The Pre-Reference Credibility Gap as Information Asymmetry: Toward a Norm of Verifiable, Honestly-Labeled Proof

1. Introduction: The Circularity of the First Sale

References can only be earned by first winning buyers, and buyers can only be won, in part, by furnishing references.

The first commercial sale is structurally the hardest, because the demand for evidence is circular. A prospective buyer rationally reduces purchase risk by asking who else the vendor has served; yet a newly founded vendor, by definition, has no prior customers to name. References can only be earned by first winning buyers, and buyers can only be won, in part, by furnishing references. The result is a deadlock that no amount of effort or product quality alone can dissolve. We term the interval between zero and the first handful of verifiable references the pre-reference credibility gap.

The founder confronting this gap faces an apparent dilemma. The honest posture — disclosing that a prospect would be among the first customers — risks reading as a warning signal of immaturity or fragility. The deceptive alternatives — borrowing logos, inventing testimonials, or purchasing reviews — promise to manufacture the appearance of an established track record. We argue that this dilemma is false. The deceptive path is now unlawful across the vendor’s principal markets and, independently, corrosive to the very trust it seeks to purchase, because counterfeit signals collapse the moment they are scrutinized. The honest path, properly instrumented, is not a confession of weakness but the construction of a credible separating signal.

This article makes three connected claims. First, the credibility gap is best understood not as a branding shortfall but as a textbook case of pre-purchase quality uncertainty under asymmetric information, and it therefore admits a known class of economic solutions. Second, for a seller who genuinely lacks references, the only signal that is at once available, credible, and lawful is verifiable, honestly-labeled, consent-backed proof. Third, the convergence of economic theory, observed market behavior, and a tightening regulatory consensus across the United States, the European Union, Germany, and Turkey delineates a distinct, previously uncategorized market segment that requires — and is beginning to acquire — a shared norm of verifiable proof.

A note on method and scope is warranted. The argument advanced here is a conceptual and theoretical synthesis, grounded in seminal works in the economics of information and in current statutory and regulatory law. Empirical figures are restricted to externally published sources and are reproduced with attribution; no statistic is generated or estimated by the authors. Consistent with this discipline, the article argues at the level of category and governing norm: it characterizes the conditions any system addressing the gap must satisfy, but it does not specify any particular implementation, and it deliberately abstains from disclosing implementation mechanics.

2. The Credibility Gap as Information Asymmetry

The pre-reference vendor is precisely the high-quality “peach” that the buyer cannot distinguish from a “lemon.”

Akerlof’s (1970) analysis of “the market for lemons” provides the foundational frame. When buyers cannot ascertain quality before purchase, and when sellers possess private knowledge of their own quality, buyers rationally price every offer at the expected quality of the pool rather than at the quality of any individual seller. This uniform discount penalizes high-quality sellers, whose true value exceeds the pooled price, and rewards low-quality sellers, whose value falls below it. High-quality sellers consequently exit or are driven to the margins, the average quality of those remaining falls, the rational discount deepens, and in the limiting case the market unravels toward collapse. Asymmetric information, in short, can degrade or destroy otherwise mutually beneficial trade.

The pre-reference vendor is precisely the high-quality “peach” that the buyer cannot distinguish from a “lemon.” Lacking any verifiable means to substantiate quality claims, the buyer applies the same skeptical discount to the capable unproven vendor and the incapable one alike. The familiar trust discount that early vendors encounter — compressed pricing, prolonged sales cycles, demands for free pilots, or outright refusal to engage — is not buyer irrationality but the buyer’s efficient hedge against unobservable quality. The gap therefore has a determinate economic shape rather than a diffuse reputational one.

Nelson’s (1970) distinction between search goods and experience goods sharpens the diagnosis. The quality of a search good can be verified before purchase through inspection; the quality of an experience good is revealed only through use. Early-stage business-to-business software and services skew heavily toward the experience-good end of this spectrum: a buyer typically cannot fully evaluate reliability, support, security posture, or fitness-for-purpose until after adoption and integration. This intensifies the asymmetry precisely where references would ordinarily substitute for direct inspection — by allowing a prospective buyer to import the verified experience of prior buyers. The pre-reference vendor is thus doubly disadvantaged: selling an experience good while lacking the experiential record that would let buyers act as if it were a search good.

Because the credibility gap has a recognized economic structure, it also has a recognized class of solutions. The economics of asymmetric information does not stop at diagnosing market failure; it specifies the instruments by which informed and uninformed parties can restore efficient trade. That this body of work is foundational rather than peripheral is reflected in its formal recognition: the 2001 Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel was awarded jointly to George A. Akerlof, A. Michael Spence, and Joseph E. Stiglitz for their analyses of markets with asymmetric information. The next section turns to the instruments their work identifies.

