The market's third failure: an impossible choice.

Every agency owner contemplating a sale runs into the same wall. The way to a strong price is exposure — put the book in front of enough qualified buyers to start a competition. But the moment word gets out, the asset starts to degrade: clients get nervous, producers update their résumés, and competitors smell blood. Reach maximizes price; secrecy preserves the thing being priced. This is the disclosure dilemma, the third structural failure, and unlike the others it cannot be solved by giving the seller more information — only by changing what information is exposed, and to whom.

§ 01 · The impossible choiceReach versus secrecy.

The two requirements are operationally incompatible under traditional models, and the operational risks of premature disclosure are concrete. Client attrition: policyholders value continuity, and a rumored sale pushes them to seek perceived stability elsewhere, degrading the recurring revenue a buyer is paying for. Employee flight: top producers are the most mobile asset an agency has, and job-security uncertainty triggers the exact departures a competitor will recruit into. Competitor exploitation: local rivals use knowledge of a pending sale to fear-monger to clients and target accounts during the agency's most vulnerable window. Every one of those is triggered by visibility — the same visibility the seller needs for price.

§ 02 · The local bubbleSecrecy at the cost of the price.

The default resolution is the local bubble: quietly shopping the agency to a small ring of known peers. It solves secrecy and creates a worse problem — zero competitive tension. A single buyer who knows they are the only buyer has no reason to bid up. The result is a silent discount driven by competitive vacuum, distinct from the information-vacuum discount of the valuation fog and additive to it: in a real unrepresented sale, both apply at once.

Why can't NDAs fix this? Because an NDA controls what a recipient does with information after they receive it — it does not prevent the exposure in the first place. In a tight local market, the bell rings the moment a name is attached to "considering a sale," and it cannot be un-rung. NDAs treat the symptom; the disease is that traditional M&A surfaces — the broker's rolodex, the conference introduction, the off-market call — are all identity-attached from the first contact. A broker's value is the identity-attached introduction, which is why the legacy model structurally cannot deliver anonymity.

§ 03 · The architectural fixSeparate the data from the identity.

The dilemma dissolves only when performance data is split from business identity — a structure Milly Books calls Shield & Hook. The buyer sees enough to gauge fit; the seller's identity stays hidden until they choose to reveal it.

Listings surface only to verified-fit buyers — matched on the non-identifying criteria above (line of business, carrier mix, geography), not broadcast to a public board. Identity reveal runs on a controlled unmasking sequence the seller drives: the seller approves the buyer, the buyer signs an NDA, the reveal happens at the seller's deliberate trigger, and only then does the secure diligence workspace open. Diligence cannot precede consent.

§ 04 · The honest versionLower leak surface, not magic.

The right claim here is a careful one. Architectural separation produces a dramatically lower leak surface — not impossibility. In very small markets, a determined observer can sometimes triangulate from the visible metrics, and that limit should be stated plainly rather than papered over. What the structure changes is the default: a seller can test the market — list the whole book, or a single fractional slice — observe real buyer interest, and withdraw with no reputational footprint and no employee anxiety, none of which the local bubble allows. For an owner just exploring, that is the difference between a risk-free question and a community-wide announcement. The friction-points playbook covers how this plays out for sellers in practice, alongside the access and execution failures.

◆