Surveillance Pricing and the Invisible Negotiation

Surveillance Pricing and the Invisible Negotiation

Venture Step E096 asks a simple question with an uncomfortable answer: is a price still an ordinary market offer when a system quietly estimates how much one person will tolerate paying?

Dalton Anderson's argument is that undisclosed individualized pricing is not the same as bargaining. In a normal negotiation, both sides know that the offer can move. They can reveal information, reject terms, make a counteroffer, or walk away. In an invisible negotiation, the seller may have a detailed model of the customer while the customer sees only a number on a screen.

The episode followed E095's investigation of grocery price testing. It widened the inquiry from one platform and one experiment to the history of visible prices, the data behind willingness-to-pay estimates, the RealPage antitrust case, and several federal proposals. That expansion produced the right questions, even though later research required corrections to several details in the recording.

The episode's strongest idea

The useful distinction is not simply fixed price versus changing price. It is shared awareness versus one-sided inference.

flowchart LR
    A["Visible negotiation"] --> B["Both parties know the offer can move"]
    B --> C["Counteroffer, comparison, or exit"]
    D["Invisible negotiation"] --> E["Seller estimates willingness to pay"]
    E --> F["Buyer sees an unexplained offer"]

A store can change a public price because inventory is low. An airline can publish different fares for different departure times. A service can offer a disclosed student discount. Those systems may still raise fairness or competition questions, but they do not necessarily estimate one person's willingness to pay from private or inferred information.

The FTC's current surveillance-pricing work focuses on the latter possibility. The agency has examined intermediary technologies that can draw on information such as location, demographics, browsing behavior, shopping history, channel, or other signals to influence individualized prices or offers.

That is where the negotiation analogy becomes useful. A price can look like a neutral fact while functioning as the output of a private assessment.

Why posted prices entered the conversation

E096 tells a clean origin story in which nineteenth-century Quaker merchants and John Wanamaker replaced universal haggling with fixed prices. The verified history is more interesting and less tidy.

Haggling existed, but historical evidence from antebellum country stores shows that some merchants used standard prices before the Civil War. In Europe and the United States, several retailers helped spread one-price policies, visible tags, advertising, returns, and large-scale store administration. Wanamaker was an influential adopter and popularizer, not the sole inventor of a universal rule.

The correction strengthens the episode's point. Posted prices were not one person's moral breakthrough. They became an operating and trust system. A visible price let a shopper compare stores without first disclosing urgency, social position, or bargaining skill to a clerk. It also let a large retailer coordinate thousands of products and employees.

Software changes the operating cost again. A digital seller can update offers, test treatments, rank products, apply discounts, and identify segments without printing another tag. The policy question is not whether commerce should return to one frozen price. It is which forms of visibility, comparability, and consent should survive when prices become programmable.

Four corrections that change the record

The first correction concerns the Robinson-Patman Act. The transcript calls it the Robertson-Patman or Robertson-Patterson Act and treats it as a broad ban on charging consumers different prices. The FTC's Robinson-Patman guide describes a narrower antitrust statute involving specified discrimination between competing purchasers of commodities. It has several elements, defenses, and a competition-injury requirement. It is not a general federal rule against individualized consumer prices.

The second correction concerns company ownership. RealPage does not own CoStar or Apartments.com. CoStar Group identifies Apartments.com as one of its brands. RealPage is a separate property-technology provider.

The third correction concerns what the RealPage case is about. The Justice Department's current case record identifies alleged collusion and agreements not to compete. DOJ's proposed RealPage settlement addresses nonpublic competitor information, rent recommendations, and features that allegedly aligned pricing. That is an algorithmic coordination problem. It is not evidence that RealPage used an individual renter's browsing history or personal profile to estimate that renter's maximum price.

The fourth correction concerns Instacart's $60 million FTC matter. The FTC did announce a proposed $60 million consumer-refund settlement, but the allegations concerned delivery advertising, satisfaction guarantees, refunds, free trials, and Instacart+ enrollment. It was not a fine for the grocery pricing test discussed in the episode.

These are not cosmetic edits. They prevent three different issues from becoming one dramatic but inaccurate story.

The bills do not do the same job

E096 discusses S. 232, H.R. 4640, and S. 3387. As of July 27, 2026, all three remain introduced.

S. 232 focuses primarily on pricing algorithms that use nonpublic competitor data, with audit and disclosure provisions. H.R. 4640 would address surveillance-based pricing and wage setting, subject to defined routes and conditions. S. 3387 would address different consumer prices for the same or substantially similar offer when informed by surveillance data, with safe harbors and a specific enforcement structure.

None should be described as enacted law or as likely to pass because an episode or enforcement action made the issue more visible. The full comparison appears in [[Federal Bills Targeting Algorithmic and Surveillance Pricing]].

What E096 leaves the reader with

The episode's durable contribution is a way to recognize an information imbalance.

A changing price tells you that something in the system moved. A surveillance-based price raises another question: what did the system decide about you?

That question cannot be answered from a screenshot alone. It requires the input data, identity or segment, decision rule, treatment assignment, comparison price, disclosure, and route to challenge an error. [[What Data Can Estimate Willingness to Pay]] follows that evidence chain. [[When Is Variable Pricing Fair]] turns it into a product and policy review.

For the canonical definition, start with [[What Is Surveillance Pricing]]. Readers following the grocery investigation should continue to E099's [[Algorithmic Grocery Pricing - What Shoppers Can and Cannot See|Algorithmic Grocery Pricing: What Shoppers Can and Cannot See]]. That page separates price testing, dynamic pricing, personalized pricing, and surveillance pricing before attributing any one practice to a company.

This article was developed from the preserved E096 transcript and current primary sources. AI assistance was used for research organization, drafting, and validation. The episode's original recording remains unchanged, and the corrections above apply to the publication package.