Retail Media and Shopper Data Checklist: Data Ownership and Basis, Supplier and Brand Terms, Measurement and Attribution Claims, Clean Room Arrangements, and Consumer Rights Handling

By Casey Scott McKay · 2026-04-02

A ten-phase working checklist for retailers building media networks, brands buying from them, agencies sitting between, and the vendors supplying the infrastructure. Phases one and two run the inventory and repair the loyalty programme documents that carry the entire structure. Phases three through five build the privacy architecture, the sensitive-category suppression, and the consumer rights machinery. Phases six and seven cover clean rooms and measurement substantiation. Phases eight through ten cover commercial terms, the interception exposure, and protecting what is genuinely proprietary. Each phase closes with a gate.

IP and Technology > Privacy Data Security | Checklist | Published 2 April 2026 - Updated 7 August 2026 | Casey Scott McKay - marksy.us

How to use this checklist

Four parties use this document and each needs a different subset.

A retailer building a network runs all ten phases, with Phase 2 on the critical path because the notice refresh has a lead time nothing else does. A brand buying media runs Phases 1, 6, 7, and 8 from the other side of the table. An agency runs Phases 7 and 8. A technology vendor runs Phases 6 and 10.

The doctrinal background is The Aisle Is an Advertising Network; the operational treatment with worked engagements is Building a Retail Media or Shopper Data Business; the cluster is assembled in the Retail Media and Shopper Data Toolkit.

Start from the position that the shopper data is not property. It is a set of facts, unprotectable under Feist Publications, Inc. v. Rural Telephone Service Co., controlled entirely by contract, technical access, and regulation. Every phase below builds one of those three.

Phase 1. Inventory

[ ] List the data sources: transactions, loyalty enrolments, web and app behaviour, in-store sensing, and third-party acquisitions. Record provenance and permitted uses for each.
[ ] Flag the third-party sources specifically, because that is where unexamined contractual restrictions live and where the client's freedom is narrowest.
[ ] Obtain the loyalty programme terms and the privacy notice as separate documents, with their dates. Expect both to predate the advertising business.
[ ] Identify what is actually being sold: on-site placements, in-store media, off-site targeting, measurement and insights. Analyse them separately; their obligations differ.
[ ] Draw the counterparty map: platforms, agencies, measurement vendors, clean room providers, identity vendors, and brands. The map is the compliance perimeter.
[ ] Collect the sales materials — rate card, methodology deck, case studies. These are representations and legal has usually never seen them.
[ ] Read the supply agreements for data terms. Frequently silent, which means the retailer holds by default and the brand's access is a concession rather than a right.
[ ] Note whether there is a private label business, since that is a structural conflict brands will eventually raise.
[ ] Record who could answer each question above, because in most organisations the answers are distributed across marketing, engineering, merchandising, and a vendor.
[ ] [Gate] The inventory is written, dated, and owned by a named person.

Phase 2. Repair the foundation documents

[ ] Reconcile three documents — programme terms, privacy notice, and the network's public description — so all three describe the same activity. Inconsistency between them is the first finding in any inquiry.
[ ] Replace generic language. "For marketing purposes" does not disclose cross-context behavioural advertising, off-site targeting, clean room collaboration, or model training.
[ ] Describe the identifier. A loyalty number functioning as a persistent advertising identifier is doing something the shopper did not enrol for.
[ ] Refresh actively rather than editing quietly. A material processing change communicated by a silent page edit is a weak position.
[ ] Build and test the opt-out where the notice says it is. Notices routinely describe mechanisms that do not function.
[ ] Version and archive the notices, since which notice was live on which date determines the analysis for any period.
[ ] Treat the refresh as the critical path, because it has a customer communication lead time that the technical work does not.
[ ] [Gate] A shopper reading the notice would recognise what the business actually does.

