When it comes to finance functions, “democratization of data” is already well underway, but the “democratization of data science” is what has recently started happening from the past few years. Data science is gradually permeating the wider society, including the finance functions of large enterprises.
Finance Functions: From BI to AI
Majority of the finance functions in large enterprises have analysts in their teams, with the basic analytics skills to work on financial data and to get the jobs done. But now, with finance functions moving from BI to AI, the finance leaders are realizing that the skills of data scientists have to become an essential part of their toolset. They desire to be able to apply advanced analytical methods, like AI and machine learning to gain deeper insights that can be applied to build true competitive advantage.
Democratization of data science in finance functions Vs. other domains
Data science has its huge application in different domains, such as marketing, sales, supply chain, etc., where it is often used to solve objective problems such as image classification, task automation, etc. However, finance functions require high interpretability and accuracy. The results of finance data science projects are often used directly by senior management to make strategic decisions that can significantly change the future direction of an organization. These help the businesses find ways to maximize profits, minimize risks, and make new investments.
A few applications of data science and ML in corporate finance
- Advanced ML for better Governance and Control – Closure of accounts is always a time-consuming process. There is lack of systems that could raise red flags on specific transactions; often there is no defined method for variance and to identify anomalies, even in the finance functions of large enterprises. Advanced ML based anomalies and outlier detection can aid in root cause analysis of variation within specific cost centres or business segments to improve governance and control in accounting.
- Analytics on at-risk customers – Within businesses/ large enterprises, there are certain customers that are not paying on time or there is a high number of low value deductions with certain customer groups, leading to working capital blockade. Advanced solutions would help in creating customer overdue profiling which in turn would help in providing actionable inputs to the business to reduce overdue proactively. Such solutions would provide a fair understanding of various components of overdue and its correlation with underlying business dimensions.
- Cash analytics – It is important for the finance teams to understand the key reasons behind the payment delays. Tracking and monitoring customer payment patterns is very critical to the accounting needs. Finance teams often lack visibility on the operational drivers behind the cash flow outcomes. Most of the time there is a lack of integrated data flow that would provide clarity to the finance teams on the end-to-end cash conversion cycle. As a result, proactive cash flow management and sufficiently accurate view on future cash flows is a struggle. AI/ML led solutions will help in identifying cash flow and working capital optimization opportunities.
- Customer segmentation – Organisations with a large customer base often face challenge to have a fit-all-size approach to manage the Order to Cash (OTC) process. Such organizations need a distinct way of grouping customers so that proper attention can be given to take proactive actions. ML based clustering algorithms will help in bucketing customers into groups viz. prompt payers, late payers, large / small customers, high order / low order value customers etc.
How large enterprises can implement successful finance analytics projects
- Align with Specific Needs: Organisations differ in terms of their maturity to adopt and use new-age technologies in the space of corporate finance. The varied nature of product categories and service offerings also mandates the use of customised finance analytics solutions for organizations. Hence, corporate finance leaders and top management must deliberate thoroughly on the type of solutions required, tools that need to be integrated, and mechanisms that should be adopted to build an effective financial analytical system in the company.
- Reliable and Relevant Data: Not only data required in the finance analytics space should be reliable, but it must also pertain to relevant dimensions on which decisions have to be taken. The financial implications of data misuse are huge and could cost an organisation up to 1% of its revenue, thereby leading to a completely devastating impact on the organisation over a longer period.
- Start with Proof of Concept (POCs): Enterprises should start modestly by developing small-scale POCs before moving on with full-fledged deployment. This involves the building of large-scale data and analytics infrastructure. The teams’ morale will remain strong if POCs are implemented successfully, and the teams are provided with the outcomes they can be confident in. This way it becomes simple to persuade shareholders and senior leadership for undertaking large-scale efforts.
The awareness and literacy of professionals handling the responsibility of Corporate Finance must be high and leaders must understand and realise the importance of these solutions for achieving higher profitability and returns on investment. These initiatives must also receive support from shareholders so that cost incurred in developing the required infrastructure should be considered as part of the investment which will yield desired results in the future.
The author is VP – Analytics at Aays Analytics
Disclaimer: The views expressed are solely of the author and ETCIO.com does not necessarily subscribe to it. ETCIO.com shall not be responsible for any damage caused to any person/organization directly or indirectly.
