successfully perform more humanlike rational-based activities. (Experts predict computers might be able to handle more complex human decisions by 2040 or 2050.) But until that happens, humans will continue to work alongside technology to help manage the continued growth of data and its interpretation.

Advanced Tax Technology In Practice

Advanced tax tools are on the market now; there’s a lot more available than what’s currently in use. Organizations that are taking advantage of cutting-edge technologies now will have an edge in data management and analysis.

CrowdReason has been on the leading edge of tax technology explicitly designed for property tax professionals—it uses machine learning in combination with human expertise to help teams manage abundant data and extract more value from it. (You can read more about how we use crowdsourcing and machine learning to handle data entry in our whitepaper, The State Of Property Tax Technology.)

In particular, property tax data is still very non-standard, with many jurisdictions, all of which have different rules and provide data in various formats. Our MetaTasker software extracts data from tax documents typically passed over by teams doing data extraction manually (simply because they don’t have time to extract it all).

Before you might have a team member looking at tax documents and noting the amounts due and account numbers, whereas, teams using MetaTasker leverage our machine learning application to cost-efficiently get all the granular information included on tax documents, such as:

  • Tax jurisdictions associated with the property
  • Tax rates
  • Taxable values
  • Exemptions
  • Penalties
  • Discounts available
  • Variability in terms of deadlines/appeals/payments, and more

Many teams use MetaTasker in conjunction with our property tax management system, TotalPropertyTax (TPT). Together MetaTasker and TPT help teams glean more insight into the data collected and apply it in their property tax practice. More data means they can make more advanced decisions, such as taking advantage of discounts that they may not have been aware of or identifying errors that would have otherwise induced penalties. TPT embeds business intelligence functionality to guide users in making decisions about certain tax positions, and a full set of depreciation tables, indexes, tax forms, and jurisdictional rules for all assessors and collectors, making tax filing and management more effortless.

For CrowdReason Users, The Future Of Tax Is Now


Technology is useful, but your people are your most valuable asset. Tax professionals do their highest-quality work when doing what they do best: strategizing around taxes, not doing mundane tasks like data entry. If you’re interested in nurturing the tax professionals of the future on your team today, get in touch. We’d love to show you how our applications can help.

Download Now: 21st-Century Property Tax Analyst: A Hiring Guide

Topics: Artificial intelligence in tax