Gartner for Finance

Digital Technology
Use Cases in Finance
Overview
Finance leaders have ambitious digital goals but an outdated view of what’s possible
with digital technology. This research underlines six critical digital capabilities and their
sample use cases to help finance leaders reset their understanding of digital potential
early in their digital journeys.

Key Findings
• Finance functions chiefly focus on applying digital technology to reduce manual
work. The 2020 Gartner Future of Finance Survey found 60% of finance functions
are implementing or have implemented RPA-driven process automation and routine
business decision automation.
• The survey also revealed only 36% of finance functions are implementing or have
implemented any digital capabilities to enhance judgment-based work.
• To meet its goals, finance must develop six digital capabilities: robotic process
automation (RPA)-driven finance process automation, routine business decision
automation, intelligent (hyper)finance process automation, business decision
augmentation, artificial intelligence (AI)-enabled analysis and decision support,
and AI-enabled nudges.
• All six digital capabilities have potential applications within finance. Even digital
capabilities utilizing forms of AI (such as machine learning [ML] and natural language
processing [NLP]) are less prohibitive for finance functions than they first appear.

Recommendations
To start expanding their use of digital technology in accordance with their functional
strategy, finance leaders should:
• Educate all finance subfunctions’ leaders on the entire range of potential
digital technology capabilities within finance, i.e., those reducing manual work
as well as those enhancing judgment.
• Discuss all six digital capabilities when setting finance’s multiyear digital
technology roadmap, rather than centering planning just on digital capabilities
most familiar to finance.
• Use existing use cases as a guide to define the characteristics of a good-fit
finance use case for each of the six digital capabilities.

2 © 2021 Gartner, Inc. and/or its affiliates. All rights reserved. CM_GBS_1584588
Introduction
Eighty percent of senior finance leaders agree that finance must significantly accelerate its
application of advanced digital technologies (RPA and AI) to effectively serve the business
in 2025.1 However, the rapid pace of digital technology evolution can make investing in
digital daunting. According to the 2020 Gartner Future of Finance Survey, only 37% of
finance functions have a clearly defined digital technology investment strategy for the
next two to three years.2

Robotic process automation (RPA) Artificial intelligence (AI) is an umbrella

refers to a set of software that uses “if/ term for a series of technologies that
then” statements to automate repetitive, apply advanced analytics and logic-
rule-based, human tasks to improve based techniques to interpret events,
speed, efficiency and accuracy. support and automate decisions, and
take actions.

Finance leaders can’t accelerate investment in digital technology without first understanding
their investment options; they need to have an up-to-date understanding of what’s possible.
The following research details six critical digital capabilities, as well as their finance-specific
use cases, to help finance leaders develop a more specific understanding of their digital
investment options.

3 © 2021 Gartner, Inc. and/or its affiliates. All rights reserved. CM_GBS_1584588
Analysis
Develop Six Digital Capabilities to Achieve
Your Digital Ambitions
We have identified six digital capabilities finance functions must develop to achieve their
ambitious digital goals:

01 RPA-driven 02 Routine 03 Intelligent (hyper)-

finance process business decision finance process
automation automation automation

04 Business decision 05 AI-enabled analysis 06 AI-enabled

augmentation and decision nudges
support

While all these digital capabilities have been adopted by the most digitally advanced
finance functions, finance functions’ overall implementation progress to date is limited.
So far, RPA-driven finance process automation and routine business decision automation—
two capabilities that help reduce employees’ manual work — have been the focus of
finance’s digital investment. Deployments of more advanced capabilities intended to
enhance finance and business judgment, such as business decision augmentation and
AI-enabled analysis and decision support, are still few and far between (see Figure 1).

Figure 1. Current Levels of Digital Implementation Within Finance

Not Currently Evaluating Piloting Implementing or

Pursuing Already Implemented

RPA-Driven Finance Process Automation 13% 21% 21% 45%

Routine Business Decision Automation 43% 26% 13% 15%

Intelligent (Hyper)Finance Process Automation 43% 39% 10% 8%

Business Decision Augmentation 41% 32% 14% 14%

AI-Enabled Analysis and Decision Support 49% 27% 15% 9%

AI-Enabled Nudges 58% 27% 10% 5%

0% 50% 100%

n = 137 senior finance leaders

Source: 2020 Gartner Future of Finance Survey
Note: Percentages may not add up to 100% due to rounding.

