This post first appeared on IBM Business of Government. Read the original article.
Engaging with tax agencies is hard: our digital experiences with other industries lead us to expect more.
In our first blog of the series, we discussed the future of tax as it relates to tax agencies. In our second, we wrote about the future of tax and the taxpayer. This final blog post discusses simplifying the tax process and reducing risk.
In many cases the rules are complicated, difficult to follow, and assistance is scarce and slow. Filing reports is time consuming, and often disconnected from where people administer their finances, whether as individuals or on behalf of organisations.
Instead, taxpaying individuals and organisations should experience superior, highly personalized, and effortless digital interactions simplified by artificial intelligence and underpinned by automated integration.
Two opportunities stand out:
- Tax agencies must excel in the digital services available to engage citizens, providing highly personalised interactions
- Tax administrations must address the disconnect between tax authorities’ software systems and applications used by taxpayers to conduct and manage their business
The challenge of human engagement with tax agencies
Taxation rules are often complex but engaging with education, guidance and support for the taxpayer can be difficult. Navigating online help is time-consuming and convoluted. Call centers with experts may be available with official reports suggesting queues of no more than 10 minutes, but both individuals and organizations regularly report much longer wait times to access support. For the 2022 filing, the IRS reported waits of 29 minutes, with only 10% of calls reaching IRS employees. Whilst at best an inconvenience for those paying personal tax, this is a costly overhead for businesses whose staff or tax advisors are forced to wait unacceptable times to access required assistance.
Artificial intelligence to expand engagement
Instead, tax authorities must provide virtual assistants that understand taxpayer intent and can predict the best way to meet their needs, with seamless hand-off to appropriate public servants where necessary.
The virtual assistants we recommend are not the common, simple, rule-based chatbots, built to represent frequently asked questions, collect information or schedule appointments. These are cognitive agents powered by deep learning to understand what the individual is attempting to achieve, and to support the complexities of taxation. They learn from conversations, execute actions, and improve themselves over time. In addition, they can recognize sentiment and adjust behavior accordingly. A cognitive agent dynamically taps into a broad and constantly growing set of relevant information to find the most accurate answer to a taxpayer query.
Acceptance of such assistants is greater than ever. Before the COVID-19 pandemic, there were some who did not like to engage with digital support channels, who instinctively pressed “zero” to speak with an operator. Human agents became scarce; many were given no option but to engage with AI for customer care and have not gone back.
Furthermore, tax agencies must increase the efficiency of the ever-decreasing workforce and address spiraling costs to government from debt, the pandemic, climate crisis and other global events. The average call center cost is $3-4 per call; bot interactions average 25¢. Furthermore, a virtual assistant is not measured on call completion or hours available. It’s always on, and it encourages taxpayers to have a conversation at their own pace thereby increasing flexibility in the process.
In fact, recently published customer success stories have shown that virtual assistants can deliver:
- 30-40% operating cost reduction
- 300 to 500% return on Investment within 6 months
- 20% increase in workflow efficiency
Indeed, in one tax agency email questions fell over a 4-month period, from 1000 emails per week to 50 after a virtual assistant was introduced.
Virtual assistants are crucial to reduce burden on both the taxpayer and the tax agency.
Consider now the second point: disconnected applications.
Extending engagement into taxpayers’ natural systems
As a minimum, tax agencies should provide gateways for taxpayers (individuals and organisations) to file reports and for businesses of all sorts to easily provide information relevant to taxpayer earnings, investments, and so on. For example, in Australia, many commercial transactions with tax implications flow into a gateway offered by the ATO. By reversing that data flow, money can be transferred to taxpaying people and organisations.
However, agencies must plan to extend the boundaries of their tax systems into the natural business systems used by taxpayers, forming digital tax ecosystems. This must cater for both the organisations that are likely supported by technology such as ERP systems, but also the individual taxpayers whose earnings may come from multiple sources, who may run “side hustles” alongside a main salary, and who are increasingly likely to use technology to manage the (financial) intersection between their lives and businesses.
These digital tax ecosystems are already taking shape in countries such as Australia and Singapore and are made possible with APIs (Application Programming Interfaces), digital interfaces that enable different software systems to communicate.
These digital touchpoints simplify interactions with tax authorities, but also naturally build compliance into tax-related processes through timely automation of the flow of information such as rules, payroll data, and financial statements. APIs are used in some countries today to allow companies to file returns directly from accounting software for income tax.
Making the change – lessons learned
Forbes tells us 70% of digital transformation projects fail due to siloed data and unreliable integration approaches. Tax authorities therefore require an integration platform that allows them to manage the lifecycle of APIs – with governance and version control – and to respond to events in real-time. They must be able to securely access data quickly, no matter the systems where that data resides, and must offer multiple integration styles to optimize for the taxpayer applications with which they are connecting. They must support digital sovereignty but also cater for connectivity with global organizations that are subject to tax rules in multiple regions.
When it comes to how the taxpayer engages, to increase effectiveness, simplify the process and reduce risk, the modern tax administration must offer digital touchpoints – with powerful, AI-enabled digital services. They must now extend their digital boundaries into the natural business systems of taxpayers, making this possible through collaboration with other government organisations, and providers of finance and management solutions to be used by taxpayers.