While digitalisation of the economy and investments in intelligent solutions have been active in recent times, the COVID-19 pandemic has accelerated this trend, which is growing across the globe. A Euromonitor International survey completed in April 2021 showed that 30% of companies globally plan to increase spending on automation tools in the next five years. For many functions within an organisation, this will involve having to adapt quickly to stay relevant and competitive.

The Finance function is no different: accountancy and finance professionals must transform themselves in response to these trends. It is with these reasons in mind that the Institute of Singapore Chartered Accountants (ISCA) embarked on a research project titled “The State of Play of Intelligent Automation in the Finance Function”. The project sought to discover the risks, challenges and benefits for an organisation in implementing intelligent automation (IA) solutions; provide insights into the prevalence of IA adoption; uncover the expectations of the C-suite before, during and after implementing IA solutions, and provide recommendations for organisations which are considering adopting IA solutions.

This study defines IA as an enhanced form of automation that combines elements of robotic process automation (RPA) and artificial intelligence (AI). Data for the study were collected from two sources: (a) an online survey of C-suite (comprising Chief Executive Officers, Chief Financial Officers (CFOs), Chief Technology Officers/Heads of Information Technology (IT) and Finance Directors, and working-level accountancy and finance professionals (referred to as “Working Level” in the article, and defined as Financial Controllers, Financial Managers, Accounts Managers, Finance Executives and Accounts Executives), and (b) indepth interviews with consultants working primarily in digital transformation, and experts in IA/AI and CFOs.

Done in collaboration with AI Singapore and National University of Singapore (NUS) Business School, the final report was launched at the Institute’s flagship annual Professional Accountants in Business Conference 2021 on August 19.


One key finding from the study was a perception gap about the efficiency of IA solutions between the C-suite and Working Level in companies that have implemented IA (also known as “IA adopters”): 61.0% of the C-suite were of the belief that IA assumed 26–50% of work, but 69.2% of the Working Level said that these solutions took over just 1–25% of work (Figure 1).

Figure 1

Cyert and March’s (2006) work suggests that this perception gap may be due to how the C-suite tends to be focused on strategic work while Working Level staff are focused on operational issues. To bridge this gap, our interviewees recommended that the C-suite cultivate an environment where employees can provide honest feedback. In addition to offering a realistic picture of what is happening at the operational level, the feedback can also be used to improve the IA solution.

This open and honest communication between the C-suite and Working Level creates a virtuous loop where the IA solution can be continuously improved and employees develop a deeper appreciation of the process and confidence in the IA solution.


According to the C-suite, the top three areas within the Finance function where IA solutions were used were Accounts Payable (61.0%), Accounts Receivable (56.6%) and Budgeting/Financial Planning and Analysis (50.7%) (Figure 2).

Figure 2

Accounts Payable and Accounts Receivable are areas that have well-established routines that are frequently executed, making them excellent candidates for automation. Budgeting/Financial Planning and Analysis is another area where IA solutions could be implemented: algorithms can process large, multidimensional data sets and start to generate insights once the accepted financial models have been put in place.

In contrast, Risk Management, Tax and Treasury are the three least popular areas for implementing IA, with only 23.5%, 19.9% and 14.0% of C-suite respondents saying that they had done so, respectively. Compared to Accounts Payable and Accounts Receivable, these three areas may be used less frequently and whose processes are less routine. As such, the “trickle effect” may be such that these will be picked for automation only after more experience is gained from automating other functions. Nonetheless, these three areas remain candidates for automation.

One potential concern of IA implementation is financial cost. However, our study showed that 82.0% of organisations kept within their expected budgets for IA implementation. As can be seen in Figure 3, 64.2% of respondents spent between $50,001 and $200,000 on their IA solution (25.2% paid $50,001–$75,000, 20.3% paid $75,001–$100,000 and 18.7% paid $100,001–$200,000). Across our respondents, the average final cost was just over $98,200.

Figure 3


Efficiency and Compliance (defined as increasing the efficiency of Finance staff/procedures, raising levels of compliance, and lowering rates of errors, etc) was rated the most important consideration by 45.5% of C-suite respondents who have IA in their Finance function. This was also reflected in how the IA solution ultimately impacted Finance employees: a significant 74.8% of C-suite respondents said that the IA solution augmented their Finance employees’ capabilities in the same role for greater efficiency and productivity. Similarly, 62.6% of the C-suite retrained their Finance employees to work in a different department at the same level of seniority (Figure 4).

Figure 4

The findings validated anecdotal impressions of the impact of intelligent solutions on Working Level staff. However, it should be noted that only 44.7% of C-suite respondents modified key performance indicators for staff within the Finance function to reflect greater efficiency. This may mean that the full return on investment in the IA solution may not be truly captured.


Our study showed that 97.4% of respondents adopted the IA solution in the Finance function as part of their digitalisation plan. Of these respondents, 89.4% and 73.2% have adopted either AI or RPA (Figure 5).

Figure 5

Of note, 25.6% of respondents who adopted either RPA or AI said that they implemented an IA solution before going on to implement RPA or AI, while 36.3% said that they opted for a simultaneous implementation of the intelligent solutions (Figure 6).

Figure 6

This trend may indicate that organisations, after having adopted either IA, AI or RPA, will eventually go on to adopt more automated solutions. These technologies themselves act as precursors to adoption of other more specialised digital technologies such as Computer Vision (65.0%), Speech Recognition (61.8%), Natural Language Processing (61.0%) and Text Mining (48.8%) (Figure 7).

Figure 7

Thus, the adoption of an IA solution may serve as a relatively simple entry point and threshold for organisations seeking to kickstart their digitalisation journeys.


