# Measurement & Attribution — Position Record What survives as advertising’s decision-makers stop being human — which metrics break, which are rebuilt around privacy, why counting a conversion still isn’t the same as causing one, and what happens once an agent is handed a single number to optimize and learns to fake it, the Scalarization Trap. > Markdown twin of https://nofluffadvisory.com/thinking/measurement-attribution/ — the Position Record for Measurement + Attribution + Agentic optimization. Generated from the verified position ledger; methodology: https://nofluffadvisory.com/thinking/ ## Current positions **Measurement** — Measurement is advertising's most institutionally mature standard, yet it is friction, not load-bearing infrastructure: a deal can still clear without touching any of it. > “The infrastructure makes the deal more defensible. It doesn't make the deal possible in the first place” — [The One Standard That Clears the Deal](https://nofluffadvisory.com/writing/the-standard-that-clears-the-deal/), Jul 31, 2026 **Attribution** — Even flawless conversion counting across platforms does not establish causation; deduplication solves the counting problem while leaving the incrementality question untouched. > “the harder question underneath all of it: even a perfect count doesn't tell you what actually caused it” — [One Event, Three Machines: Orchestrating Conversions Across PMax, Advantage+, and OpenAI Ads](https://nofluffadvisory.com/writing/one-event-three-machines/), Jul 23, 2026 **Agentic optimization** — Which layer of agent (copilot, platform optimizer, or agent-to-agent negotiator) is executing a spend decision shouldn't change where accountability lands — it lands on whoever authored the objective function, not on the layer. > “The poll left "agents" unspecified on purpose, because the answer shouldn't depend on the layer — whatever executes, the objective function is where accountability lands.” — [Business Outcomes Isn't a Number](https://nofluffadvisory.com/writing/business-outcomes-isnt-a-number/), Jul 24, 2026 25 positions tracked · held since Jul 31, 2026 · record spans Sep 18, 2015 – Jul 31, 2026 **Verified tensions:** none currently recorded between the positions on this page. That is a statement about what the ledger has verified — not the result of an exhaustive contradiction analysis. ## Coverage gap, disclosed Every verified position below is either still holding or was confirmed by a later event — no retracted position yet. That's not because Evgeny's thinking never reverses or leaves a question open elsewhere on this site; this topic's own record simply hasn't reached one yet. Rather than manufacture a loss or an open call to look complete, this page stays visibly partial until a real one earns its place. ## The timeline — newest first ### Jul 31, 2026 — Framework — Current · Measurement Measurement is advertising's most institutionally mature standard, yet it is friction, not load-bearing infrastructure: a deal can still clear without touching any of it. > “The infrastructure makes the deal more defensible. It doesn't make the deal possible in the first place” — [The One Standard That Clears the Deal](https://nofluffadvisory.com/writing/the-standard-that-clears-the-deal/), Jul 31, 2026 - Status: current — no change since Jul 31, 2026. ### Jul 24, 2026 — Coinage — Current · Measurement Any single-scalar optimization metric handed to an agent gets gamed, because a scalar's cheapest route up is faking it; Business Outcomes resists this precisely because it cannot be scalarized. > “Call the failure the Scalarization Trap — the moment you collapse "what we actually want" into one number an agent can watch and chase, you’ve handed it a target that’s cheaper to fake than to earn” — [Business Outcomes Isn't a Number](https://nofluffadvisory.com/writing/business-outcomes-isnt-a-number/), Jul 24, 2026 - Status: current — no change since Jul 24, 2026. ### Jul 24, 2026 — Claim — Current · Agentic optimization Keeping Business Outcomes outside an agent's loss function defeats real-time metric-gaming, but it does nothing about the deeper problem of an agent's owner grading its own agent's outcome with no independent party checking the grade. > “It does nothing about the owner grading their own agent's homework after the fact, with no independent party checking the grade.” — [Business Outcomes Isn't a Number](https://nofluffadvisory.com/writing/business-outcomes-isnt-a-number/), Jul 24, 2026 - Status: current — no change since Jul 24, 2026. ### Jul 24, 2026 — Claim — Current · Agentic optimization The instinct that makes Business Outcomes resist scalarization also shows up at company scale, in Perplexity's and Anthropic's refusals to let advertiser incentives inside the loop that