Ask most business owners to name the most enjoyable part of their work and few would answer with “accounts.”
Bookkeeping is one of those tasks that can eat up a lot of time, requires constant attention and can divert you from the nuts and bolts of your business. Fortunately, machine learning technology is here to help and promises to take most of this tedious work off your hands.
Machine learning defined
While many people may equate machine learning with artificial intelligence (AI), the term refers to the practice of using computer algorithms to examine data for numerical patterns and, as a result, taking action or making a decision. You may not realise it, but machine learning is already helping to power Google’s intimate knowledge of you, Facebook’s targeting of your interests and every product that Amazon recommends – and it’s about to get even more powerful.
Machine learning works when software algorithms running on powerful computers in the cloud can ingest and process large data sets, and then learn from the connections between items and begin to take corresponding actions for themselves, continuously improving with each action.
Integration into SMEs
This all begs the question: how could machine learning help in accounting firms? The answer is simple – by increasing your accounts’ accuracy and by decreasing the time you spend on the books.
In a recent Small Business Survey conducted by Pandle in the UK, 49% of respondents said that they had lost money because of errors in their self-assessment submission. These errors could include miscategorising expenses, wrongly treating value added tax (VAT) on certain transactions, or incorrectly matching deposits to an actual sales invoice, and they arise because busy business owners – who are neither experts in, nor excited by bookkeeping technicalities – make understandable mistakes.
Learning from experience
Already today, some cloud-based accounting services use machine learning to take the pain out of expense categorisation. At first, you may have to manually reconcile yesterday’s £6.99 purchase from your office supplier as “stationery” – but with a machine learning algorithm, running on a system with many other customers, your system can and will learn from the categorisation they have already carried out, and simply file these types of transactions away for you automatically.
This type of automated action will compound in the future, meaning that actions will get even more accurate, while new variants of machine learning will be applied to other areas of your accounting practices.
In time, it is likely that almost every part of your bookkeeping will become automated. Long gone are the days where you or your accountant will need to manually input, categorise and submit things for yourself – now an intelligent accounting platform will carry out most of these tasks, only prompting you with occasional alerts where your clarification may be needed.
What will be the effect of automating your accounting? Business owners will save significant amounts of time that could be better spent in more impactful areas of their business, like producing over-arching reports, examining cash flow, planning for the future and achieving scale and growth.
For your accountant, the view will look a little different. Many companies will be able to manage and submit their taxes without having to enlist the help of an accountant at all. Whilst some businesses with specific needs and nuances will continue to require a personal service, others will simply need advice, pre-submission checks or none of these services at all. It is likely, in fact, that the role of the accountant will change to provide a higher-touch, higher-value service than at present.
Automating the future
In a world where algorithm-powered automated software drives your accounts, you’re free to carry out other work and act as an additional layer of certainly – on top of machines and software acting as the primary point of contact for customers. This will dramatically invert the way that accounting – and many other professions – are carried out; turning consultancy in to the tail that wags the dog, rather than the other way around.
Machine learning is only going to get more sophisticated. Today, many services are writing and deploying rudimentary machine learning capabilities using their own code. However, more of us are planning to move our services to large and powerful cloud services such as Google Cloud, which are rapidly increasing the processing power and sophistication available to these algorithms.
Any accounting software that doesn’t move forward into this automated future will be at a significant disadvantage with customers that care about reclaiming their time to work business rather than personal or professional finance. The likes of Sage’s old desktop accounting software, for example, will be challenged for not living online – being unable to draw helpful connections between its customers’ anonymised accounting data.
So, business owners should be salivating at the prospect. We stand on the threshold of being able to reduce those submission errors from 49% to 5%, or perhaps even 0%, saving customers large amounts in the process.
In modern-day banking, transactions are still a laborious process—sending money across the globe involves time, effort and risk. Payments moving across borders are slow, as they typically hop from one correspondent bank to another, each sitting on the funds, ccollecting afloat for who-knows-how-long.
Once there is KYC blockchain, the technology will be at the forefront of helping to identify those who present a greater risk of criminality, argues David Poltorak, chief technology officer at Fortytwo Data.
The UK’s Prompt Payment Code will have a significant impact on the relationship between large businesses and their suppliers. What does the Code mean for your business? And how can you navigate this change effectively?
The recent NotPetya cyberattack underlined the need for organisations to address their exposure and how to mitigate the risk.