Our vision

The forecast just became free.

A point of view on where enterprise planning goes in the age of AI — what gets commoditized, what gets more valuable, and the kind of product that wins when forecasting is no longer the hard part.

For thirty years, planning software sold one core thing: the ability to build a model. Sophisticated, multi-dimensional, consultant-grade models that a spreadsheet couldn't hold. That capability was the moat, and it was worth a lot.

AI just made it abundant. A large language model will write the formula, generate the scenario, draft the variance explanation, and assemble the board deck — in seconds, in plain English, with no model builder required. The thing planning vendors charged the most for is becoming the cheapest part of the stack.

The value didn't disappear. It moved. The question is where — and what kind of product is built to catch it.

First, an honest reframe

Planning software never really sold spreadsheets.

The model was the smallest part of the value. What enterprises actually bought was everything around it.

A financial system of record

Budgets, forecasts, headcount, comp, capex, and sales plans — in one trusted place.

Collaboration workflows

Budget requests, approvals, submissions, and variance explanations across hundreds of owners.

Governance

Version control, security, audit trails, and approval chains a board and an auditor can rely on.

Data integration

ERP, CRM, HRIS, payroll, and the warehouse, connected into one operating picture.

The shift

What gets cheap, and what gets precious.

When a capability becomes abundant, its price falls to zero and the scarce thing next to it captures the value. Forecasting is becoming abundant. Here's what sits on either side of the line.

AI makes abundant

  • First-pass forecasts
  • Model building
  • Variance write-ups
  • Board-deck drafts
  • Formula authoring

The analyst labor layer and the modeling moat both erode. A first-pass forecast goes from two days to thirty seconds.

What stays scarce

  • Enterprise context
  • Judgment under uncertainty
  • Accountability & audit
  • Cross-functional buy-in
  • The confidence to act

Context, judgment, and accountability don't commoditize. If anything, they get more valuable as outputs get cheaper.

Where it goes

Planning becomes a question. The product becomes a decision.

The interface stops being a grid of cells and becomes a conversation. Instead of opening a workbook, a CFO asks: why did EMEA gross margin decline? What happens if we freeze hiring? Can we hit Rule of 40 next year? Show the board the downside.

The planning platform still exists underneath — it has to, because the answers have to be grounded and the changes have to be governed. But the product moves from recording what you decided to recommending what you should decide: here are the five moves most likely to improve EBITDA, ranked, with the work already drafted.

That's the category we're building toward. Not a better model builder — decision intelligence.

The real contest

The scarce asset isn't the forecast. It's context plus workflow.

Once forecasting is abundant, the advantage belongs to whoever has the deepest operational and financial context and the strongest position in how decisions actually get made. That's why the interesting threat to standalone planning isn't a chatbot — it's the suites that already sit on the data.

But owning data in silos isn't the same as helping someone decide. Real decisions cross finance, revenue, people, and operations at once, and they have to be explainable and reversible. We think the winning product is decision-first, sits across systems rather than inside one, and is fast enough to adopt that the people who own the plan can actually use it.

Build the context layer. Make the interface a question. Recommend the decision. Keep it all governed. That's the bet.

Why we're building Finicast

Built for the world after free forecasting.

Grounded in context

Connected to your operational and financial data, so recommendations reflect how your business actually runs.

Decision-first

It doesn't just store the plan — it surfaces the moves most likely to change the number, ranked by impact.

Governed by design

Branching, approvals, and a full audit trail make it safe to let AI near the plan in the first place.

Plan for where this is going.

If you think planning is becoming a question and a decision rather than a spreadsheet, we should talk. Bring a plan you run today and see it in Finicast.