Phase 3. Build the privacy architecture by product

[ ] On-site: confirm advertising is identified as advertising and that ranking basis is disclosed, against the deception standard at 15 U.S.C. § 45.
[ ] Off-site: treat the activity as a share and probably a sale under the state comprehensive statutes, since the consideration is the advertising revenue whether or not money changes hands for data.
[ ] Implement universal opt-out signal recognition and confirm it reaches the advertising stack.
[ ] Apply opt-outs at audience build, not campaign launch, because an audience constructed from opted-out records is already the disclosure.
[ ] Read the platform's role characterisation critically. Terms casting the retailer as controller and the platform as service provider should be tested against what the platform does with the data.
[ ] In-store: gate every sensing deployment with a written description of what it captures, retains, and infers, reviewed before the purchase order.
[ ] Screen cameras carefully. Counting, estimating demographics, and recognising individuals are three different legal positions separated by a software setting. See Your Face as Data, Building a Biometric Compliance Program, and the Biometric Data Checklist.
[ ] Treat in-store location as precise geolocation, whatever its source.
[ ] Treat staff as data subjects, since sensors that observe shoppers observe employees.
[ ] Complete any required assessment for targeted advertising or sensitive processing before launch rather than after.
[ ] [Gate] Each of the four products has its own written basis, and none relies on another's.

Phase 4. Sensitive-category suppression

[ ] Accept that a basket is a health record, and sometimes a religion, pregnancy, or immigration-status proxy, and that the statutes reach inferences rather than only stated attributes.
[ ] Suppress at segment creation, not campaign launch, because the segment exists before it is used and its existence is the processing.
[ ] Map product categories to inference risk, maintained alongside the merchandising taxonomy rather than against a static list, since the taxonomy changes continuously.
[ ] Set a minimum segment size and enforce it in the platform rather than in a policy document.
[ ] Screen children's exposure under 15 U.S.C. § 6501 and 16 C.F.R. Part 312, plus the state age-appropriate design duties reaching general-audience services. See Building for Someone Who Cannot Consent.
[ ] Screen the marketing channels separately under 47 U.S.C. § 227 and 15 U.S.C. § 7701, where consent records are the perennial gap. See Permission to Reach Someone and the Marketing Communications Toolkit.
[ ] Screen for the eligibility trap. If shopper data ever informs credit, insurance, or employment decisions, 15 U.S.C. § 1681 may convert the dataset into a consumer report with a different regime attached.
[ ] [Gate] No audience can be built that the suppression policy would have prohibited.

Phase 5. Consumer rights machinery

[ ] Map every intake channel — website, app, call centre, in-store, email, browser signal — and confirm each reaches the same workflow.
[ ] Test end to end, including the channel nobody uses, because that is the one that will be tested by a regulator.
[ ] Confirm deletion actually propagates to the advertising stack, the clean room, the measurement vendor, and any third-party recipient, rather than only to the customer database.
[ ] Confirm access requests can be answered including inferences and segments, which most systems were not built to surface.
[ ] Set and meet the statutory response windows, and record the timestamps.
[ ] Flow the obligations to processors in contract, and confirm each has a mechanism rather than a promise.
[ ] Keep the request log with outcomes, since the log is the evidence of the programme. See the State Privacy Law Applicability and Readiness Checklist and the Marketing Privacy Compliance Checklist.
[ ] [Gate] A deletion request submitted through any channel removes the individual from every downstream system within the statutory window, provably.

Phase 6. Clean rooms and derived data

[ ] State plainly what the clean room does and does not do. It permits collaboration without raw data exchange. It does not create a legal basis, and regulators have said so.
[ ] Document the configuration: minimum aggregation thresholds, permitted query types, query review, output review, and any noise injection — with a record of who may change them.
[ ] Negotiate output ownership and post-termination retention for both parties.
[ ] Negotiate derived model ownership, which the standard agreements handle badly. A model trained on combined data has more than one claimant.
[ ] Negotiate the contamination question expressly: may the retailer use what it learned from Brand A to advise Brand B, and within what limits? See the Competitive Intelligence Checklist.
[ ] Establish the vendor's processing role against what it actually does, not against what its template says.
[ ] Prohibit re-identification, require reporting of attempts, and attach a termination right.
[ ] Retain query and output logs for a stated period, because in any dispute the logs are the evidence.
[ ] [Gate] The configuration is documented, the ownership of every output and model is stated, and the contamination boundary is written down.

