The case for integrated planning — and why AI makes the absence of it more dangerous, not less.
There's a story I use when I'm talking to CFOs about integrated planning. It's not a hypothetical. It's a company called Cochlear — a publicly listed medical device business. In a single trading day, the company lost approximately 40% of its market value recently.
The reason wasn't fraud. It wasn't a product failure. It was a disconnect. Sales were declining. Inventory was building. The two facts were sitting in different systems, owned by different teams, measured against different KPIs. Nobody connected the dots in time. The market did.
When I ask CFO audiences how many of them are operating with the same structural vulnerability, the room gets uncomfortable. Because the honest answer, for most mid-to-large organisations, is: yes.
Most finance functions know they have a data fragmentation problem. What they underestimate is the risk that fragmentation creates — particularly now that AI is being layered on top of it.
Here's the dynamic I see repeatedly: an organisation invests in AI tools to improve forecasting, automate reporting, or generate commentary. The tools are good. But the data feeding them is fragmented. Sales live in one system. Inventory in another. Headcount in a spreadsheet. Finance in the ERP. There's no single connected model that shows the cause-and-effect relationship between these inputs.
When AI is applied to that fragmented foundation, it doesn't fix the fragmentation. It produces confident, well-articulated outputs that are built on incomplete information. That's arguably more dangerous than a spreadsheet, because it has the appearance of rigour.
If you don't have integrated planning — if you don't understand the cause and effect between your departments — you're not ready to layer AI on top of it. That's your baseline.
I want to give you a concrete picture, because 'integrated planning' is a term that gets used loosely. Let me describe what we built for one of the investment banks in our client portfolio — a globally operating institution with more than 800 legal entities and over 1,000 active users of the planning application.
When we started the engagement, their budget cycle took two months. By the time it was complete, it was already out of date. Rolling forecasts — the aspiration — were impossible at that cadence.
After implementing a properly integrated planning environment, with AI infused across the workflow, they can now complete a rolling forecast in four to six days. More importantly, the system responds to real-world events. When macro conditions shift — supply chain disruption, geopolitical instability, commodity price movement — the model recalibrates. Frontline decision-makers get live insight, not a stale plan from last quarter.
75%+
Reduction in forecast cycle time — from 2 months to 4-6 days
800+ global entities, 1,000+ users
When integrated planning is working properly, with AI layered appropriately, it delivers four things that individually sound incremental but together are genuinely transformational:
I want to be clear about something: I'm not dismissing the sophistication of finance teams who have built complex Excel environments. Many of them are genuinely impressive pieces of engineering. But they have a structural limitation that AI cannot fix: they are disconnected by design.
A spreadsheet that sits in finance doesn't know what's happening in supply chain. An Excel model in the sales team doesn't feed into the headcount plan. The connections are manual, maintained by people, and dependent on someone remembering to update something. When that person is on leave, or when they leave the business, the connections break.
Integrated planning replaces those manual connections with a live data model. AI then operates on top of that model — not on top of a series of disconnected files.
If you're reading this and recognising your own organisation in the description above, the path forward is cleaner than it might seem. In our experience, the most effective starting point is not a full transformation program — it's a targeted proof of concept that connects two or three currently disconnected planning domains and demonstrates what a live, integrated model looks like.
Once the board and leadership team can see a rolling forecast that updates automatically when assumptions change, the conversation about broader transformation becomes significantly easier.
The Cochlear story is an extreme example. But the structural vulnerability it illustrates is common. And with AI amplifying the outputs of whatever data foundation you're sitting on, the cost of not addressing it is rising.
In the next blog in this series, I'll address something most CFOs are completely unaware of: the cost of AI tokens — and why it's about to become a CFO-level line item.
Amendra Pratap is the Founder and Managing Director of Octane Solutions, an IBM Gold Partner specialising in AI-powered finance transformation across Australia, New Zealand, and the Pacific.