Automation killed the accountant

Automation has generated disruptive changes to employment during recent years, and the trend is that its impact will be even deeper in the near future. However, the timing and repercussion of automation varies depending on the geography and industry.

Accounting is a very popular occupation, employs a large number of professionals worldwide. At the same time, it’s on the top of the lists of global declining roles because has a high percentage of repetitive tasks (very easy to automate) and many of its inputs are digital (used to be manual). Also, significant investment has gone into the sector, especially by large firms, to reduce costs by optimizing accounting operations.

Impact felt already

In most of the occupations, the potential impact of automation is quite uncertain. Usually, there are a number of tasks in each role that can be automated, but it’s hard to conclude when and at which extent technology will actually replace workers.

In accounting, this is a process that has already started a few years back. A large number of highly time-consuming tasks have been automated and are no longer performed by employees.

According to The Hackett Group, since 2004 the median number of full-time employees in the finance department at large companies have seen a 40% decline.

Robotic Process Automation

Robotic Process Automation (RPA) is the most likely technology to change accountancy tasks in the short term.

RPA can be applied to the automation of transactional rule-based tasks where structured data and clear pre-defined rules and parameters are used. Usually are designed to require low investment and are easy to integrate.

Blue Prism, a pioneer robotic process automation software, has developed an RPA tool to create invoices automatically. To build the bot, Blue Prism analysed the different steps required to create, and invoice via SAP and replicated it with RPA:

1.- Open SAP and log in

2.- Open e-mail and scan attached excel template

3.- Perform checks to determine the quality of the data from the excel template

4.- The software robot runs the transaction code for the creation of the invoices in SAP and the invoice data from the excel template is copied into the relevant fields in SAP

5.- Send request via email to invoice requestor to adjust specific invoice data

6.- Create invoicing document and correctly save for sending

7.- Send email of confirmation and close all the applications.

This is a very simple example of how RPA works, but this same logic can be deployed in thousands of ‘robots’, all coordinated, and each programmed to automate a different task.

According to Gartner, by 2020, RPA market will top £800 million and 40% of large enterprises will have adopted an RPA tool (vs. 10% from today).

Unfortunately, education and training programmes are not preparing professionals for this new reality. Experienced accountants will still be needed for advising and consulting roles, however, low skilled and entry-level professionals will struggle, because the competition will be huge for a limited number of not well-paid spots.

Optical Character Recognition

Optical Character Recognition (OCR) technology converts scanned documents, PDFs and photos, into data. The technology is not new, however, in recent years there have been developed many data entry cloud-based applications focused on automating accounting tasks.

OCR has completely changed the Accounts Payable process. Not too long ago, companies used to receive all the invoices from suppliers by mail, have a bookkeeper to post each invoice into the accounting system and then archive them in an organized way, easy to find in case you need to look at it again. Recently, tools like AutoEntry, DataMolino or ReceiptBank have completely automated all the accounts payable process. The invoice can be received in any format (scanned documents, PDFs, photos), then the tool extracts the important data from the invoices automatically (invoice date, supplier, amount, tax) and archives a copy of the invoice in the cloud. All synchronized with accounting systems, such as Xero, Sage or QuickBooks.

Computer Assisted Audit Tools and Techniques (CAATT)

CAATT is using technology and analytics to automate audit processes. CAATT refers to any software tool, from excel formulas to data analytics software such as ACL or IDEA.

During their works, auditors would randomly select a statistically valid sample and analyse it. If there is a problem with any of the samples selected, auditors might ask for more information, otherwise they would assume the rest of the data is correct or that the possible issues are not material.

CAATTs allows auditors to analyse large volumes of data in seconds and perform a complete review of all the transactions of the company for the selected period of time.

Other examples of tasks automated

Not long ago, banks used to send their statements on a monthly basis and companies needed a data entry to record all the data into the system. Nowadays, most of the banks have the information online and the accounting systems are able to synchronize using encrypted logins to pull the transactions information from the bank account. Also, are able to compare the transactions and automatically reconcile the movements.

Unleashed is a real-time inventory control and reporting software tool. It’s easy to integrate with most of the accounting systems. It gives real-time stock visibility, accurately calculates the costs of good to provide a better understanding of production costs and margins.

There are tools to completely automate the expense management process like Expensify, Xpenditure or Concur. In this same area, Spendesk is a service that combines prepaid credit cards with an expense report solution.

There are apps like HubDoc and PandaDoc to automate all the management (generation, approval, tracking, eSign, storage) of documentation like quotes, proposals, purchase orders, contracts and invoices. Also sends reminders and tracks calls and tasks.

What are the ‘big four’ doing?

PwC has partnered with (a company that brings AI to enterprises) to build, which is the first module of PwC’s algorithms examine every recorded transaction, user, account, etc. (basically a huge amount of data) in seconds to detect anomalies in the general ledger, without bias or variability. The bot improves its performance the more it’s used.

Two cognitive solutions have been so far the outcome of a partnership between KPMG and IBM. One is a contract abstraction tool that simplifies the management and controls of lease contracts for compliance with IFRS 16. The second solution is a research tax credit service which has the ability to review relevant documents to collect evidence of R&D projects eligible for credit subsidies. For both solutions was used IBM Watson artificial intelligence technology.

Deloitte has developed a long list of AI projects in tax, audit, consulting, risk and financial advisory services. They cover areas like transfer pricing, database analysis, machine learning to assess risk for insurance policies, payment behaviour and fraud.

EY plans to invest $1 billion in ‘new technologies, client services, innovation and the EY ecosystem’ over the next 2 years. As part of this plan, they will create a team of 2,000 people to expand its digital solution services across blockchain, AI, automation and tax technologies.


Blockchain has the potential to eliminate the need for an audit and drastically reduce bookkeeping functions.

Blockchain groups the transactions into blocks. The blocks are linked and ordered chronologically. Each block has a hash of the previous block, a timestamp and the mentioned transactions. Once recorded, the blocks cannot be modified retroactively without changing the subsequent blocks. This guarantees that the underlying transactions are valid, real, accurate and complete.

At some point, fully automated audits will become a reality and auditors instead of analysing and reviewing the data of the company, will verify that the algorithms and code has been correctly used.

Is it possible to completely automate the accounting processes with today’s technology?

Despite all the examples listed above of accounting automation tools, companies need human input to manage the accounting functions. Companies still need an accountant to review the data, take care of complex accounting situations or deal with tax authorities.

However, it’s clear that the need of accounting professionals will continue decreasing over time.

In the report The Future of Work: Jobs and Skills in 2030, they say ‘there is a possibility that in a near future, accountancy services will be entirely automated and a typical role in demand in such firms is ‘meta-accountability’, which is a human audit and verification process to check the machine algorithms have been applied properly.’

Unfortunately, education and training programmes are not preparing professionals for this new reality. Experienced accountants will still be needed for advising and consulting roles, however, low skilled and entry-level professionals will struggle, because the competition will be huge for a limited number of not well-paid spots.

read original article at——artificial_intelligence-5