Phase 7. Measurement substantiation

[ ] Write the methodology document before the claim: attribution model, lookback window, control group design, assumptions, limitations, and period covered.
[ ] Distinguish attribution from incrementality in the language used. Last-touch attribution is not a measure of incremental sales, and incrementality without a holdout group is not incrementality.
[ ] Constrain the sales materials to what the document supports, since case studies presenting the best campaign as typical are the classic exposure.
[ ] Version the methodology and notify changes, because a brand comparing periods across a silent model change has been misled even if nobody intended it.
[ ] Expect the competitor route under 15 U.S.C. § 1125, with standing under Lexmark International, Inc. v. Static Control Components, Inc. and no compliance defence following POM Wonderful LLC v. Coca-Cola Co.. See False Advertising Under the Lanham Act and Bringing and Defending a Lanham Act False Advertising Claim.
[ ] Expect the regulatory route under 15 U.S.C. § 45, which reaches business-to-business representations.
[ ] Address the marking-your-own-homework problem by moving towards independent verification, which is where the market is heading and which is a commercial advantage to adopt early.
[ ] Set aggregation thresholds on insights, since an insight granular enough to identify an individual is personal data whatever it is called.
[ ] [Gate] Every number in the sales deck traces to a methodology document that supports it.

Phase 8. Commercial terms

Retailer side

[ ] Settle the data question in the supply agreement, not only in the media agreement, because that defines the brand's baseline access.
[ ] State what the brand receives: granularity, frequency, format, permitted uses, and whether it may combine the data with other retailers' data.
[ ] Restrict post-termination retention with deletion and confirmation.
[ ] Reserve aggregate category learning above a stated threshold, expressly, so the network's core value is not argued about later.
[ ] Address the private label conflict before a brand raises it, with a firewall description that could be verified.
[ ] Scope the brand's mark licence to the campaign, subject to approval, terminating cleanly. See Two Brands, One Product and the Co-Branding Agreement Checklist.
[ ] State creative ownership for assets the retailer's studio produces. See Who Owns the Work and the Agency Engagement Checklist.
[ ] Build one standard agreement rather than negotiating each deal, since this is the highest-leverage document in the engagement.

Brand side

[ ] Negotiate audit rights at the point of the spend commitment, which is the only moment they are obtainable.
[ ] Scope every asset and data grant to the campaign: no retention, no derivative models, no use to inform private label or competitors.
[ ] Establish your processing role in writing rather than accepting the retailer's characterisation.
[ ] Ask whether a holdout group exists before accepting an incrementality claim.
[ ] Require methodology change notification.
[ ] Ask how the private label firewall is verified, and record the answer.
[ ] Plan for incremental improvement across renewals where the retailer controls category access, since a single hard negotiation will not succeed. See the Channel Partner IP Checklist.
[ ] [Gate] Both sides can state what the other may keep, learn, and reuse.

Phase 9. Interception and pixel exposure

[ ] Produce the tag inventory for web and app: what is on each page, what it sends, to whom, and why. Most clients cannot, and producing it is the single most diagnostic early deliverable.
[ ] Assess the interception theory under 18 U.S.C. § 2511 and, more importantly, the state analogues with statutory damages, which drive the class exposure.
[ ] Check the video question under 18 U.S.C. § 2710, since product video plus an advertising pixel is the recurring fact pattern.
[ ] Expect partial help from standing arguments in Spokeo, Inc. v. Robins and TransUnion LLC v. Ramirez, since claims with common-law analogues survive the concrete-harm analysis.
[ ] Note the constitutional backdrop at Sorrell v. IMS Health Inc., useful occasionally and rarely the lead argument.
[ ] Remediate forward while defending: remove unnecessary tags, gate the rest behind consent where required, and date the change.
[ ] Repeat the exercise for the mobile app, which is frequently forgotten and frequently worse.
[ ] Set a recurring inventory cycle, because marketing adds tags continuously and nobody removes them.
[ ] [Gate] Somebody can state, page by page, what the client's properties transmit and to whom.