mediates what an AI assistant tells users. > “keep the thing you actually care about — trust, in their case — outside the number the system is optimizing, because the moment it's inside, it's for sale to whoever prices it best” — [Business Outcomes Isn't a Number](https://nofluffadvisory.com/writing/business-outcomes-isnt-a-number/), Jul 24, 2026 - Status: current — no change since Jul 24, 2026. ### Jul 24, 2026 — Call — Unresolved · Agentic optimization Ev's Week 7 ballot: Business Outcomes, reasoned mechanically — it structurally resists the Scalarization Trap because it isn't a live number an agent can see and chase, not because it's a nobler goal. > “I voted Business Outcomes, and the reasoning is the mechanical one above, not the values-based one.” — [Business Outcomes Isn't a Number](https://nofluffadvisory.com/writing/business-outcomes-isnt-a-number/), Jul 24, 2026 - Status: unresolved — a falsifiable call awaiting an event. When it resolves, it becomes a confirmation or a retraction here. ### Jul 24, 2026 — Claim — Current · Agentic optimization Which layer of agent (copilot, platform optimizer, or agent-to-agent negotiator) is executing a spend decision shouldn't change where accountability lands — it lands on whoever authored the objective function, not on the layer. > “The poll left "agents" unspecified on purpose, because the answer shouldn't depend on the layer — whatever executes, the objective function is where accountability lands.” — [Business Outcomes Isn't a Number](https://nofluffadvisory.com/writing/business-outcomes-isnt-a-number/), Jul 24, 2026 - Status: current — no change since Jul 24, 2026. ### Jul 23, 2026 — Claim — Current · Attribution Because no platform will dedupe against another's identifier, a first-party hub capturing each platform's click identifiers is the only architecture that reconciles conversions and routes down-funnel value back to each platform. > “Each garden is built to be the definitive ledger of its own traffic, and refusing to reconcile against a competitor’s ledger isn’t an oversight — it’s the business model.” — [One Event, Three Machines: Orchestrating Conversions Across PMax, Advantage+, and OpenAI Ads](https://nofluffadvisory.com/writing/one-event-three-machines/), Jul 23, 2026 - Status: current — no change since Jul 23, 2026. ### Jul 23, 2026 — Claim — Current · Attribution Even flawless conversion counting across platforms does not establish causation; deduplication solves the counting problem while leaving the incrementality question untouched. > “the harder question underneath all of it: even a perfect count doesn't tell you what actually caused it” — [One Event, Three Machines: Orchestrating Conversions Across PMax, Advantage+, and OpenAI Ads](https://nofluffadvisory.com/writing/one-event-three-machines/), Jul 23, 2026 - Status: current — no change since Jul 23, 2026. ### Jul 23, 2026 — Claim — Current · Agentic optimization A live campaign confirmed the Scalarization Trap outside the AI-shopping-agent context: PMax, handed a conversion action that barely fired, optimized toward whatever cheap inventory satisfied the thin signal rather than toward real outcomes. > “The machine did roughly what I told it to: find the cheapest way to look busy against a conversion event that essentially never happened.” — [One Event, Three Machines: Orchestrating Conversions Across PMax, Advantage+, and OpenAI Ads](https://nofluffadvisory.com/writing/one-event-three-machines/), Jul 23, 2026 - Status: current — no change since Jul 23, 2026. ### Jul 23, 2026 — Claim — Current · Agentic optimization The training signal fed to a bidding algorithm has to be actively protected, because every unfiltered spam lead that reaches the conversion event is a poisoned training example teaching the optimizer that bots are customers. > “Every spam lead that reaches your conversion event is a poisoned training example — you are teaching the machine that bots are your customers.” — [One Event, Three Machines: Orchestrating Conversions Across PMax, Advantage+, and OpenAI Ads](https://nofluffadvisory.com/writing/one-event-three-machines/), Jul 23, 2026 - Status: current — no change since Jul 23, 2026. ### Jul 13, 2026 — Framework — Current · Measurement Under agentic commerce a metric breaks if it presumes a human eye or browser session, survives if it observes money or causality, and measurement authority migrates to whoever authenticates the agent. > “a metric breaks if it presumes a human eye or a browser session; it survives if it observes money or causality; what emerges is whatever must observe the agent itself” — [Measurement for Agentic Commerce](https://nofluffadvisory.com/writing/measurement-for-agentic-commerce/), Jul 13, 2026 - Status: current — no change since Jul 13, 