3. Signaling and Screening: The Economics of Credible Disclosure

A signal that a low-quality actor can cheaply replicate conveys no information and collapses into a pooling equilibrium.

Spence’s (1973) theory of signaling identifies the first mechanism by which asymmetry can be overcome. The informed party — here, the seller who privately knows its own quality — can transmit credible information by sending a signal that is costly to produce, and, critically, differentially costly: cheaper for a high-quality actor than for a low-quality one. Spence’s requirement is that the cost of the signal decline in ability; the formal single-crossing (Spence–Mirrlees) condition is a later generalization of this differential-cost idea. When that condition holds, only high-quality sellers find it worthwhile to send the signal, and the market sorts into a separating equilibrium in which the signal reliably distinguishes quality. The decisive property is not expense as such but non-imitability: a signal that a low-quality actor can cheaply replicate conveys no information and collapses into a pooling equilibrium, in which the signal is sent by all types and discriminates among none.

Rothschild and Stiglitz (1976) describe the complementary mechanism. Where signaling proceeds from the informed party, screening proceeds from the uninformed one: the screener designs instruments — a menu of options, conditional terms, staged commitments — that induce informed actors to self-sort by revealing their type through the options they choose. In Rothschild and Stiglitz the screening party is the uninformed insurer offering a contract menu to privately-informed applicants; in business-to-business procurement the structurally analogous role is played by the uninformed buyer, who screens via reference requests, trials, security questionnaires, and pilot conditions. The two mechanisms are not rivals but a matched pair. A coherent account of the credibility gap must therefore explain not only what signal a seller should send but how that signal interlocks with the verification the buyer is already attempting to perform.

Shapiro (1983) supplies the dynamic complement. Reputation — an accumulated stock of credible past signals — supports a price premium and disciplines ongoing quality precisely in markets where quality is hard to observe before purchase. The premium is the return to reputation, and the prospect of losing it deters quality-shading. The pre-reference vendor’s defining problem is, in these terms, the absence of this stock: it has not yet accumulated the credible history on which a reputation premium rests, and it cannot conjure that history retroactively without fabrication.

These results yield the article’s central proposition. For a pre-reference seller, the credible separating signal cannot be a customer testimonial: such testimonials are, by hypothesis, absent, and — as Section 5 establishes — fabricating them is now unlawful. The signal must instead be verifiable, honestly-labeled proof of what the seller can actually demonstrate today: a working demonstration, an inspectable sample, a measured outcome from a consented pilot, or an on-the-record founder statement, each presented as exactly what it is and nothing more. The value of such proof depends entirely on its being costly to misuse — that is, on its verifiability and its honest labeling. An explicit honesty-of-label condition is what converts the disclosure from a pooling signal, which any actor could emit by relabeling weak material as strong, into a separating one, which only an actor with genuine, checkable substance can sustain under scrutiny. The claim here is theoretical: the honesty condition is the load-bearing element of the signal, not an optional refinement of it.

4. Market Evidence: Scale, Buyer Behavior, and the Empty Quadrant

Each new entrant begins with no references, so every additional solo founder reproduces the same pre-reference asymmetry.

The theoretical case is reinforced by observable conditions on both sides of the market, reported here strictly from externally published sources. On the demand side, the population subject to the credibility gap is large and growing. As reported by Carta, the solo-founded share of new startups has risen to roughly 36 percent, up from approximately 24 percent in 2019 (Carta, Solo Founders Report, 2025). Because each new entrant begins with no references, every additional solo founder reproduces the same pre-reference asymmetry, expanding the affected population over time. The gap is thus not a residual edge case but a structural feature of an increasingly atomized founding landscape.

Observed buyer behavior, in turn, rewards exactly the checkable evidence that new vendors lack. As reported by G2, 86 percent of software buyers consult peer-review sites during the purchase process (G2, Software Buyer Behavior Report, 2021). This is a direct empirical instance of screening in the sense of Section 3: buyers actively seek third-party-verifiable signals before committing, and they weight precisely the form of proof that a pre-reference vendor cannot supply. The behavior confirms that the discount applied to unproven sellers is deliberate and information-seeking rather than incidental.

The content of buyer scrutiny points the same way. As reported by Capterra, security is the leading factor in business-software purchasing, cited by roughly 50 percent of buyers — ahead of any other single consideration (Capterra, Security Features Survey, 2023). Security is a domain in which straight, checkable answers dominate borrowed credibility: a buyer can verify a clear, specific representation in a way it cannot verify a logo or an unsourced endorsement. The pattern across these sources is consistent — buyers reward verifiable proof and discount its absence.