4 © 2021 Gartner, Inc. and/or its affiliates. All rights reserved. CM_GBS_1584588
01
RPA-Driven Finance Process Automation

What it is Implementation progress

RPA is used to automate routine, Forty-five percent of finance functions
highly manual tasks across both are implementing or have already
transactional and judgment-based implemented RPA-driven process
finance processes. automation, while 21% are piloting it,
21% are evaluating it and 13% are not
currently pursuing it (see Figure 1).

Overview
Using RPA to automate certain manual tasks can substantially reduce the time spent
by finance staff on highly manual work. RPA is best suited for automating five types
of processes: reporting, alerting, migration, validation and calculation. In each case,
a software bot is programmed to follow instructions to complete highly repetitive tasks
at a much higher speed and error-free rate. Automating such processes can also improve
control, as all RPA bots’ actions can be tracked and reviewed for compliance.
This application of digital technology is among the most common in finance, especially
as only one or two robots are needed for RPA to start generating a return on investment.
Currently, the most common use cases include:
• Report automation. The bot downloads statements from multiple financial systems
and combines them into a digital package. The bot emails this package according to
user-access restrictions.3
• Budget data prepopulation. The bot logs into a financial system, extracts data,
performs data manipulations and inputs data into a corporate budget that is then
submitted to budget contributors as a starting point.4
• Embedded derivatives accounting. The bot pulls a regional BI report, performs
data manipulations and sends updated data to a corporate team that runs further
transformations. The bot then sends the updated data to the general ledger team.5

5 © 2021 Gartner, Inc. and/or its affiliates. All rights reserved. CM_GBS_1584588
02
Routine Business Decision Automation

What it is Implementation progress

Digital technologies (primarily RPA with Fifteen percent of finance functions
select uses of AI) completely execute are implementing or have already
frequent, rule-based decisions without implemented routine business decision
human intervention. automation, while 15% are piloting it,
26% are evaluating it and 43% are not
currently pursuing it (see Figure 1).

Overview
Many manual, routine business decisions are subject to human bias and subjectivity.
Using digital technologies to perform routine business decisions helps eliminate this
subjectivity. In most cases of routine business decision automation, RPA bots are taught
decision rules to automatically apply a particular action to a given item. As finance’s
customers increasingly expect shorter decision times, routine business decision
automation allows finance to quickly advise on complex decisions by freeing up time
previously spent on relatively simple decisions. Currently, the most common use
cases include:
• Credit decision automation. Establish decision rules to govern the credit decision
process, including perimeters for an automatic denial or an automatic approval based
on the company’s level of acceptable risk. Finance staff members only make credit
decisions for which the automated process is inconclusive.6
• Automated cash allocation. Use an accounts receivable (AR) software to automatically
read, match and record open items against payments received in real time, shortening
the average collection period and decreasing days sales outstanding (DSO).7

6 © 2021 Gartner, Inc. and/or its affiliates. All rights reserved. CM_GBS_1584588
03
Intelligent (Hyper)Finance Process Automation

What it is Implementation progress

Digital technologies automate end-to- Only 8% of finance functions are
end processes that use unstructured implementing or have already
data, are highly variable and require implemented intelligent (hyper) process
judgments RPA cannot manage. automation, while 10% are piloting it,
39% are evaluating it and 43% are not
currently pursuing it (see Figure 1).

Intelligent (hyper)finance process automation uses a combination of automation tools

(typically RPA and ML are combined) to generate predictions to help finance better execute
complex finance processes. ML models are particularly useful for predicting a quantity or a
category and should be applied to business processes with potential to generate significant
cost savings or revenue impact if improved.

Machine learning (ML) is a subset of artificial intelligence that

solves problems using algorithms and statistical models to extract
knowledge from data.

Although some finance leaders may assume ML is prohibitively resource-intensive and

requires advanced computer science skills beyond their finance staff, developing an
ML solution is simpler than it appears. It primarily requires choosing a precise business
prediction, acquiring historical data and using open-source ML tools. The ML model is
trained on historical data to create future-looking predictions.