2026. ### Jul 13, 2026 — Claim — Current · Attribution Browser-native, differentially-private attribution upgrades a number's provenance but not its logic — it still assigns credit, not cause, while relocating measurement's rules into infrastructure no buyer signs. > “The privacy engineering upgrades the provenance of the claim, not its logic. Credit is not cause, even when a browser signs the credit.” — [Measurement, on the Browser's Terms](https://nofluffadvisory.com/writing/measurement-on-the-browsers-terms/), Jul 13, 2026 - Status: current — no change since Jul 13, 2026. ### Jul 13, 2026 — Claim — Current · Measurement Retail media iROAS is a negotiated artifact, not a measurement: defensible methodology choices alone flip most campaigns between profitable and money-losing, so buyers must audit the methodology negotiation. > “The same campaign, the same spend, the same shoppers — profitable under one defensible recipe, money-losing under another.” — [iROAS Is Not a Number, It's a Negotiation](https://nofluffadvisory.com/writing/iroas-is-not-a-number/), Jul 13, 2026 - Status: current — no change since Jul 13, 2026. ### Jul 13, 2026 — Coinage — Current · Measurement Ev coins "agent share of choice" — the successor to Share of Model, measuring not whether a brand is mentioned by an AI but whether it was actually selected or defaulted to in a completed agentic purchase. > “No standards body or vendor has standardized that term as of July 2026; call it agent share of choice — coined here, labeled as such.” — [Measurement for Agentic Commerce](https://nofluffadvisory.com/writing/measurement-for-agentic-commerce/), Jul 13, 2026 - Coins the term **“agent share of choice”** — supersedes “Share of Model” as the operative term. - Status: current — no change since Jul 13, 2026. ### Jul 13, 2026 — Claim — Current · Measurement Under agentic commerce, measurement authority is shifting from whoever watches the shopper to whoever cryptographically authenticates the agent — the same kind of power shift that made panel ownership decisive in the human web. > “measurement authority is migrating from the parties that watch the shopper to the parties that authenticate the agent. In the human web, whoever owned the panel owned the truth. In the agentic web, whoever verifies the signature does.” — [Measurement for Agentic Commerce](https://nofluffadvisory.com/writing/measurement-for-agentic-commerce/), Jul 13, 2026 - Status: current — no change since Jul 13, 2026. ### Jul 10, 2026 — Framework — Current · Measurement When agents become the primary buyers, inference metrics (Attention first, then Reach & Frequency and Attribution) lose their referent because the decision path becomes directly readable; ROAS survives by accident. > “a metric survives agentic decisioning only if the event it measures remains unobservable in the system now making the decision — inference has no work left to do once you can just read the record” — [The Metric That Was Never About You](https://nofluffadvisory.com/writing/which-metric-breaks-first/), Jul 10, 2026 - Status: current — no change since Jul 10, 2026. ### Jul 10, 2026 — Coinage — Current · Measurement The instrument that replaces broken inference metrics is the Delegate Audit Record: a structured trace of the agent's spend decision — constraints, alternatives considered, predicted vs actual outcome. > “a structured trace of why an agent made a specific spend decision — its constraints, the alternatives it considered, and its predicted outcome measured against the actual one” — [The Metric That Was Never About You](https://nofluffadvisory.com/writing/which-metric-breaks-first/), Jul 10, 2026 - Status: current — no change since Jul 10, 2026. ### Dec 1, 2025 — Call — Current · Confirmed · Measurement Campaign reporting dies as a category, replaced by live diagnosis with human-in-the-loop resolution. > “Campaign reporting dies, replaced by live diagnosis with human-in-the-loop resolution.” — [Part 4: Why Your Campaign Reports Are Already Obsolete](https://nofluffadvisory.com/writing/part-4-why-your-campaign-reports-are-already-obsolete/), Dec 1, 2025 - Confirmed: Snowflake x Meta governed loop — Jul 21, 2026 — “The thesis confirmed — bounded autonomy on governed data, human gate — but shipped inside one walled garden” - Status: current — no change since Dec 1, 2025. ### Dec 1, 2025 — Claim — Superseded (refined) · Attribution Attribution as practiced is theater — last-click and multi-touch models distribute credit by arbitrary rules with no basis in causation, and the industry knowingly made million-dollar decisions on it. > “We made million-dollar decisions based on correlation masquerading as causation.” — [Part 4: Why Your Campaign Reports Are Already