The supply side reveals a structural vacancy. The adjacent categories that ostensibly address vendor credibility each presuppose that proof already exists. Deal rooms and sales rooms are built to organize case studies, return-on-investment documents, and security materials that an established team already possesses. Testimonial and review tools are designed to capture feedback from customers the new vendor does not yet have. Review marketplaces require an existing customer base before a vendor can be meaningfully listed. Security trust centers attest compliance posture rather than commercial trust or founder credibility. None of these instruments generates a seller’s first credible signals; each begins where the credibility gap ends. The conclusion is that an addressable and structurally underserved segment exists at the earliest stage — an empty quadrant in the proof-tooling landscape. We restate that every figure cited above is drawn from an externally published source and that no statistic has been generated by the authors, consistent with the same evidentiary discipline the proposed norm would impose.

5. The Regulatory Landscape: Why a Verifiable-Proof Norm Is Both Needed and Lawful

The lawful path and the economically efficient separating signal coincide.

The economic case for verifiable proof is matched by a legal one that has, in recent years, hardened against the deceptive alternative across the vendor’s principal markets. In the United States, Section 5 of the Federal Trade Commission Act (15 U.S.C. §45) prohibits unfair or deceptive acts or practices in commerce, the long-standing basis for treating misrepresentation as actionable. This general prohibition has been given specific and severe form by the Commission’s 2024 rule on the use of consumer reviews and testimonials (16 CFR Part 465), which expressly bans fake reviews and testimonials, including those generated by artificial intelligence, the purchase of positive or negative reviews, and the creation of company-controlled websites that falsely purport to be independent. Relatedly, the Commission’s Endorsement Guides (16 CFR Part 255) require the disclosure of material connections between endorsers and sellers. Fabricated social proof is therefore not a gray area subject to reputational risk alone but conduct exposed to direct enforcement.

In the European Union, data-protection law channels any proof that involves a third party toward a consent-based, revocable model by operation of law. The General Data Protection Regulation requires a lawful basis for processing personal data (Art. 6) and, where consent is that basis, imposes strict conditions on it, including the requirement that consent be as easy to withdraw as to give (Art. 7, and specifically Art. 7(3)). The right to erasure (Art. 17) further entitles a data subject to have personal data removed. Together these provisions make consent-backed, revocable proof not merely a courtesy but a structural requirement: third-party quotes, outcomes, or identifying material may lawfully be published only on a stored consent basis that the subject can withdraw, triggering removal.

The European Union’s Artificial Intelligence Act adds a transparency dimension directly relevant to proof artifacts. Article 50 imposes transparency obligations that include marking AI-generated or AI-manipulated content in a machine-readable form, disclosing when a person is interacting with an AI system, and disclosing deepfakes. To the extent that any proof artifact is AI-generated or AI-manipulated, its provenance must therefore be disclosed; a prudent reading extends the same candor to materially AI-assisted artifacts. This reinforces, from a second legal direction, that representations to buyers must be accurate about what they are.

Germany supplies a further reinforcing layer through its Act Against Unfair Competition. Section 5 of the UWG prohibits misleading commercial actions, and Section 5a prohibits misleading omissions — the withholding of material information that the recipient needs to make an informed decision. The pairing is significant: §5a means that honesty is not satisfied merely by avoiding false statements but requires the affirmative disclosure of material facts, which is the legal analogue of mandatory honest labeling. Turkey completes the picture across the third target market: the Law on the Protection of Personal Data (KVKK, Law No. 6698) provides consent and data-protection requirements broadly parallel to the GDPR, while the E-Commerce Law (No. 6563) and the Consumer Protection Law (No. 6502) constrain misleading and deceptive commercial practices in online and consumer-facing contexts.

The synthesis is striking. Four independent legal systems, proceeding from different doctrinal starting points — consumer-protection enforcement in the United States, data protection and artificial-intelligence transparency in the European Union, unfair-competition law in Germany, and a composite of data-protection, e-commerce, and consumer law in Turkey — converge on the same prescribed behavior: proof presented to buyers must be verifiable, accurately disclosed, consent-based where it implicates third parties, and non-deceptive. None of these regimes compels a vendor to produce proof; what they jointly do is foreclose the deceptive alternative and constrain how proof may lawfully be presented. A verifiable-proof norm is therefore not a trade-off between what is lawful and what is effective. It is the rare case in which the lawful path and the economically efficient separating signal coincide.

6. A New Category and the Case for a Verifiable-Proof Norm

The norm is imitable in name and not in fact — which is precisely what makes adherence to it informative.

Three independent lines of argument converge on a single conclusion. The economic analysis of Sections 2 and 3 establishes that the credibility gap is an information-asymmetry problem whose resolution requires a costly, non-imitable separating signal. The market analysis of Section 4 establishes that the demand population is large and growing, that buyer behavior actively rewards verifiable proof, and that no existing tooling category serves the earliest stage. The regulatory analysis of Section 5 establishes that the law across the vendor’s principal markets forecloses precisely the deceptive behavior the economics already condemns. Taken together, these forces delineate a distinct and previously uncategorized market segment: a pre-reference proof layer that sits before, and is structurally different from, every category that assumes proof already exists.

