Mind the gap

Chapitre

Capital narration

The business plan will no longer be a document

For fifty years, the business plan was a stable object: a binder, a presentation, three scenarios, a five-year horizon. That object is disappearing — not because business owners will stop projecting the future, but because the raw material of the projection is changing in nature. Declared data is giving way to observed flows, the frozen document to a model you can query, and credibility is shifting from the polish of the presentation to the traceability of the assumptions.

Editorial series Read · 9 min Business owners · sale, acquisition, financing

Fig. 0 — The forecast stops being a deliverable and becomes an access right.

The signal: the data-driven business plan is already here

Visible · today

Three observable facts are already converging.

The first is regulatory and French: the roll-out of mandatory e-invoicing, under way for all VAT-registered businesses in France, turns every transaction into structured, time-stamped, usable data. What the forecast used to estimate — actual payment terms, fine-grained seasonality of revenue, customer concentration — becomes a directly observable flow. The same applies to open banking: a company's bank data can now be accessed, with its consent, by third-party analysts. A lender or a buyer no longer needs to believe a spreadsheet; they can plug the spreadsheet into reality.

The second signal comes from investors themselves. Private equity funds and private debt platforms are equipping their analysis teams with AI models that rebuild a challenger forecast from the target's raw data — before they even read management's version. The business plan handed over by the seller is no longer the single reference: it becomes one version among others, tested against versions generated by the other side.

The third signal is editorial: the documents themselves are changing shape. Static data rooms are giving way to live spaces where projections are recalculated at every month-end close. Some young technology companies no longer hand over a file at all: they open read access to their management dashboard, customer cohorts and recurring revenue included. The forecast is no longer a deliverable; it is an access right.

The mechanics: inputs, display and purpose of the forecast

Invisible · today

Beneath these signals, a deeper mechanism is taking shape, affecting the three components of the object: its inputs, its structure and its purpose.

InputsThe input data is changing regime. The traditional business plan rested on declared assumptions — a stated growth rate, a target margin, an estimated addressable market. The business plan that is coming rests on four ranked layers of data, from the verifiable to the wagered.

Fig. 1 — Four ranked input layers. The value of the forecast lies in the wiring between the layers and in the documented source of every figure.

DisplayThe display structure abandons the triptych. The three ritual scenarios — cautious, base, ambitious — were a reading convention, not an honest representation of uncertainty. Philip Tetlock's work on forecasting and Nassim Taleb's on extreme events have embedded a simple idea in investor culture: a future is not presented in three lines, it is presented as a distribution. The new format shows probability-weighted ranges, simulations on the sensitive variables and, above all, an explicit display of dependencies: which assumption breaks which result, and from what threshold.

Fig. 2 — From the conventional triptych to the probability-weighted range: the median carries the commitment, the fan carries the honesty.

Second display shift: versioning. Like source code, the forecast keeps its history — who changed which assumption, when and on what basis. For a buyer, reading the revision history of a plan often says more than the plan itself: a company that revises its assumptions as soon as reality diverges inspires more confidence than one that sticks to its curve. Third shift: queryability. The reader no longer receives a file, they question a model — “what happens to cash if the largest customer accounts for 40% of revenue and extends its payment terms by thirty days?” — and get the recalculated answer, not a footnote.

Fig. 3 — The revision history becomes part of the file: the discipline of updating can be read, and it is priced in.

PurposeThe narrative function, however, does not disappear. This is the point automation enthusiasts underestimate. A business plan has never been merely a forecast: it is an alignment tool, a consensual fiction in the literal sense — a story that owner, investor and lender agree to share in order to make commitment possible. Rita McGrath has shown this for strategy: in environments where competitive advantage no longer lasts, the value of a plan lies less in its accuracy than in its ability to organise discovery — framing testable assumptions, setting the checkpoints at which you decide to carry on or pivot. The data-enhanced business plan does not remove the story; it disciplines it. Conviction remains, but it is now isolated, named and kept apart from what can be observed.

