Automation has been at the forefront of global accountancy for decades, and professionals are constantly hungry for new tech innovations that will enable them to improve effectiveness and efficiency levels for clients. Yet while demand has continued to grow for the availability of smarter and more dynamic accounting systems, adoption has been pretty slow – particularly when it comes to artificial intelligence (AI). It looks like that’s finally starting to change.
Computer scientists have been tinkering with AI for more than half a century, and the concept itself is quite simple. By utilising a series of defined algorithms, AI-powered applications are able to draw from big data sources and use deep, machine learning functionality to process data, perform high-volume tasks, recognise patterns and even make decisions – and while that may sound like space-age science fiction to some practitioners, AI-powered apps and modules have already started to make a serious impact across the global accounting sector.
According to researchers at Sage, 58% of accountancy professionals are expecting to automate tasks using AI solutions within the next three years. What’s more, four out of five practitioners say they believe AI will inevitably increase their company’s productivity, deliver added value and offer numerous competitive advantages. Yet while enterprising fintechs have yet to unlock the full potential of AI, the truth is there are already plenty of AI tools available right now that can make an accountant’s job easier.
Streamlining data analysis
Everybody knows that time is money – and data entry takes an excruciatingly large chunk of time out of an accountant’s day. They’re often forced to comb through multiple spreadsheets and documents in order to stay on top of transactions and provide meaningful reports for clients. According to a survey by McKinsey Global Institute these repetitive data tasks needlessly eat up six or more hours per week. Thanks to AI, accountants are starting to reclaim some of that time and apply it to higher value processes.
Apps powered by machine learning can automatically extract data from receipts, accounts and spreadsheets and then classify that data based on pre-defined categories in order to populate dynamic custom reports on-demand. Better yet, AI reporting tools are then able to use deep learning in order to assess patterns from various data streams and go on to provide unprompted insights around budget forecasting, spending and investment opportunities.
Plenty of mainstream accounting solutions providers have already brought these sorts of products to market. For example, QuickBooks Auto Categorization is an easily enabled product feature that learns from manually recorded transactions and then automatically repeats those categorisations and organises future transactions accordingly.
So-called smart assistants like chata.ai, Hey Xero and Sage’s Pegg have even taken that functionality to the next level by pairing AI data entry ability with natural language processing so that accountants can receive automated insights and manage transactions by having conversations with scripted chatbots.
Pegg is a free service that’s compatible with both Facebook Messenger and Skype – and earlier this year, ERP provider Acumatica announced the beta release of a new product that allows clients to ask the Amazon Echo bot ‘Alexa’ for verbal accounting insights on the go. This enables accounting professionals to use voice commands in order to look up and approve expenses, update CRM and ERP entries, ask for spending patterns and more.
While asking Alexa for a spending analysis might feel a bit frivolous to some practitioners, other artificial intelligence applications focus on the very critical task of fraud prevention.
Spending across many corporates is becoming more and more decentralised, which has made it difficult for some accountancy professionals to mitigate risks of fraud and failing to meet various regulatory compliance targets. As with data entry, time is a key factor when it comes to auditing expenses.
According to the Association of Certified Fraud Examiners, corporates lose an average $130,000 per case when it comes to instances of fraud. Yet because most organisations have only got the time and personnel to audit just 10% of expense reports manually, the vast majority of potential fraud cases simply go undetected.
By contrast, scalable AI-powered solutions can conduct a full audit of up to 100% of corporate expense reports – automatically identifying and highlighting anomalies and predicted patterns. Better yet, smart auditing apps aren’t prone to the same sort of human error that often results in time wasted and funds unrecovered.
California startup AppZen is one innovative solution accountants are turning to. It integrates with existing finance systems and uses artificial intelligence to audit expenses, invoices and contracts immediately at the point of submission. It then extracts words, amounts, names, rates and other data from documents in order to contextualise that information and then cross-check it against company policy, regulatory responsibilities and contracts.
Finally, the app automatically assigns a risk score to each transaction and flags lone, high-risk items deemed to require a more serious manual review. In turn, accountants are more likely to catch duplicates, suspicious merchants and more.
Because the AI system understands common terms and brand names, it can even draw from keywords within line items in order to highlight expense violations like alcohol, personal credit card usage and travel add-ons. AppZen scans the internet for a restaurant’s average prices and compares it to the total amount cited within a given expense report in order to identify excessive spend.
On the flipside, this sort of app not only highlights violation of corporate spending policies, but it can also flag whether certain policies are actually working for the company. For example, if employees are regularly overspending on travel, repeated audits may suggest companies agree to raise thresholds to a more reasonable level.
Not only does this auditing save clients time and money, but it also assists corporates in meeting anti-bribery obligations like the Foreign Corrupt Practices Act.
Addressing the data challenge
There appears to be a certain sense of inevitability when it comes to AI in accountancy. Although just over half of practitioners are expecting to implement an AI-based system within the next three years, 90% of accountants agree that there appears to be a cultural shift taking place across the sector. That being said, it’s worth pointing out there are still one major barrier stifling mass adoption.
Availability and access of data is absolutely critical to the success of any AI-powered accounting solution. That’s because they require huge volumes of data in order to spot patterns, carry out process automations and learn to analyse information – and that level of information can only be facilitated via cloud computing.
The cloud enables data to be fully accessible and easily collated by artificial intelligence solutions without having to deal with the friction of incompatible systems or old-fashioned data outputs. But according to researchers at IDG, more than one in four organisations still don’t have any cloud-based infrastructure. A further 10% of corporates say they do not plan on migrating within the next three years.
Without that access, artificial intelligence solutions cannot hope to offer 100% accuracy when auditing or carrying out data entry processes because the system may not be guaranteed full access to data sources. That being said, there is a certain sense of inevitability where migration to the cloud is concerned – which is why such a high proportion of accounting professionals are already poised to begin integrating AI technology within their existing systems.
The artificial intelligence revolution won’t happen overnight. Instead, most firms are likely to experience an incremental move towards AI-based systems as existing accounting solutions providers slowly introduce bold-on modules and system updates that embrace machine learning and automated functionality. Yet at the end of the day, it appears the sector is certainly bracing for change – and if AI continues to expose greater potential, life as an accountant could soon become a whole lot easier.