Phase 10. Protect what is proprietary

[ ] Enumerate the identity graph — the resolution logic linking loyalty account to device to household to transaction — as a trade secret under 18 U.S.C. § 1839, with the access controls that requires and the claim under 18 U.S.C. § 1836 available.
[ ] Enumerate segment definitions and measurement methodology likewise, and manage the disclosure tension with tiered disclosure under confidentiality rather than by publishing or refusing.
[ ] Consider patenting the technical implementation under 35 U.S.C. § 101, remembering that a claim drafted as matching records across sources will not survive Alice Corp. v. CLS Bank International and the Mayo Collaborative Services v. Prometheus Laboratories, Inc. framework. See What Can Actually Be Patented.
[ ] Register the network's own brand under 15 U.S.C. § 1052 with ordinary clearance and portfolio management. See the Trademark Portfolio Management Toolkit.
[ ] Record model provenance, since a model's ownership depends on where its training data came from.
[ ] Do not attempt to claim ownership of the transaction log, which is a set of facts under Feist and the thin-compilation rule at 17 U.S.C. § 103, with the organising system excluded by 17 U.S.C. § 102(b). See Selling Something You Cannot Own and the Data Licensing and Rights Toolkit.
[ ] [Gate] The derived assets are enumerated, access-controlled, and distinguishable from the raw data that is not protectable.

Launch readiness

Before an off-site product goes live, all of the following should exist with a named owner.

[ ] A reconciled notice and programme terms describing the actual activity.
[ ] An opt-out working through every channel the notice names, plus universal signal recognition, propagating at audience build.
[ ] A sensitive-category suppression list applied at segment creation.
[ ] A tag inventory for web and app.
[ ] A processing role map with contracts matching the roles.
[ ] A methodology document for every claim in the sales materials.
[ ] A standard media agreement with data, retention, contamination, mark, and audit terms settled.
[ ] A clean room configuration record, where applicable.
[ ] A completed assessment, where the applicable statutes require one.
[ ] An in-store deployment gate, where any physical sensing exists.
[ ] [Gate] Every item above has a named owner, because a launch list without owners is a document rather than a control.

The annual review

[ ] Re-read the notice against what the network now does, since products launch faster than notices are updated.
[ ] Re-run the tag inventory.
[ ] Test the opt-out end to end through every channel.
[ ] Review the sensitive-category list against the current merchandising taxonomy.
[ ] Check for new counterparties in the data flow and confirm each has a contract matching its role.
[ ] Re-read the sales materials for claims that have drifted beyond the methodology document.
[ ] Confirm the methodology version in use and whether changes were notified.
[ ] Audit the in-store estate for sensing added since the last review.
[ ] Re-check clean room thresholds against the agreed configuration.
[ ] [Gate] Nothing in the file is more than twelve months old and unverified.

Failures this checklist prevents

[ ] The off-site product launched on a notice written for a coupon programme. Phase 2.
[ ] The opt-out that reaches the customer database and not the advertising stack. Phases 3 and 5.
[ ] The sensitive-category audience built before anyone reviewed it. Phase 4.
[ ] The class action over a pixel nobody knew was on the page. Phase 9.
[ ] The brand dispute over an incrementality claim with no holdout group. Phase 7.
[ ] The clean room collaboration followed by a private label launch, with no contamination clause. Phases 6 and 8.
[ ] The camera that started recognising faces after a vendor software update. Phase 3.
[ ] The measurement model changed silently mid-year, leaving brands comparing incomparable periods. Phase 7.
[ ] The deletion request that removed the record from one system out of six. Phase 5.

Party-specific short forms

The retailer

[ ] All ten phases, with Phase 2 on the critical path.
[ ] Phase 9 as the fastest diagnostic, since the tag inventory reliably surfaces both litigation exposure and vendor relationships nobody in legal knew about.
[ ] Phase 8's standard agreement as the highest-leverage document, drafted before the sales team signs the first ten deals on purchase orders.

The brand

[ ] Phases 1, 6, 7, 8 from the other side.
[ ] Negotiate at the spend commitment, which is the only point of leverage.
[ ] Ask three questions and record the answers: is there a holdout group, how is the private label firewall verified, and who owns a model trained on our data.

The agency

[ ] Phases 7 and 8.
[ ] Do not restate network claims as your own; attribute them in writing.
[ ] Establish your processing role in both directions, since agencies are frequently processors one way and controllers the other. See Managing an Agency Relationship and the Advertising Agency and Creative Services Toolkit.