The players' bet: how the business plan changes sector by sector

Announced · tomorrow

Each sector plays out this gap in its own way, and concrete examples show the scale of the shift.

Manufacturing

Key input · digital twin

The forecast draws on the digital twin of the production facility. A precision engineering supplier preparing its succession will no longer present an assumed capacity utilisation rate: it will present machine-by-machine history, the actual energy efficiency curve and a simulation of the impact of a fixed-price electricity supply contract on its gross margin. The buyer will be able to test for themselves the scenario of energy costs doubling or of losing the second-largest customer. The bet of manufacturers who equip themselves: turn transparency into a valuation premium, because a readable asset commands a better price than an opaque one.

Healthcare & social care

Key input · enforceable demographic data

The dominant inputs are demographic and regulatory: population projections by catchment area, tariff paths set by the health authorities, staffing pressures on healthcare professionals. A group of clinics or a network of laboratories will build its plan on enforceable public data — which shifts the conversation with the funder from “do you believe in our growth?” to “do you share our reading of the catchment area?”. The sector's particular difficulty: regulatory risk does not lend itself neatly to probabilities, and the new format will have to show political uncertainty in some other way than a sensitivity line.

Software & subscription

Key input · continuous cohorts

Here the shift is almost complete: monthly recurring revenue, net revenue retention and customer acquisition cost by cohort are flow data, available continuously. The business plan literally becomes an access: the software company raising funds opens up its measuring instrument, and the discussion turns to the interpretation of cohorts rather than the credibility of the figures. The sector's bet lies elsewhere: as AI agents take over part of production, the link between headcount and revenue loosens, and the plan will have to justify a cost path that no longer follows the hiring path — a complete reversal of the traditional reading grids.

Retail & distribution

Key input · geodata, till flows

Inputs become geospatial and behavioural: till data hour by hour, catchment areas recalculated from mobility data, price elasticity measured continuously. A network of outlets up for sale will present a plan in which each site carries its own micro-forecast, aggregated and stress-tested. The advantage is clear for a multi-site seller: value is no longer discussed on average but site by site, which protects the good assets from contagion by the weaker ones.

Energy & environment

Key input · price curves, simulated output

The forecast follows forward price curves and long-term power purchase agreements. A renewable asset developer does not sell a promise of output: it sells an hour-by-hour simulated production profile, backed by historical weather data and crossed with price scenarios. This is the sector where the probabilistic business plan is already the norm — and it foreshadows what the others will become.

The limit of the bet

The players are betting that observed data will replace declared conviction as the currency of credibility. This bet has its limits: a model plugged into reality can create false quantitative comfort — the precision of the decimals masking the fragility of the structural assumptions. And the asymmetry is reversed: when the buyer has its own models, a poorly equipped seller negotiates against a counterparty that knows its business better than it does.

Crossing the gap: preparing your business plan for a deal now

In your hands · now

For a business owner considering a transaction within two to five years, three workstreams are in their hands right now.

Plumbing

Structure your management data so that it can be extracted and kept with its history — the switch to e-invoicing is the natural opportunity to do this properly, rather than in a rush when a data room opens.

Method

In any existing forecast, separate what is observed, what is contextual and what is wagered, and document the source of every assumption. This discipline alone transforms the quality of a discussion with a lender or a buyer, whatever the format of the document.

Stance

Accept that the forecast you hand over will be tested against counter-models, and prepare for it by having the challenger forecast produced yourself — know the pessimistic version the other side will build, before they build it.

The business plan is not dying. It stops being a document and becomes infrastructure — and like all infrastructure, it is prepared long before it is used.

AI in the value chain: the EBITDA upside to negotiate before you sell your business

Venture capital leaves the station: tokenised securities, retail savings and porous borders

The business plan will no longer be a document

Dealmakers: the end of M&A league tables, the rise of proof

E-invoicing and Open Finance: when cash flows become programmable