Case in Point: Simple Machine Learning Pilots (Iron Mountain)

When late customer payments prevented finance from making reliable predictions and
plans based on expected revenue, Iron Mountain used ML to develop a targeted customer
treatment plan. First, it identified an opportunity to improve its A/R process by predicting
the probability of a customer settling an invoice late. Second, it collected customer data
required for an ML training set. Third, it deployed the trained ML model. Finally, it targeted
high-risk, high-value customers for direct intervention. The result was a 40% reduction in
settlement time for accounts within the ML pilot. The average time to settle for high-risk
customers decreased from 68 days to 40 days.8

7 © 2021 Gartner, Inc. and/or its affiliates. All rights reserved. CM_GBS_1584588
04
Business Decision Augmentation

What it is Implementation progress

Digital technologies augment Fourteen percent of finance functions
complex, ambiguous business are implementing or have already
decisions by producing insightful implemented business decision
recommendations humans can augmentation, while 14% are piloting it,
choose from. 32% are evaluating it and 41% are not
currently pursuing it (see Figure 1).

Overview
Like routine decision automation, business decision augmentation also helps reduce
human bias in business decisions. However, it typically uses AI instead of RPA bots to
augment more complex decisions that require significant degrees of judgment. For
example, while RPA bots can provide simple “yes” or “no” answers, AI’s self-teaching
capability provides more tailored answers based on large amounts of data.
Business decision augmentation is particularly useful in situations that demand swift
or large-scale classification and prediction, and is most likely to be helpful when data
is well-defined and of good quality.

Case in Point: Pricing Simulation Tool (Food and Beverage Company)

A brewery used an AI-enabled simulation tool to help decision makers understand
the likely impact of various approaches to pricing in different markets worldwide. First,
the brewery performed an impact assessment to understand the relationship between
price and demand in various markets. Second, the brewery used an AI-enabled
simulation tool employing discrete choice modeling techniques and economic and
demographic indicator data. This enabled decision makers to test out and visualize
the impact of potential pricing strategy on sales, as well as sales lost to competitors.
The brewery discovered opportunities to optimize prices by 16%, resulting in a potential
improvement of $250 million in revenue.9

8 © 2021 Gartner, Inc. and/or its affiliates. All rights reserved. CM_GBS_1584588
05
AI-Enabled Analysis and Decision Support

What it is Implementation progress

Digital technologies automate the Only 9% of finance functions are
generation of customized reports implementing or have already
and analyses. implemented AI-enabled analysis and
decision support, while 15% are piloting
it, 27% are evaluating it and 49% are
not currently pursuing it (see Figure 1).

Overview
AI-enabled analysis and decision support capability enables finance to automatically
generate analysis that is more complex than the reports typically enabled by RPA bots. In
practice, finance functions seeking this capability may need to enlist multiple forms of AI
tools in a multistep process. For example, finance may need a tool with the functionality
to fix inconsistent and unstructured data, identify data trends by correlating and grouping
data, generate suggested corrective actions, or all of the above.
This capability can significantly enhance finance’s self-service offerings, freeing finance
analysts to focus on more hands-on decision support and more customized analysis.

Case in Point: Machine-Learning-Powered Trend Analysis (UBS)

UBS uses ML and NLP to analyze employee expense data and automate corresponding
reporting. First, ML automatically corrects inconsistencies within varied and often unstructured
employee travel receipts. Second, additional ML identifies spending trends within the cleaned
data and automatically pulls them into a visualized presentation for stakeholders. Finally, NLP
generates summary text to describe the nature of the trends. Armed with a clearer and more
immediate understanding of its employees’ travel activities and needs, UBS is able to negotiate
improved booking rates and travel terms with suppliers/agencies globally.10

Natural language processing (NLP) is a subset of AI that involves

the ability to turn text or audio speech into encoded, structured
information, based on an appropriate ontology.

9 © 2021 Gartner, Inc. and/or its affiliates. All rights reserved. CM_GBS_1584588
06
AI-Enabled Nudges

What it is Implementation progress

Digital technologies identify finance Only 5% of finance functions are
or decision maker behaviors that implementing or have already
historically have had a negative financial implemented AI-enabled nudges,
impact, and suggest an alternative while 10% are piloting the capability,
decision with superior outcomes. 27% are evaluating it and 58% are not
currently pursuing it (see Figure 1).