The technology vendor

[ ] Phases 6 and 10.
[ ] Do not let marketing position the product as a compliance answer while the contract disclaims responsibility for legal basis.
[ ] Settle the processing role honestly, since a vendor deriving its own analytics is determining purposes.
[ ] Design the logging to help rather than harm, since in any dispute the vendor holds the record.

Three worked applications

The launch six weeks out

A regional chain plans to launch off-site targeting on loyalty data.

[ ] Phase 2 first and immediately. The loyalty terms almost certainly describe a coupon programme and will not support the product, and the remediation involves a customer communication with a lead time.
[ ] Phase 5 in parallel, testing the opt-out through every channel rather than assuming it works.
[ ] Phase 4 before the first audience is created, not before the first campaign.
[ ] Phase 3's platform review, reading the platform's role characterisation against what it actually does.
[ ] Tell the business the launch date is a function of the notice refresh, not the technical integration. Unwelcome, correct, and far better delivered six weeks out than six weeks late.
[ ] Consider a sequenced launch: on-site first, which is the least exposed product, holding off-site until Phase 2 completes.

The brand disputing reported lift

A manufacturer believes a network's reported sales lift is implausible.

[ ] Read the contract first for methodology, audit, and remedy provisions. Usually thin.
[ ] Ask whether a holdout group existed. Without one there is no incrementality measurement, whatever the deck says.
[ ] Ask when the methodology last changed, since a silent model change explains many implausible movements.
[ ] Assess the leverage honestly. The brand needs category access more than the network needs this brand, which makes disclosure, a verification commitment, and better renewal terms the realistic outcome.
[ ] Fix it for next time by negotiating the audit clause at the point of the next spend commitment.

The pixel demand letter

A retailer receives a demand alleging its site's tags intercepted communications.

[ ] Phase 9's inventory before anything else. The answer determines the entire response.
[ ] Separate the federal and state theories, since the state analogues with statutory damages carry the exposure.
[ ] Check the video question under 18 U.S.C. § 2710, because product video plus pixel is the pattern.
[ ] Remediate forward with a dated record while defending the past.
[ ] Extend the exercise to the app, which is usually worse and usually forgotten.
[ ] Set the recurring inventory cycle as part of the settlement of the internal problem, whatever happens to the claim.

The adjacent networks

Every business with a customer relationship and a transaction record is building one of these, and the framework transfers with the sensitivity of the data.

[ ] Travel and hospitality. Booking and stay data with a loyalty programme already in place. See Points, Flags, and Bookings, Running a Travel or Loyalty Brand, and the Travel, Hospitality, and Loyalty Programme Brand Toolkit.
[ ] Financial services. The richest signal and the tightest constraints, with 15 U.S.C. § 1681 live from the outset.
[ ] Pharmacy adjacencies. Health data changes the analysis in kind, and a general retailer with a pharmacy counter operates both regimes in one building.
[ ] Marketplaces and delivery. Transaction plus location plus a seller population that will ask the questions brands ask.
[ ] Connected vehicles. See the Connected Vehicle IP and Data Checklist.
[ ] Logistics. The data is about shippers rather than consumers, removing the privacy layer and leaving the rights questions. See Where the Box Went and the Logistics and Supply Chain Technology IP Toolkit.

What good looks like

[ ] The notice describes what the business does, and a shopper would recognise it.
[ ] An opt-out submitted anywhere reaches everywhere, provably, within the statutory window.
[ ] No audience exists that the suppression policy would have prohibited.
[ ] Somebody can state what every tag transmits, page by page.
[ ] Every number in the sales deck traces to a methodology document that supports it.
[ ] Every clean room arrangement states who owns the outputs, who owns the models, and what may be learned across brands.
[ ] The derived assets are enumerated and access-controlled, and nobody is pretending the transaction log is property.

Seven statements. The networks that can make all seven will still be operating in this form when the enforcement wave finishes.