Obsolete](https://nofluffadvisory.com/writing/part-4-why-your-campaign-reports-are-already-obsolete/), Dec 1, 2025 - Refined by: “Browser-native, differentially-private attribution upgrades a number's provenance but not its logic — it still assigns credit, not cause, while relocating measurement's rules into infrastructure no buyer signs.” (Jul 13, 2026) ### Apr 21, 2025 — Claim — Current · Measurement Transformer-based MMM (Google's NNN) supersedes parametric Bayesian MMM (Meridian) by learning temporal carryover and cross-channel synergy directly from aggregate data, keeping measurement privacy-safe post-cookie. > “NNN signals a transformative shift in how we measure and optimize media in a privacy-first, post-cookie world” — [From Meridian to NNN: How Transformers Are Redefining Marketing Mix Modeling](https://nofluffadvisory.com/writing/from-meridian-to-nnn-how-transformers-are-redefining-marketing-mix-modeling/), Apr 21, 2025 - Status: current — no change since Apr 21, 2025. ### Apr 21, 2025 — Claim — Current · Measurement Attention is not a spotlight but a relevance engine: paying attention is the act of filtering an overwhelming world for what might matter, which is why equal attention produces uneven outcomes across contexts. > “attention is not a spotlight — it's a relevance engine.” — [Relevance Realization: The Cognitive Science of Attention in the Age of AGI and Advertising](https://nofluffadvisory.com/writing/relevance-realization-the-cognitive-science-of-attention-in-the-age-of-agi-and-advertising/), Apr 21, 2025 - Status: current — no change since Apr 21, 2025. ### Apr 7, 2025 — Framework — Current · Measurement Attention metrics alone do not predict marketing outcomes; measurement needs a relevance layer — the EARO framework's 22 metrics across Exposure, Attention, Relevance, Outcome. > “Traditional metrics (impressions, CTRs) can’t answer these questions. They tell you what happened, not why.” — [Attention Isn’t Enough—The Future of Marketing is Relevance](https://nofluffadvisory.com/writing/attention-isn-t-enough-the-future-of-marketing-is-relevance/), Apr 7, 2025 - Status: current — no change since Apr 7, 2025. ### Apr 7, 2025 — Coinage — Current · Measurement Coins EARO — Exposure, Attention, Relevance, Outcome — an open-source framework for measuring marketing effectiveness holistically, built on the premise that attention alone is not predictive. > “I'm building EARO (pronounced "Hero") — an open-source framework to measure marketing effectiveness holistically.” — [Attention Isn’t Enough—The Future of Marketing is Relevance](https://nofluffadvisory.com/writing/attention-isn-t-enough-the-future-of-marketing-is-relevance/), Apr 7, 2025 - Coins the term **“EARO”**. - Status: current — no change since Apr 7, 2025. ### Aug 2, 2021 — Claim — Current · Measurement Cookie loss does not kill measurement's pillars but rebuilds each around privacy: 1:1 attribution shrinks to samples while clean rooms, lift analysis, and MMM/Markov modeling carry the load. > “measurement will have to leverage multiple solutions to make up for the loss of cookie tracking and the ability to deconstruct customer journeys” — [Marketing is Dead, Long live the Marketing](https://nofluffadvisory.com/writing/marketing-is-dead-long-live-the-marketing/), Aug 2, 2021 - Status: current — no change since Aug 2, 2021. ### Sep 18, 2015 — Claim — Superseded (refined) · Attribution Attribution is fundamentally a data-centralization problem: unify first-, second-, and third-party data in a DMP and the full conversion path becomes visible and modelable. > “Understanding the entire conversion path across the whole marketing mix eliminates the accuracy challenge of analyzing data from siloed channels” — [Data is the New Currency to power Attribution](https://nofluffadvisory.com/writing/data-is-the-new-currency-to-power-attribution/), Sep 18, 2015 - Refined by: “Cookie loss does not kill measurement's pillars but rebuilds each around privacy: 1:1 attribution shrinks to samples while clean rooms, lift analysis, and MMM/Markov modeling carry the load.” (Aug 2, 2021) ## Where the verified record ends The verified record for Measurement + Attribution + Agentic optimization currently ends at **Jul 31, 2026**. 3 further rows are in the ledger as unverified candidates and are not shown — candidacy is not evidence. Nothing beyond that date, and nothing outside these verified rows, is claimed here. --- 25 of 111 total verified ledger rows shown here, filtered to Measurement + Attribution + Agentic optimization. Built against corpus revision `d5aebe0` — every row passed `verification.state === 'verified'` before it was allowed to render. Receipts are verbatim; superseded and retracted rows are marked, never presented as current. Methodology: https://nofluffadvisory.com/thinking/