--- title: "Stated Demand Is Not Paid Demand: The Institutional Cost of Hypothetical Intent" description: "Willingness to pay stated in a hypothetical setting sits systematically above actual purchase behaviour, because at the moment of the statement no budget constraint, no opportunity cost, and no internal approval friction is engaged. Treated not as measurement error but as a named, auditable conversion factor with a documented accuracy record, the deviation stops distorting capacity and valuation models." url: https://www.beirek.com/en/blog/hypothetical-bias-in-demand-forecasting canonical: https://www.beirek.com/en/blog/hypothetical-bias-in-demand-forecasting published: 2025-09-13 modified: 2025-09-13 category: "Marketing & Consumer Behaviour" category_url: https://www.beirek.com/en/blog/category/marketing-consumer-behaviour language: en-US reading_time_minutes: 8 publisher: BEIREK LLC publisher_url: https://www.beirek.com license: "© BEIREK LLC — citation with attribution and link permitted" keywords: ["hypothetical bias","stated willingness to pay","demand forecasting","letter of intent valuation","pipeline conversion discipline"] topics: ["Consumer research calibration and stated preference reliability","Demand assumptions in capacity and working capital planning","Non-binding pipeline treatment in diligence and deal architecture"] alternate_language_url: https://www.beirek.com/tr/blog/hypothetical-bias-in-demand-forecasting --- # Stated Demand Is Not Paid Demand: The Institutional Cost of Hypothetical Intent > **In short:** Willingness to pay stated in a hypothetical setting sits systematically above actual purchase behaviour, because at the moment of the statement no budget constraint, no opportunity cost, and no internal approval friction is engaged. Treated not as measurement error but as a named, auditable conversion factor with a documented accuracy record, the deviation stops distorting capacity and valuation models. *Demand expressed in a survey, a focus group, or a letter of intent runs consistently above the demand that ultimately clears at a price, and the direction of that gap is remarkably stable. What matters is not the existence of the deviation but its migration, uncorrected, into capacity plans, inventory positions, and valuation models.* --- Place the consumer research presented to a pricing committee next to the actual sales curve of the same product six months after launch, and the distance between the two curves is rarely random; its sign, moreover, is almost always the same. Willingness to pay expressed in a survey or a focus group exceeds the price at which the market actually clears, and stated purchase intent for the same product sits visibly above the demand that converts. The pattern is not confined to a single category or a single research instrument, repeating with identical sign across fast-moving goods, durables, and subscription models alike. What is striking is not the magnitude of the gap but its reproducibility: a systematic deviation with a fixed direction is not measurement noise but a displacement produced by the conditions of measurement themselves. The same pattern surfaces far from consumer markets, in industrial and infrastructure settings, through instruments that carry no binding force. Volume expressed in a capacity reservation form, a letter of intent, a pilot memorandum, or a distributor's annual purchase forecast exceeds the volume that eventually converts into contract, and the gap is independent of the counterparty's good faith or commercial seriousness. An operations director signing an MOU may genuinely intend, at that moment, to take the stated volume; what that director does not know while signing is what the budget committee will prioritise in the following period, how the capital expenditure queue will be sequenced, or where the price of alternatives in the buyer's own supply chain will settle. This asymmetry of information and authority between the moment of statement and the moment of payment is the true origin of the gap between the two numbers. The deviation has a name — hypothetical bias, the systematic divergence between a preference expressed in a hypothetical setting and the same preference exercised against a real cost — and its mechanics are unremarkable. A respondent answering in a hypothetical setting carries no budget constraint; the answer subtracts nothing from any pool of resources under that respondent's control, so no opportunity cost calculation is engaged. At the same time the question asked and the question answered diverge: formally the question reads as whether the respondent would pay a given amount, but the question actually answered is whether the proposition is a good idea, and endorsing an idea is incomparably cheaper than assuming a cost. Layered onto this is the courtesy economy of the interview setting itself, in which the affirmative answer is the preferred answer precisely because it makes the relationship easier to sustain. Reading this tendency as irrationality forecloses the possibility of managing it. The hypothetical statement is a low-cost shortcut for filtering a wide set of propositions, and in that function it genuinely works: a product concept nobody wants fails to attract endorsement even in a hypothetical setting, so the direction of the statement — which option is preferred over which, which attribute precedes which — carries information and can