We propose a norm to govern this segment. In its essential form, the norm has two clauses. First, every public proof claim must carry exactly one honest label drawn from a fixed taxonomy — an illustrative instantiation of which spans the spectrum from in-production results, through clearly-marked demonstrations and inspectable samples, to measured pilot outcomes and first-person founder statements. Second, any element of proof that implicates a third party must rest on stored, revocable consent. The first clause operationalizes the honesty-of-label condition that Section 3 showed to be the load-bearing element of a separating signal; the second clause operationalizes the consent and erasure requirements that Section 5 showed to be legally mandatory.

In signaling terms, a mandatory honest taxonomy is what converts a reference shortage into a trust advantage. By forcing each artifact to declare exactly what it is, the taxonomy makes it impossible for a demonstration to masquerade as a live customer result, or for a founder statement to pose as an independent endorsement. It thereby enforces separation: a buyer can reliably read the strength of each claim from its label, and a vendor with genuine substance is no longer pooled with one that merely relabels weak material as strong. The norm does not ask buyers to extend unearned trust; it equips them to verify, which is what the economics of screening predicts they will do regardless.

The contribution is best understood at the level of the field rather than any single offering. A nascent category becomes legitimate when its participants adopt a shared norm that buyers can rely on across vendors; absent such a norm, each honest vendor’s signal is undermined by the noise of those who cut corners, reproducing the very pooling equilibrium that defeats trade. This article articulates that norm and the conditions any system addressing the gap must satisfy to instantiate it: mandatory honest labeling, consent-backed and revocable third-party proof, and disclosure of AI-generated content.

Two features distinguish the instrument this norm calls for. First, it operates at the pre-reference stage that the adjacent categories skip. Second, and more consequentially, it does not merely organize proof a seller already holds; it is designed to help the seller EARN its first reusable, verifiable proof — converting first engagements into durable, consent-backed references that compound over time. A system built to generate a seller’s first credible signals, rather than to display signals that already exist, is a categorically different instrument from the proof-display tooling surveyed in Section 4, and it is the instrument the pre-reference gap requires.

Two objections merit a brief answer. The first holds that honest labeling is self-defeating, because openly marking a result as a demonstration rather than a customer outcome forfeits persuasive force. The reply is that this candor is the cost of the signal, not a flaw in it: an honest label is credible precisely because a low-quality actor gains nothing by adopting it, which is the defining property of a separating signal in Spence’s (1973) sense. A signal that imposes no cost on misrepresentation cannot separate. The second objection holds that a norm confers no durable advantage because it can be copied. The reply is that the value of this particular norm lies in the fact that faking is structurally precluded: a competitor can announce the same labels, but cannot cheaply manufacture the verifiable substance the labels assert without actually possessing it. The norm is imitable in name and not in fact, which is precisely what makes adherence to it informative.

7. Conclusion

If you have no references, do not fake them — send verifiable, honestly-labeled proof.

The argument of this article reduces to a single chain. The pre-reference credibility gap is not a marketing deficiency but a structural information-asymmetry problem in the tradition of Akerlof (1970): buyers who cannot distinguish capable unproven sellers from incapable ones rationally discount all of them, suppressing trade. Such asymmetry is resolved, as Spence (1973), Rothschild and Stiglitz (1976), Shapiro (1983), and Nelson (1970) jointly establish, by costly and non-imitable separating signals — never by counterfeit proof, which pools rather than separates and collapses under scrutiny.

For the pre-reference seller, that separating signal is verifiable, honestly-labeled, consent-backed proof. The distinctive feature of the present moment is that economics and law now point in the same direction. The signal that the theory of information asymmetry identifies as efficient is the same behavior that consumer-protection, data-protection, artificial-intelligence-transparency, and unfair-competition law across the United States, the European Union, Germany, and Turkey now make the only lawful form of proof. Verifiable, honestly-labeled, consent-backed proof is therefore simultaneously the efficient equilibrium and the lawful one.

This leaves a standing question for the field. A functioning trust market for early-stage vendors does not arise spontaneously; it requires the adoption of a shared norm of verifiable proof, so that an honest vendor’s signal is legible to buyers and not drowned by those who fabricate. The precondition for the category is the norm, and the precondition for the norm is collective adoption.

The norm itself reduces to a single principle, which is at once a research claim and a practical rule: if you have no references, do not fake them — send verifiable, honestly-labeled proof. We offer this as a version-one statement, with explicit scope limits. It is a conceptual synthesis grounded in seminal economics and current law rather than an empirical study; it restricts its figures to externally published sources and invents none; and it confines itself to the public case for the category and its governing norm, leaving questions of implementation to separate work. We invite scholarly comment, empirical testing, and critique.