A note on proportion

[ ] Scale to the product mix. A retailer selling only on-site placements on its own site needs Phases 1, 2, 3, 7, and 9, and can defer the rest.
[ ] Never defer Phase 2 where any product uses shopper data, because the notice is the document that makes everything else defensible or indefensible.
[ ] Never defer Phase 9, because the tag inventory is cheap, fast, and diagnostic, and the exposure it surfaces is the sector's most active litigation.
[ ] Treat Phase 6 as conditional on there actually being a clean room, and resist the vendor framing that one is required.
[ ] [Gate] Client and adviser have agreed in writing which phases are in scope and why the others are not.

Diligence: buying, selling, or investing in a network

Retail media businesses are acquired, carved out, and taken to investment committees, and the diligence questions are distinctive enough to be worth listing.

[ ] Ask for the notice history, not the current notice. Which version was live during each period determines whether the revenue in that period was lawfully earned.
[ ] Ask for the tag inventory and its date. Absence is itself a finding.
[ ] Ask for the request log with response times, since it is the only real evidence of whether the rights machinery works.
[ ] Ask how many claims and demands have been received on interception, video privacy, or marketing channel theories, and what happened to them.
[ ] Ask for the standard media agreement and the deviations from it. A network that negotiated every deal separately has a hundred different contamination positions.
[ ] Ask who owns the models, and trace each back to its training data provenance.
[ ] Ask for the measurement methodology and its version history, and compare the current sales deck against it.
[ ] Ask whether any brand has raised the private label conflict, and what was said in response.
[ ] Ask what happens to the loyalty relationship on a change of control, since the notice, the terms, and the shopper's expectation may all point different ways.
[ ] Price the remediation rather than the exposure, since in this sector the cost of fixing forward is usually knowable and the cost of past exposure is not. See the IP Due Diligence Toolkit.
[ ] [Gate] The buyer knows which revenue was earned under which notice.

Evidence habits worth building

[ ] Archive every notice version with its live dates. This single practice answers more questions in an inquiry than any other record.
[ ] Timestamp the rights request log, including the channel of receipt and the date of completion in every downstream system.
[ ] Keep the tag inventory under version control, so that additions are visible and attributable.
[ ] Log clean room queries and outputs with the configuration in force at the time.
[ ] Version the measurement methodology and keep the superseded versions, since brands will ask about historical periods.
[ ] Record the deployment gate decisions for in-store sensing, including what was rejected and why.
[ ] Keep the segment creation log, including segments proposed and suppressed, which is the evidence that the suppression policy operates rather than exists.
[ ] [Gate] Seven records exist, are maintained by named owners, and would survive a platform migration.

Conversations that recur, and how to answer them

[ ] "The platform said we are the controller." Test the characterisation against what the platform actually does with the data. A platform improving its own models on the retailer's audience is not a service provider, whatever the paperwork says, and the retailer should not carry a label that benefits only the counterparty.
[ ] "The clean room means we do not need consent." Architecture is not a legal basis. Put the correction in writing once, clearly, because this belief is widespread and is the one most likely to produce a finding.
[ ] "Our notice covers it — it says marketing purposes." It does not. Cross-context behavioural advertising, model training, and third-party collaboration are not what a shopper understands by that phrase.
[ ] "Nobody else in the market does this either." Industry practice is not a defence, and being early in a market about to be examined is a commercial advantage rather than a cost.
[ ] "We cannot delay the launch." Then sequence it: launch the on-site product, which is least exposed, and hold off-site until the notice is fixed. The business usually accepts a sequenced launch when the alternative is framed as launching the highest-risk product first.
[ ] "The brand agreed to it in the insertion order." An insertion order is not a data agreement, and a signature on a media buy does not settle contamination, retention, or model ownership.
[ ] "It is aggregated, so it is not personal data." Only above a threshold that somebody has set and somebody enforces. Ask what the threshold is and where it is configured.
[ ] [Gate] Each of these has been answered once, in writing, so it does not have to be answered again in a deposition.

Four questions to open every file with

[ ] What does the notice say, and does it match what you do? Phase 2.
[ ] What do your tags transmit, and to whom? Phase 9.
[ ] What is in your segments, and who checked before they were built? Phase 4.
[ ] What does the deck promise, and what document supports it? Phase 7.

Four questions, answerable in a fortnight, and between them they describe every way this business goes wrong.