Overview
Originally established as a behavioral economics concept, nudges are subtle signals
designed to steer individuals’ behavior in a predictable way, without limiting choice or
implementing additional economic incentives. For finance functions, nudges represent
a powerful tool to indirectly or subtly encourage employees to make “better” (i.e., more
financially sound) decisions. This acts as an extra layer of quality or compliance control
when finance staff are unable to provide this same level of oversight.
This capability is particularly useful for processes where human judgment cannot be
eliminated but still needs to be improved. For example, AI-enabled nudges can identify
the most high-risk decision points (those more likely to produce errors in human
judgment) and share digital prompts to signal either the risk of continuing with a
particular decision and/or the benefit of pursuing an alternative decision.

Case in Point: Anomalous Transaction Risk Training (Financial Services Company)

A fund used an AI capability to coach its analysts to spend more time assessing risky
anomalous transactions. To begin, the AI scored potential anomalies by risk and potential
cost. It used a time tracker tool to monitor how much time analysts were spending on
anomalies of various levels of risk, which revealed analysts were spending too little time
on the riskiest anomalies. Then, the AI employed an algorithm used in chess-playing
programs to “challenge” the analysts’ assessments of anomalies in real time. If an analyst
made a conclusion on a risky anomaly too quickly, the AI prompted the analyst to take
another look. The fund was able to slow down analysts’ assessments of transaction errors,
enabling the company to catch a greater percentage of high-risk transaction errors by
combining artificial and human intelligence.11

10 © 2021 Gartner, Inc. and/or its affiliates. All rights reserved. CM_GBS_1584588
Take Three Steps to
Get Started
As finance leaders consider how and where to introduce these six digital capabilities
within finance, they should start with the following steps:

1. Consider not just whether targeted finance processes and/or set of activities can
be improved, but also how they can be improved (e.g., automating a manual task,
automating a decision, identifying patterns, generating predictions).

2. Develop a finance digital technology roadmap to guide digital technology

investment prioritization decisions.

3. Consult with the organization’s IT support partners (e.g., finance IT, corporate IT,
chief digital officer, chief information officer) to determine whether a particular
application of digital technology is appropriate, particularly when considering
more advanced digital capabilities that use AI.

Evidence
1
2020 Future of Finance Survey, n = 137 finance leaders
2
2020 Future of Finance Survey, n = 137 finance leaders
3 Cross-Functional RPA Use Cases
4 These Are the Bots You’re Looking For, Strategic Finance.
5 Cross-Functional RPA Use Cases
6 Ten Dimensions of Decisioning.
7 How to Maximize AR Efficiency With Cash Allocation Software, Cashbook.
8 Case Study: Simple Machine Learning Pilots (Iron Mountain)
9 CPG Industry Use Case — Price Optimization Analytics
10
Machine Learning Powers UBS’s Travel Reporting, Business Travel News.
11
Can AI Nudge Us to Make Better Choices?, Harvard Business Review.

The organizations profiled in this research are provided for illustrative purposes only, and do not constitute an
exhaustive list of examples in this field nor an endorsement by Gartner of the organization or its offerings.

11 © 2021 Gartner, Inc. and/or its affiliates. All rights reserved. CM_GBS_1584588
Actionable, objective insight
Explore these additional complimentary resources and tools for
Finance leaders:

Webinar Research
The Gartner 2022 Finance Top Strategic Technology
Digital Plans & Priorities Trends for Finance
Watch our experts review our 2022 Identify nine key technology trends
Finance Priorities survey results. to optimize finance operations.

Watch Now Download Now

Conference Resource
Gartner CFO & Finance Gartner for Finance
Executive Conference 2022 Discover how we can help you to
Join us for this unique event dedicated achieve your critical priorities.
for CFO and finance leaders.

Register Now Learn More

Already a client?
Get access to even more resources in your client portal. Log In

12 © 2021 Gartner, Inc. and/or its affiliates. All rights reserved. CM_GBS_1584588
 Get More.
Get actionable, objective insight to deliver on your most critical
priorities. Our expert guidance and tools enable faster, smarter
decisions and stronger performance. Contact us to become a client:

U.S.: 1 855 811 7593

International: +44 (0) 3330 607 044

Become a Client

Learn more about Gartner for Financial Leaders

gartner.com/en/finance

Stay connected to the latest insights

© 2021 Gartner, Inc. and/or its affiliates. All rights reserved. CM_GBS_1543562

© 2021 Gartner, Inc. and/or its affiliates. All rights reserved. CM_GBS_1584588

13 © 2021 Gartner, Inc. and/or its affiliates. All rights reserved. CM_GBS_1584588