Tech innovation bright ideas

We will need to take our firms through a technology transition in the next five years that is unlike anything that has come before. Prevent disruption to your firm by being proactive on artificial intelligence, machine learning and robotic process automation.

By Randy Johnston

While you may believe that the transition from DOS to Windows, the arrival of the Internet or the use of cloud technology were big things, artificial intelligence (AI), machine learning (ML) and other emerging technologies such as robotic process automation (RPA) will deeply change the practice of accounting. How can you help keep your firm on the right track? It certainly won’t be done by doing the same old thing, but the risk won’t be mitigated by jumping to new technology too early, either.

The VSCPA has been proactively scheduling CPE courses to help you investigate emerging technology options. However, we each use different technologies, we have different client needs and we have different strategies for our firms. Remember that you are still well served having a strategic plan and tactical objectives for your firm, which are then incorporated into a technology strategic and tactical plan. We still must get day-to-day work done in accounting, payables, tax, audit and other core services. Each business area and application has development in progress that will apply AI, ML and RPA. The key question you should ask is: When does my area of responsibility need to apply an emerging technology?

Today’s top AI and automation opportunities

Wherever there are a lot of transactions or data, there is an opportunity to automate. You have already seen the opportunity to reduce the amount of effort needed to use bank or credit card feeds to capture all transactional data if you use online accounting software. Some of you are using expense management software to capture transactions and reduce the effort to manage expense reporting. These simple examples leverage the benefits of centralized cloud computing.

Over time, computing power has gone from centralized to distributed back to centralized and is preparing to shift back to more distributed capabilities. The new wave of distributed computing is a result of widespread use of mobile computing power and from the reduced cost of graphics processors, which can accelerate machine learning and artificial intelligence. Recognition capabilities, commonly referred to as Optical Character Recognition (OCR) have continued to improve, but adding AI, ML and other techniques improve accuracy. Having a local graphics processor can speed up AI and ML. If a document is presented in paper form, the accuracy of translating the image’s numbers and words continues to improve, approaching 100 percent. Of course, it is better to capture the data in the original format, which means we will download the transactions from a bank or from an accounting software system. Another rule to apply is that if an accountant is keying data, there is a broken process somewhere. Your team should not be keying data at any level.

We have seen and experienced the first wave of computer automation with document management, expense management and accounts payable management with tools like eFileCabinet, Concur and Altec doc-link. While some of these products are approaching 10 years old, their capabilities were largely built with manual forms and compute power, not with AI and other advanced techniques. New generation tools like Receipt Bank Extract, Zoho Expense and Bill.com are leveraging emerging technology to improve their capabilities and recognition rates.

Consider what happens if we can ingest a large amount of data for a tax return or an audit, perhaps even all transactions, to look for irregularities? What happens if all, or at least most, business transactions can be accurately captured and classified? Both functions are happening now with first-generation products such as Inflo and VSCPA partner MindBridge Ai. The algorithms in use are improving in accuracy as more data is processed, and the rules in the algorithms are improved. Mainstream and start-up vendors are beginning to sell these capabilities. Imagine what can happen if we can apply blockchain techniques to ensure that all recorded transactions can’t be changed (=immutable). We are seeing this type of product with tools like AuditChain, which is also leveraging XBRL reporting.

Tools available today

Multiple publishers are marketing products claiming that they have solved these problems. While their techniques are beginning to work, don’t expect 100 percent solutions. Examples of products that are working for audit include Wolters Kluwer Audit Accelerator and TeamMate Analytics. MindBridge is making progress on ingesting and classifying a large amount of data, while new audit tools are being promoted by the competitors of AuditFile, MyWorkPapers and the American Institute of CPAs’ OnPoint PCR.

We can bring back feeds and expense documentation together with a wide variety of expense reporting tools from Zoho Expense to Tallie to Expensify to Nexonia. We can automate payables with over a dozen different products from CloudX to Anybill to Aviid Exchange.

Note that if a process can be handled by a human, there are tools that can automate these processes as well. Reports from large organizations demonstrate that RPA is paying off very well financially. RPA is an emerging form of business process automation technology based on the notion of software robots or AI workers. RPA can be attended or unattended (similar to machine learning). When RPA 2.0 arrives, we will have more integration to AI. Today the tools are more like programming and scripting. The benefits of RPA include productivity, cost savings, risk management, innovation and employee engagement. The use cases (and more) include QA, patch deployment, load testing and data entry.

Possible tools to consider include (pay particular attention to the platform across the tool chosen): UiPath, Automation Anywhere, Auto IT (scripts, local workstations) and Blue Prism. What is done well with RPA? Any manual processes that can be automated, repetitive tasks, rule-based activities, changeable tasks that are parameter-driven, electronically readable format content, mature/stable processes, processes with low exception rates, areas that have high volume and any places where you can achieve FTE savings from automation.

There are three levels of complexity with RPA: low (recorder, power users), medium (some simple scripting) and high (developers). To get started on a project, you’ll need to engage the stakeholders (tax/finance), have solid process mapping/prioritization, look at the return on investment, find and develop a tech/proof of concept (use case before, vendor build) and repeat to improve. We suggest that you target work with least realization or areas that you believe will gain benefit from optimization, standardization, quality or innovation.

We are not done yet!

What can you do now? Recognize that if you keep your technology current, vendors will make new offerings available to plug into your technology infrastructure. Consider services offered by your firm today or ones you’d like to offer in the future. Create an innovation lab, or sandbox, that can be used for testing of new applications and techniques. Listen to your clients or customers’ needs carefully while you are trying to assemble an offering. Test the offering with a few trusted clients and expand the offering into more of your base. Remember that you don’t want to give away the efficiencies gained by reducing fees or prices. Over time, poor competitors are likely to compete on price, but initially you should compete on value. Use the extra time made available by the efficiencies of the application(s) to provide a higher level of service, perform more business development, reduce the workload of your team or allocate the time to other worthwhile projects in your firm.

Expect some of the applications to not work out as you expect, and act quickly to set these options aside. You may want to hold your work for deployment at a later time once the quality of the products improves and/or become more affordable. Remember to package service offerings as a product and have a marketing documentation and procedures to support the offering. Revise the procedures and offering frequently until you have it finely honed to fit your client base.

There is broad FOMO (fear of missing out) and vendors are selling many partially baked offerings that aren’t really what is advertised. The keys to the offerings are: Do they work sufficiently well today to be useful? Are they affordable? Do clients and/or your firm have a better experience while using them? If so, you can implement innovative products today and be ready for the emerging technologies of tomorrow.

Randy Johnston is a shareholder in K2 Enterprises, LLC, a leading provider of CPE to state CPA societies. He also owns Network Management Group, Inc., a managed services provider that provides around-the-clock support from Boston to Honolulu. Concepts for this article were extracted from the Emerging Technology session produced as part of the 2018 K2 Technology Conferences and from Johnston’s own experience working with technology at various firms in the United States. Ask for help at NMGI by emailing helpdesk@nmgi.com or call (620) 664-6000.