support a decision. What it does not carry is level; the stated price, the stated volume, the stated conversion rate are not magnitudes that can be exported at their own scale. The problem lies not in the shortcut but in its persistence at unchanged weight once conditions change — that is, once the figure leaves a research presentation and enters a cell in a model. In institutional settings the deviation carries a second layer, more insidious than its consumer-research form: the person who expresses the preference and the person who authorises the payment are not the same person. An engineer in the technical unit of a corporate buyer who tests a product and returns favourable feedback, not owning the purchase decision, delivers feedback that carries no budgetary consequence; that answer is honest with respect to the engineer's own knowledge and only a weak indicator with respect to the institution's payment decision. The measure of progress in a pipeline record is therefore not the warmth of the contact or the number of meetings held, but which person, holding which authority, produced which record within the counterparty organisation. Intent recorded without its authority layer makes the sales cycle look shorter than it is and the conversion probability higher than it is. Counter to intuition, the balance sheet expression of the deviation does not appear in the price line; it appears in fixed cost committed to capacity and in working capital committed to inventory. A company that accepts stated demand uncorrected sizes its line capacity, shift plan, warehouse footprint, and first production run against that figure, and when demand clears at its real level the result is not merely lost revenue but a fixed cost commitment that takes months to unwind and a durable slowdown in inventory turnover. That slowdown lengthens the cash conversion cycle, widens the gap between supplier payment terms and collection terms, and typically becomes visible in the quarter in which the working capital or leverage covenant of the credit agreement is tested. A single uncalibrated assumption, in short, echoes separately across three financial statements. The second cost surfaces at the capital markets and M&A table. Where a revenue projection presented in an information memorandum rests on a pipeline composed of non-binding intent documents, the buy-side diligence team separates the two categories — signed binding contracted volume against volume that remains at the level of intent — in nearly every transaction, and declines to price them at the same multiple. In deal architecture that separation typically takes one of three forms: a discount applied to the headline price, an earn-out trigger tied to realisation of the projection, or a pre-closing condition requiring a defined volume to be converted into binding contract. The consequence of all three is identical; demand that remains at the level of statement stays on the seller's side as risk, and an uncalibrated forecast transfers a portion of negotiating leverage directly to the counterparty. The third cost is institutional rather than quantitative, and it is usually the last to be noticed. Once a hypothetical figure has entered a model, it becomes the anchor for subsequent periods; what is debated in budget discussions is no longer the real level of demand but the extent of deviation from the prior forecast, and to the degree the frame shifts in this way, the underlying question is never posed at all. To this is added an ambiguity of ownership: where the record does not show whether the figure originated in research, in sales, or in management's final adjustment, the source of the realisation gap cannot be established retrospectively either. Learning does not institutionalise while the identity of the deviation's source remains unknown, and each new product, each new market entry, reproduces the same deviation from zero. The mechanism that neutralises this tendency is not individual scepticism but the design of measurement and record, and it separates into four components. The first is incentive-compatible test design: constructing a setting in which the statement carries a cost — a refundable reservation fee, a limited allocation of pre-orders, a revocable but monetary commitment — measurably narrows the distance between statement and payment. The second is recording the question: which figure was obtained through which question form, in which context, and from whom, stored alongside the figure itself. The third is the conversion factor: every hypothetical magnitude enters the model only through a named adjustment factor accompanied by that factor's historical accuracy record. The fourth is the authority distinction: each pipeline item is classified according to the level of authority that produced it within the counterparty organisation, and that class forms the basis of weighting. In investment and development mandates, the way BEIREK operates this layer runs through a commitment register in which every demand magnitude entering the model is recorded together with its evidentiary chain. Each line in that register requires three data points: the documentary form of the commitment — binding contract, conditional contract, non-binding intent, verbal statement — the counterparty authority that produced it, and the conversion factor applied to that class together with the reasoning behind the factor. The same register is closed