When was the last time you or anyone on your property tax team went home at 5 P.M. on a workday during tax season?
Sounds impossible, right? Well, that was the case with one of our clients, whose office I stopped into recently at the end of a workday. The place was deserted. When I asked where everyone was, the response was, “They went home. They’re done!”
That small miracle could be attributed to technological advancements in robotic process automation (RPA), machine learning, and artificial intelligence (AI)—and the team’s wise decision to embrace these concepts and apply them to their own tax work using CrowdReason software.
RPA, Machine Learning, & AI Defined
There are definite distinctions between these technologies, though sometimes the lines get blurred.
Robotic process automation (RPA) is an application that performs highly logical automated tasks. These types of tasks can be easily performed if there are clear conditions associated with carrying out the task; for example, “If this is true, do this. If this is false, do that.”
Machine learning refers to the science around teaching computers to progressively improve their performance on a task. In contrast with RPA, which is logical and condition-oriented, machine learning requires the computer to have some degree of cognitive capability. The computer must be trained to detect data patterns or relationships that will then help it to draw conclusions.
Artificial intelligence (AI) refers to computer systems that can perform human-like tasks. There’s no manual teaching component involved here like machine learning, but rather it's about feeding a neural network with large amounts of diverse data to create cognitive like answers.
Why should these concepts be applied to the tax industry?
Having been an advisor in the property tax industry myself for 25 years where I lead a compliance practice and valuation advisory, I’m intimately familiar with the challenges faced by property tax teams. Not only do property tax practitioners have to keep up with multiple deadlines and due dates, but they also have to accommodate a wide range of the skill sets. The tax process involves an extraordinary amount of complexity, and a massive number of potential pitfalls. Simply getting through the tax season without any mistakes or missteps is a feat.
At CrowdReason, our belief is that much—if not eventually almost all—of tax compliance can be automated. That’s why we found a way to incorporate RPA and machine learning into our property tax software. Filing returns, tracking notices, processing tax bills—automating all these things greatly simplifies your ability to comply (and keep up) with jurisdictional requirements, making your job easier.
And yet the value of these technologies goes beyond simply making things easier. Being bogged down by data entry and information tracking keeps highly-trained tax professionals from utilizing their skills to the fullest extent. Rather than focusing on meeting deadlines and avoiding penalties, their skills are best applied to more high-value activities. For example, they are better tasked with answering questions such as: Is our business property overassessed? If so, how can we get it reduced? What data do we need to collect and present to do that? We’re passionate about helping tax professionals do the job they were meant to do—and the one that will benefit their organizations the most financially.
How can machine learning and robotic processing automation be applied to the tax process?
1. You can use RPA to automate repetitive tasks.
Many property tax professionals perform the same tasks repeatedly for asset management and report generation. They constantly click the same buttons to generate reports, for instance, or follow the same series of actions to retrieve a list of assets each time one is needed. But thanks to RPA, manual, repetitive tasks like these can be automated.
Using RPA, you can program your computer to go to certain websites or legacy applications (like Oracle) to run tasks. For example, if you traditionally input account numbers or asset ID numbers onto a spreadsheet then run a report with particular filter criteria, you can automate the process so those numbers are filled in ahead of time in a grid. RPA will then mimic your actions of clicking on buttons and setting up filters, and generate the report for you.
RPA can overlay a legacy application as described above; it may also be embedded within an application, if that application is advanced enough.
Want to see how CrowdReason software simplifies the tax cycle using embedded RPA and other advanced automation techniques? Schedule a free demo to see it in action.
A number of property tax processes would benefit from RPA. In data validation, you can check a list of fixed assets against the information in a general ledger; and in data entry, you can input data into a tax return application.
2. You can use machine learning to extract key data from tax documents.
Your team probably spends a lot of time on three foundational activities: classification, taxonomy, and data extraction. Before anything else in the tax cycle can be done, your team needs to identify the documents that come in and who they came from, define what useful information is on the document, and extract the data.
CrowdReason’s MetaTaskerPT software simplifies these tasks. It’s designed to use a combination of machine learning and crowdsourcing (outsourcing work to a large pool of people online) to make data extraction easier, faster, and more accurate. Here’s how it works:
- To classify the document. MetaTaskerPT poses questions to on-demand labor platform workers, asking them to identify the document: What is the document type? Who is the collector for this tax bill? Who is the assessor for this assessment notice? An answer is deemed correct if is produced by three separate people. (This tactic also improves the accuracy of the data collection as opposed to relying on just one team member to get it right.) This classification process helps determine the type of document being extracted.
- To define the taxonomy of the document. Again, MetaTaskerPT poses several questions to the crowdsource workers, including: Is there an account number? How many payments are on this tax bill? Is there a discount available on this tax bill? The taxonomy process defines what is on this type of document. Once we have a consensus in triplicate, we move on to the third stage.
- To extract the required data from the document. MetaTaskerPT now understands what the document is and what’s on it, so the software can ask those same crowdsource workers for the specific information we need.
So we build algorithms around the gathered data to help the computer system get answers to questions that would otherwise require people to make those determinations. Over time, MetaTaskerPT begins to recognize documents and is able to replace those human data processes with machine-generated answers.
Elevate Your Property Tax Team’s Value With Robotic Process Automation & Machine Learning
Whether you provide property tax compliance services or work as part of an organization’s tax team, you can benefit from using machine learning and RPA. In our experience, organizations get greater value and financial return from a tax team that spends more time on strategic knowledge work instead of administrative tasks; tax professionals also appreciate the fact that they can contribute to the bottom line in a more substantial way.
If you’re interested in learning more about CrowdReason property tax software, including MetaTaskerPT for data extraction and TotalPropertyTax (TPT) for tax management, contact us. Or, request a demo to see it in action. We’d love to discuss your team’s unique needs for property tax software and how we can help.