[ ] And a fifth, for the physical estate: what is in the stores, what does it capture, and did anybody review it before it was installed? In-store is where the practice is thinnest, the questions are newest, and the answer is least likely to exist anywhere in the organisation already.
[ ] [Gate] All five have been asked of a person who could actually answer, and the answers are written down with dates.

The underlying discipline is the same throughout: in a business built on documents rather than property, the documents are the business, and a document nobody has read since it was drafted is a liability wearing the costume of a control.

That is the whole discipline of this checklist: it converts documents nobody has read into documents somebody owns, on a cycle, with a date.

Everything else in this sector — the technology, the measurement, the commercial negotiation — sits on top of that one habit.

Build the habit first, and the rest becomes ordinary practice rather than crisis management.

Key Authorities at a Glance

The foundation is negative: shopper data is not property. Feist Publications, Inc. v. Rural Telephone Service Co. denies copyright in facts, 17 U.S.C. § 103 leaves compilations thin, and 17 U.S.C. § 102(b) excludes the organising system. Contract, technical control, and trade secrecy under 18 U.S.C. § 1836 and § 1839 substitute for the missing property right.

The privacy layer supplies the binding constraints: sale and share definitions in the state comprehensive statutes, children's obligations at 15 U.S.C. § 6501 and 16 C.F.R. Part 312, marketing channels at 47 U.S.C. § 227 and 15 U.S.C. § 7701, and the eligibility trap at 15 U.S.C. § 1681. Private litigation runs on 18 U.S.C. § 2511 and 18 U.S.C. § 2710, with standing from Spokeo, Inc. v. Robins and TransUnion LLC v. Ramirez, against the backdrop of Sorrell v. IMS Health Inc..

Measurement claims are governed by 15 U.S.C. § 45 and 15 U.S.C. § 1125, with Lexmark International, Inc. v. Static Control Components, Inc. on standing and POM Wonderful LLC v. Coca-Cola Co. closing the compliance defence. Protecting the stack runs through 35 U.S.C. § 101, Alice Corp. v. CLS Bank International, Mayo Collaborative Services v. Prometheus Laboratories, Inc., and 15 U.S.C. § 1052.

| Authority | Phase | | --- | --- | | Feist Publications, Inc. v. Rural Telephone Service Co. | 1, 10 — no property in facts | | 17 U.S.C. § 102(b) | 10 — organising systems excluded | | 17 U.S.C. § 103 | 10 — thin compilation protection | | 18 U.S.C. § 1836 | 10 — trade secret claim for derived assets | | 18 U.S.C. § 1839 | 10 — reasonable measures | | 15 U.S.C. § 45 | 3, 7 — deception, on-site and business-to-business | | 15 U.S.C. § 6501 | 4 — children's data | | 16 C.F.R. Part 312 | 4 — the operative children's rule | | 47 U.S.C. § 227 | 4 — calls and texts | | 15 U.S.C. § 7701 | 4 — commercial email | | 15 U.S.C. § 1681 | 4 — the consumer report trap | | 18 U.S.C. § 2511 | 9 — interception theory | | 18 U.S.C. § 2710 | 9 — video privacy claims | | Spokeo, Inc. v. Robins | 9 — concrete harm | | TransUnion LLC v. Ramirez | 9 — standing limits in data cases | | Sorrell v. IMS Health Inc. | 9 — commercial data and the First Amendment | | 15 U.S.C. § 1125 | 7 — competitor claims on performance | | Lexmark International, Inc. v. Static Control Components, Inc. | 7 — who may sue | | POM Wonderful LLC v. Coca-Cola Co. | 7 — standards compliance is no defence | | 35 U.S.C. § 101 | 10 — eligibility of the technical stack | | Alice Corp. v. CLS Bank International | 10 — abstract idea screen | | Mayo Collaborative Services v. Prometheus Laboratories, Inc. | 10 — the two-step framework | | 15 U.S.C. § 1052 | 10 — registering the network brand | | 17 U.S.C. § 201 | 8 — ownership of campaign creative |

Further reading is collected at retail media data rights, universal opt-out signal, data clean room agreement, pixel wiretap litigation, and incrementality measurement claims.