Hi Robert,
Thank you - this is an insightful share, especially the 4D framework and the Shell examples.
While we increasingly employ AI-enabled dashboards and digital tools to enhance finance processes, I sometimes wonder whether we are merely evaluating productivity improvements rather than the true financial benefit in measurable bottom-line terms.
AI is clearly a productivity enhancer and efficiency driver in operations, where ROI is visible and immediate. However, in finance, the cost–benefit equation appears more fungible and often harder to quantify.
Convincing Boards of the tangible financial contribution can therefore be challenging. Costs are typically front-loaded, while benefits accrue gradually over time. Meanwhile, as new operational and strategic challenges emerge, the original productivity gains risk being absorbed into business-as-usual expectations.
Perhaps the real question is:
Is AI in finance a short-term productivity tool - or a long-term investment in decision making architecture and resilience
Curious to hear contrarian perspectives.
------------------------------
Venkat Pillai
CFO
SLK Global
Kakkanad
+919482008992
------------------------------
Original Message:
Sent: 03-06-2026 06:58
From: Robert Weissen
Subject: ROI in finance is sub-optimal
Venkat, another great subject to debate!
I raised your question with one of our guest presenters, who runs an AI consulting firm, at next week's CIMA Scotland AI for Finance Conference titled "from Quills to Quantum".
All their clients are laser-focused upon ROI and exhibit professional skepticism before they commit. They run a 4-D approach. Discovery and Define phases involve workshops with internal stakeholders reviewing end-to-end processes, identifying pain points, before proposing AI-enabled solutions and moving into Develop and Deliver. The Define phase identifies the benefits / business case (e.g. two thousand man-hours saved per annum) & time to pay-back (3, 6 or 12 months). They have both operational and finance process examples.
Within Shell Finance, we are anticipating productivity gains from AI and digital tools, including in finance processes. For example, leveraging Microsoft Copilot within Power BI reporting apps. Recent trainings on how to prepare data, removing noise from databases (data cleansing), replacing acronyns (column headings) with full text, prior to fine-tuning the AI (for example, train AI to understand that a user request for Actuals is referring to Year-to-Date Value of Work Done), towards achievement of consistent, reliable management reporting.
------------------------------
Robert Weissen
Shell UK Ltd
United Kingdom
------------------------------