Nonetheless, challenges to adopting IA solutions remain: 55.3% of respondents said that their organisations have no plans to adopt IA in the Finance function, and a mere 17.1% said they had plans, but had yet to execute them. This group of IA non-adopters felt that Efficiency and Compliance (66.7%) was the greatest challenge they faced. This aligns with the main reason that organisations elected for an IA solution.

However, General Competitiveness (51.4%) and Integration (49.5%) (defined as an organisation’s general competitiveness, and training staff/adapting processes to use IA solution time/effort required to incorporate into current systems etc, respectively) are the next two concerns for the IA non-adopters (Figure 8). Taken together, these three challenges may suggest that IA non-adopters may be unsure about the benefits of automated solutions in general.

Figure 8

A significant portion of IA adopters (28.5%) and IA non-adopters (44.8%) believed that Financial Cost was an important factor and a key challenge, respectively (Figure 8). Closer examination of the data showed that 74.2% of IA non-adopters overestimated the cost of adopting an IA solution in the first four estimated budget bands (from less than $10,000 to $75,000), compared to the actual expenditure of IA adopters (Figure 9). This hints at another aspect of hesitance towards adoption of IA solutions: without a true understanding and estimation of the actual financial costs, it would be difficult to justify taking any concrete steps towards adoption.

Figure 9

A third concern that IA non-adopters may have was the impact of the IA solution on their Finance employees. Figure 10 shows that 51.4% of IA non-adopters would train their Finance employees to work with the IA solution in the same role, and only 36.2% of them said that Finance employees would be retrained to work higher up the value chain. A similar portion of respondents (35.2%) said that they would let natural attrition take its course once the IA solution was implemented – an unwelcome side effect of digitalisation by any measure.

Figure 10

Taken in totality, the trifecta of doubts surrounding the benefits of automation, an overestimation of the financial costs, and uncertainty about the impact of IA on their Finance employees made up the main challenges and concerns among IA non-adopters.


To help organisations begin their digitalisation journey with IA solutions, the project included eight recommendations, abbreviated here from the final report.

1) Start small.

Do not overreach, and start by automating simple processes. It is important to take baby steps to acquire experience and, more critically, confidence, when adopting IA. If early projects do not demonstrate any return on investment, or if users do not see the benefits, the project is likely to be unsuccessful.

2) Leadership from the top is critical.

The role of the CFO, and in general the C-suite, is to set the direction and provide the input for how the existing processes and Finance talent should change and evolve with the technology being contemplated. The top management also decides on the pace of automation. Therefore, any IA adoption must receive the endorsement and support from the very top.

3) Buy-in from all levels is essential.

Once endorsement from the top is achieved, the C-suite needs to convince the remaining levels to come onboard the IA adoption. Since the Finance function employees are the end users of the intelligent solution, they are the people intimately aware of what the process entails. With their hands-on knowledge, they can be tapped on to provide a basic blueprint of the process as well as the list of requirements which the IA solution should fulfil. At the same time, the IT department should be involved as well. They will need to understand the expected level of IT involvement should changes or tweaks need to be made regarding how the solution has been set up.

4) Manage the transition to automation carefully.

Automation has sometimes been perceived as a threat to job security. The C-suite will need to manage perceptions and allay fears when transitioning to automation, such as by offering reskilling and/or upskilling opportunities as well as relevant resources to employees to address their concerns.

5) Communication must be a two-way channel, and taken seriously.

Automation can be both a boon and a threat. Communication in both directions is key to ensure that everyone’s voice is heard at all stages of the IA adoption journey.

a) The C-suite has to foster an environment where the Working Level staff are able and encouraged to provide honest and constructive feedback about the solution, updated processes, outputs and whether the pain points have been resolved.

b) To garner buy-in, the company needs to continuously emphasise, explain and promote the strategic vision and objectives of the solution to all levels after implementation. This is especially beneficial for staff who are deeply involved operationally, to be able to occasionally take a step back and assimilate the strategic vision of the company.

6) Consider the opinions of Working Level staff, particularly during process improvement phases.

As Working Level employees are familiar with the processes, their feedback about operational concerns are often critical sources of information which can be used to improve the IA solution. Recalibration and retraining of the solution are essential in ensuring that its performance is maintained, to keep up with changes in data quality.

7) Ensure that expectations across all levels are realistic and specific.

The following (non-exhaustive) list shows some of the common expectations when considering IA adoption:

a) Investment cost should commensurate with expected return/value.

b) The solution can address a specific pain point/problem that it is intended to address.

c) The extent and type of post-implementation support offered to users should be comprehensive (for example, training, troubleshooting, maintenance, etc).

d) The solution is easy to implement and integrate into existing systems.

e) The experience and capabilities of implementation consultants are appropriate and specific.

Expectations may vary across different strategic and operational levels. To minimise the gaps in perception, the C-suite should also set realistic expectations that are informed by the Working Level.

8) Seek to continuously improve on the process and the solution.

Implementing an IA solution is not an end in itself. Companies should adopt an IA solution with the understanding that it is a continuous and ongoing process that needs constant improvements, and that the solution will grow with the organisation.


The benefits of automation are well-documented but it requires a concerted effort across the entire organisation. For organisations which are keen or even just curious about IA, it would be best to first identify a process that is well-established and frequently used, and explore how an IA solution can help relieve some or all of the tedium associated with the process. Open communication channels are essential not only for ensuring that the IA solution can integrate with current systems and processes but also for continuous improvement and maintenance of the IA solution. Organisations would do well to keep in mind that automation is just a first step towards long-term sustainability of its operations.

The full report is available online

This article was written by ISCA, AI Singapore and NUS Business School, National University of Singapore.