Accountants

How will automation affect the accounting industry?

Low Value Automation

The most recent Practice of Now research from Sage showed that nearly half of all accounting professionals (45%) would like to automate repetitive, time-consuming accounting tasks, such as data entry and number-crunching.

20% of respondents said they were already investing in automation, while 43% said they plan to in the next three years.

Make no mistake. Just like desktop computing in the 1980s and 90s, and the Internet in the 2000s, we’re entering a time when automation is impacting the industry – whether we like it or not.

But this raises questions we aim to answer in this blog. And we start with the simplest question of all: Why automate?

Why automate accounting?

Automation for an accountant is all about value. It offers accounting and bookkeeping practices the opportunity to focus on delivering more for clients, with less expense.

Why should any staff member spend time collating, organizing, and processing large amounts of data when automation can do it, and in an efficient way that will help you be more successful, achieve your ambitions and perform at your best?

There’s a handful of specific benefits to automation that need calling out.

  • Structured, clean data: Data outputted by automation tends to be of higher quality than human-keyed output. Automation doesn’t suffer from fat finger syndrome! This accuracy allows you to understand and do more. Automated data is well-structured, too, so can also cleanly flow between systems, meaning that automation can be explored and adapted to the very end of a process, not just in part.
  • Data at the heart of client conversations: Traditional client accounting based on previous periods moves to the background of what an accountant offers. Instead, automation delivers real-time data, allowing for pre-emptive problem solving. This opens-up conversations around tax planning, as one example. Reporting tools can provide live impressions on the most pressing and relevant insights and feed in directly to management decision-making.
  • Data security: The world’s awash with data protection regulations – from the NY SHIELD Act, to the California Consumer Privacy Act (CCPA). Data flows within automation processes provide reassurance around the quality of your procedures, ensuring an additional level of client-side protection against fraud or data breaches.
  • Better workflows: This is true even if the client doesn’t want to change! Automation often allows you to adapt your workflows to benefit those clients that do not want to adapt or learn new processes. This could mean having better ways of handling clients who still rely on you to work through their (disorganized) paper records, or removing administrative tasks from payroll clients like the distribution of pay checks and handling of basic queries.

How automation is being used today

There are three key areas where automation is already being used today within practices. This indicates the path that automation will take in future too.

  • Automatic data creation: This is the lowest-hanging fruit and is perhaps the most obvious too. Issuing invoices from within an accounting application means the data is already there, with no need to input it manually. Similarly, issuing a purchase order within the application allows seamless tracking of expenditure. The data can then seamlessly continue a journey, such as being reconciled against bank statements later, or being used for audit purposes.
  • Document fetching: Next, there’s the data that is digital but that isn’t already in your system. Examples might include supplier PDFs, such as those from utility companies. Automation here means the data is reliably recognized, coded, and published directly to the ledger. This kind of data might arrive by email or some other electronic means.
  • Capturing data: The next step is to capture data from sources that are print based, in order to get it into your systems. This might be data from printed receipts you or your staff are handed (e.g. from stores), or from printed invoices or purchase orders you’re sent. Paper continues to make the world go around, even if this is the 21st century!

Through use of the above, automation is proven to drive down the time taken on manual tasks, and in turn the whole cycle of work. It reduces processes that might have taken two weeks to two days, and from two days to two hours. This is a significant resource saving, but automation also speeds up the ability to then process that data (management accounts, tax reporting, year-end processes, and payroll).

The focus therefore does not have to be restricted to a single part of the process, but can be opened up to thinking about streamlining end-to-end workflows across different departments or services within the practice.

Let’s take a look at employee expense management, as one example. In an automated world, employees capture receipts on their cellphone using its built-in camera. Managers can then approve the expenses on similar mobile devices, and the data continues its journey to the accountant for processing. Because the data is already in the system, reconciliation is also automated.

That’s all there is to it.

Consider the manual process, without automation.

The employee has to keep their receipt upon purchasing something (a difficult task in itself!). They then hand it to their manager, complete with a form providing details. Once it’s cleared the manager’s approval, the receipt and form are then sent onto the account where, with maybe 100s more, they have to be manually inputted into the system. Then this data has to be manually sorted, split into document types, manually inputted into one or more systems, and reconciled against bank statements. All of this is labor intensive and, in our modern world, entirely unnecessary.

The future of automation in accounting

To predict future applications of automation in accounting, you can look at your existing workflows to see if three indicators are present:

  1. Processes or workflows are resource intensive, which is to say, they involve significant amounts of relatively unskilled human-only effort (e.g. interpreting paperwork). Imagine what employees could do if free from these tasks – and imagine the caliber of individuals you could recruit if this drudgery wasn’t an expectation for those new to the profession.
  2. There’s substantial amounts of data that, in a perfect world, needs to live inside a computer system, rather than be scrambled across many systems or pieces of paper. This might be because there’s a need to create a more easily-followed audit trail, for example.
  3. There’s a need for data security that isn’t being addressed. This is becoming increasingly the case as government regulations force businesses to take care handling data for individuals. Payroll is a perfect example, and an area with a serious need for automation.

Machine learning and AI in accounting

At the present time, the automation buzzword is machine learning. But what does that mean for accountants?

Machine learning is a specific technique that allows computer software to learn both from experiences and data, and this self-learning is what underlines the new era of AI.

Machine learning is built on advances in cloud computing that have provided the huge data samples that machine learning typically needs, as well as reduced the cost of processing power and increased processing capacity.

By teaching machines to learn for themselves and to do the mundane tasks, we can enable ourselves to have time to focus on what we humans are best at – creativity, strategy and interacting with other people.

A good example of machine learning in accounting is automatic bank reconciliation and ledger coding. Here’s a situation where tasks are easy for most humans yet also very mundane – and disliked for this reason. Sure, computers since the early days have been able to match numbers – a $33.93 client receipt vs a $33.93 debit from an account is an obvious match… Right? But these old school computers don’t know what ledger or code to use, or even whether that particular $33.93 entry is the one that should be consolidated, compared to another $33.93 debit that occurred a week later.

This is where machine learning provides a solution. It learns from watching humans process data. It sees the inherent patterns in bank reconciliations and uses a balance of probabilities to autocomplete human tasks. Often, this feels spooky to humans, because we’re so used to computers being mechanically predictable. When machine learning is effective, computers seem to have gained a sixth sense.

Machine learning gets better the more data it has access to, and this is why the cloud is so invaluable – it’s a huge library of data that machine learning can consume.

Conclusion: How to start automating

What other tasks can you see within your own work or that of your colleagues that machine learning might one-day enhance? Because you can bet that, in the depths of the programming labs of all software companies, somebody is already working on it. The future is now.

This means that accounting professionals have a duty to keep an eye on the technological horizon. Those that embrace these machine learning advancements have a huge competitive advantage, in terms of cost savings and client satisfaction, over those that don’t.