11: aop of tomorrow
Back when I started my career as a Financial Analyst, the period of August - December represented thousands of excel files, revisiting financial models, scrubbing actuals, multiple department leaders either unaware that the AOP (Annual Operating Plan) was underway or trying to get their budget requests in, all supported by long office hours.
Many years later, I’ll say not everything has changed.
But some things have and for the better from a process and a technology standpoint.
We now have:
Could ERP (Enterprise Resource Planning) systems enabling “almost” real time data access, centralization and integration across departments.
Tight monthly close processes followed by flux and B/FvA (Budget/Forecast versus Actual) analyses.
Robotic Process Automation that have automated repetitive tasks in AR/AP (Accounts Receivable/Accounts Payable), reconciliations and reporting.
Rolling forecasts, dynamic scenario planning and driver-based modeling FP&A (Financial Planning & Analysis).
Connected planning tools like Datarails, Vena and others that enable cross-collaboration.
Advanced BI (Business Intelligence) platforms that enable simulation and scenario modeling.
Recurring quarterly shareholder reporting.
More modes of payment than just wires, paper checks enabling real time analysis and better liquidity management.
Recent additions have been:
Generative AI in automating tasks, improving decision-making and generating insights.
Agentic AI in autonomously making decisions or executing tasks on its own.
Adoption of Generative AI is a bit further along than Agentic AI in the area of xx.
So with all that is at disposal, what should AOP actually look like?
Here is how we see the AOP process of tomorrow:
1. Aligning on Company’s Strategy
The era of “I believe” or “I think” has lost its luster. Data has gained more importance and Generative AI tools can now summarize external data
2. Base-line forecasting
An independent view of the future sits with Finance especially due to current times being more uncertain than before. ML models can ingest macro and market trends to create a baseline forecast. An organization's planning calendar can span over a few accounting periods so agents can be set up to update assumptions based upon incoming internal and external updates.
3. (Re)Tooling driver based models
Do unexpected variances occur despite bottoms-up drivers based forecasting? When that happens, confidence in drivers fades and dependent variable forecasting gets further complicated. Some variation of average rolling forecasts is the usual fallback.
4. Departmental Planning
NLP based connected planning platforms let users test “what-ifs” and benchmark metrics versus industry peers.
5. Scenario Planning
A growth, conservative and a mid set of scenarios is a framework most organizations are used to. What is the probability of success? That is the question management is most interested in understanding.
Generative AI can create a set of scenarios. ML models can stress-test non-linear impacts.
6. Re-balancing budgets
Most financial models experience difficulty in pushing down changes and therefore create an imbalance between departments. An optimized ML model rebalances budgets.
7. Communication
Generative AI can build board visuals and narratives.
8. Tracking
Agreed budgets need monitoring. This is a relatively untapped area for agents to monitor spend in real time, generate alerts on deviation in agreed spending and take actions.
Armed with new tools and a new process, the next question that CFO’s might want to tackle:
What should the composition of my Finance team now be?
