As artificial intelligence becomes more advanced, fear of automation has left people feeling nervous about their future careers, and whether robots will replace them. Grant Thornton’s Mark O’Sullivan and Kabir Dhawan explain how the profession can best prepare for change.

Automation in all its forms has been absorbed by businesses to such an extent that tech experts and even governments are starting to ask questions about it. For example, how might society cope with an excess of workers whose jobs have become obsolete? What happens when this leads to a potentially reduced tax take?

Bill Gates was among the first to suggest that a tax for robots could work. One thing these discussions tell us is that, despite the concerns, people are starting to accept automation is a reality. Those in organisations not already preparing for a more automated future could find themselves at a disadvantage.

How is automation affecting the profession?

KABIR DHAWAN: Automation, as it is traditionally understood, is the automation of transactional processes. Most finance packages, for example, allow for workflow-based, hands-free invoice processing, where people get involved only when things go wrong. Businesses that don’t do this today are below what we think the median position should be. New automation is an emerging area, dominated by the field known as robotic process automation (RPA). The IT Faculty’s recent publication in the Tech Essentials series, The essential guide to Robotic Process Automation, goes in-depth, but RPA spans two broad areas: automation and artificial intelligence (AI).

The automation element is largely concerned with replicating human behaviour, often through complex rules-based and learning mechanisms. There are a number of small providers of software who do this brilliantly. We partner with a couple that are disrupting the legal services industry where they have software that replicates the contract review work that a paralegal might undertake.

On the AI side, Amazon offers algorithms via Amazon Web Services (AWS) which can learn and predict off multiple time series simultaneously (see tinyurl.com/CH-AMS18). These are not yet embedded in business tools commonly, but when something becomes available as an AWS service it’s a strong indicator that demand for a given service exists, and we can expect to see more of it in the future.

RPA capabilities have been hyped up by parties with vested interests, and there is not a defined set of best practices yet. There are no clear leaders in the field, but several vendors are seriously focused on it. A lot of the use cases still need to be defined, or invented, by finance teams (and the consultants working with them).

The Grant Thornton blog Three steps to prepare your finance team for AI talks about people at the start of their career. What do junior staff need to do to develop their skills?

MARK O’SULLIVAN: Junior finance staff will undoubtedly require additional core skills that previous generations have had the time to develop. Alongside the core accounting skills there will be a far greater emphasis on demonstrating commercial skills around sophisticated analytics and critical thinking.

Greater technology understanding is clearly also integral – being able to knit together a query to interrogate a data warehouse will not be a mythical dark art. These skills are already quickly becoming foundation level competencies for junior finance team members.

The rate at which skills develop will not fundamentally change, however the emphasis on the skills that are prioritised by employers will. The ability for junior accountants to be able to distill insights and trends from data sets will be what marks individuals out – rather than the speed with which they can crunch through data sets.

There is an element of risk for junior staff in that they are not developing the skills that will be needed in future roles. However, the onus lies heavily on employers and educational institutions to ensure that the skills that they teach are in line with the requirements needed from future finance leaders.

What implications do you think a reduced pipeline of talent might have, especially for smaller practices?

KD: Recruitment decisions are always important but, given the number of people in finance will likely reduce as a function of automation over time, an incorrect hire has a larger impact. Smaller teams in smaller organisations are already affected in this way today.

Practices currently use highly intelligent and capable graduates to do intensely manual activities such as audit filing, calling accounts, etc. AI releases some of that time to do more value-creating activities, harnessing the more innovative ideas and capabilities that millennials entering the workplace possess. The impact of this should create a more enriched work experience.

Freeing up junior team members in this way should allow smaller accounting firms to compete in more ways with other firms, rather than just on the audit and tax side (due to automation driving labour efficiency in these spaces). This will allow junior team members to develop advisory skills earlier on. We see there being greater opportunity for finance professionals to go out into the business and act as business partners to their internal stakeholders.

What role can and should education be playing in preparing the next generation of finance professionals for work?

MO: In the UK, the introduction of the apprenticeship levy presents new opportunities for schools and universities to work with the profession to develop courses aligned to the development of some foundational skills that employers will be looking for.

The challenge around school and university programmes is that they often focus heavily on theory rather than application. Commercial interrogation, building analytical skills, understanding ways to represent data and understanding how finance fits into an organisational value chain are all examples of where practical work-life skills can be covered.

Education could and should take a greater role in equipping accountants with skills they are less likely to develop in practice, such as attention to detail, the ability to identify the keystone of the data presented and the ability to filter through a large volume of data points to extract the ones with the most meaning.

How will business/finance absorb new entrants to the profession?

KD: When considering automation, accountants need to be equipped so that they are not “checking the machine”, but understanding “what it is doing and why”.

This will remain critical even as automation starts changing roles in finance functions. Students emerging from accounting qualifications need to have the commercial awareness to understand how things connect to each other. For example, when we close the books at the end of each month and generate the management reports, how is the data we produce used to guide business decisions? Underpinning principles of business finance, audit and tax still need to be understood, but the accountancy bodies have a role to play in preparing people for a digitally enabled environment. A greater understanding of the operating model of a business is critical – eg, if you’re auditing a process that has gone through RPA, where is the segregation of duties? What’s the control environment in a process managed by a machine? How do you assure it?

The next generation of auditors who come through accountancy training will likely be the first ones to experience RPA and AI as a core component of finance functions. Audit is a progressive industry (in terms of responsibility increasing as people move upwards through hierarchies), where they become the leaders of audits. This means that there will be a point where this next generation is the first to grapple with the question of how to audit a process that is machine driven. Partners and senior leaders in their firms will start looking to them for practical experience in dealing with these challenges, which is a huge opportunity for more junior staff.

Are there any good starting points for automation in practice?

MO: Most organisations making successful use of automation have started simple. This means identifying a sample of finance sub-processes that currently require intense manual effort, happen outside of standard system workflows and are highly repeatable. It doesn’t take long to build up a clear picture of how much effort is going into these sub-processes and what the potential efficiency gain could be through the use of automation.

The more technical part is then through design and implementation of the solution. There are some significant considerations beyond simply proving that there is a business case. Implementing automation creates a digital workforce that will require organisations to reconsider their finance operating model.

There is also an opportunity for finance teams to be more influential in guiding the external market to where automation solutions might be most beneficial. At the moment automation is being driven by the providers, as opposed to the individuals who are benefiting most from its application.

Are AI and automation still farther away for some organisations than others?

KB: Small businesses will not be able to leverage economies of scale and lots of automation and AI work is focused on replacing scale economy opportunities via process efficiency.

Price will not be the differentiator, but rather the level of maturity of the solution. An automation solution for a FTSE 100 firm may be far too complex to be leveraged or utilised by a £30m organisation looking to have a more effective billing process.

It doesn’t take away from the fundamental need to constantly analyse the finance operation model and constantly ask if your organisation is as optimised as it could be across the different value drivers of finance, given the resources at your disposal. It’s not looking to companies in a different strata and what they do, it’s about extracting the marginal gains in what they do day to day.

• Grant Thornton’s guidance on steps for preparing a finance team for AI