out retrospectively as realisation data arrives, which converts the factor from an assumption into a measure grounded in the institution's own historical accuracy. This structure does not eliminate forecast deviation — such a claim would not be credible in any case — but it renders the direction, the magnitude, and the source of the deviation knowable, and that is precisely what proves defensible in financing discussions. The second mechanism is embedded in the decision rhythm: at every threshold requiring a capital commitment — line investment, first production run, headcount expansion, FID — the demand assumption underlying the decision is broken out as a separate item, and the evidentiary class from which that assumption derives is entered into the investment committee minute. That minute is taken at the moment the proposal is made rather than at the moment approval is granted, because a record created at approval cannot arrest the tendency to rewrite the rationale in the light of the outcome. Producing the counter-argument to the assumption is likewise assigned to a defined role, and that role sits outside the team advancing the proposal; the purpose is not to slow the project but to ensure that the distance between statement and payment is named at least once at the table. When these two records operate together, forecast error resolves into the next revision of the factor rather than into a surprise. What ultimately determines the quality of a demand forecast is not the figure itself but whether the cost conditions under which that figure was stated are known; a yes given at no cost and a yes resting on a payment commitment, written into the same cell at the same weight, leave the model carrying an aspiration rather than a measurement. The question standing in front of the investment decision is therefore not what the customer said, but what the customer put at risk in saying it. ## Key Points - The direction of a hypothetical statement carries information while its level does not; ranking and preference ordering remain broadly reliable, but a stated price point should never enter a model at face value. - In most institutional buyers the person who expresses intent is not the person who authorises payment, and unless the pipeline record separates expressed interest from budget authority, the deviation compounds rather than cancels. - The deviation shows up on the balance sheet not in the price line but in fixed cost committed to capacity and in working capital committed to inventory, where it lengthens the cash cycle. - At the diligence table a pipeline built on non-binding intent is repriced through a headline discount, an earn-out trigger, or a pre-closing conversion condition, transferring negotiating leverage to the buyer. - The neutralising mechanism is not individual scepticism but a recording discipline that admits every hypothetical figure into the model only alongside a named conversion factor and that factor's historical hit rate. ## Questions ### What is hypothetical bias and why does it arise? It is the systematic divergence between a preference expressed in a hypothetical setting and the same preference exercised against a real cost. At the moment of statement no budget constraint, no opportunity cost, and no internal approval friction is engaged, so the cost of an affirmative answer is close to zero. The respondent also tends to answer whether the proposition is a good idea rather than whether the stated amount would actually be paid, and the distance between those two questions forms the core of the deviation. ### Does this make survey-based pricing research unusable? It remains usable for direction rather than level. Ordering information — which option is preferred over which, where an attribute sits in the priority sequence — holds up broadly well even in a hypothetical setting. Stated price points, stated volumes, and stated conversion rates, by contrast, are not suited to direct transfer into a model; those magnitudes become meaningful only alongside a named conversion factor and the documented historical accuracy of that factor. ### How do non-binding letters of intent affect company valuation? Buy-side diligence teams separate the pipeline into signed binding volume and volume remaining at the level of intent, and decline to price the two at the same multiple. That separation typically resolves into one of three forms: a discount applied to the headline price, an earn-out trigger tied to realisation of the projection, or a pre-closing condition converting a defined volume into binding contract. The outcome is the same in every case: uncalibrated demand remains on the seller's side as risk. ### What is the practical way to narrow the gap between stated and actual purchase? Constructing a measurement setting in which the statement carries a cost. A refundable reservation fee, a limited allocation of pre-orders, a revocable but monetary commitment, or a partial advance payment narrows the distance between statement and payment to the extent it removes the answer from the category of costless. In institutional selling this is supplemented by recording the authority level on the counterparty side; favourable feedback from a technical unit and a signature carrying budget authority do not carry equal weight. --- Source: https://www.beirek.com/en/blog/hypothetical-bias-in-demand-forecasting Publisher: BEIREK LLC — https://www.beirek.com