WritingReading path
Programmatic, Marketing Mix Modeling (MMM), attribution, embeddings, and the data infrastructure beneath modern advertising — from DMPs and OpenRTB to NNN and Transformer-based measurement.
28 essays in this path
of 95 published · 2015–2026
Google's NNN brings Transformer self-attention to Marketing Mix Modeling — moving past Meridian's parametric assumptions to learn temporal dependencies, cross-channel synergies, and creative nuance directly from aggregated data. ~22% lift in predictive accuracy on early benchmarks.
By capturing long-term effects, creative nuances, and channel interactions within one framework, NNN offers advertisers a clearer, more actionable view of their media mix.
Advertisers already own the cheapest hedge in media: the pause button. So the open web's first real performance insurance will protect the sell side, not the buy side — the publisher's receivable, the intermediary's spread, the counter's restatement — assembled from narrower protections around payment, counting and residual performance risk rather than one policy. The near-term contract names its counting source, escrows the money and bonds the count. The parametric policy on the landlord event itself comes last, once the loss history exists.
AppLovin and Criteo show why measuring an outcome is easier than putting a balance sheet behind the miss. The open web can define an outcome, measure it and settle a campaign. What it cannot do is guarantee that outcome across company boundaries, and neither, it turns out, can the walled gardens: AppLovin's 10-K says no fixed price per action, and Meta's own documentation says the cost cap is not guaranteed. AppLovin's advantage is a closed loop that lets one company observe performance, adjust price and keep the spread. Criteo came closest to carrying real risk on the web, buying impressions and selling clicks, and its loop depended on identifiers controlled by browsers. The open web is not dead. It still lacks a credible counterparty behind the miss.
This site's ask box declined a question on how to measure DOOH and place-based media eight times, in what turned out to be its own test runs. The answer: a DOOH number is one near-observed event, one auditable claim, and three models stacked on top — a chain the channel has been honest about since 1933. Run the Instrumentation Test in reverse and the punchline lands: the channel everyone calls unmeasurable is the one where reach and frequency still mean something.
Curation went from an SSP side product to the organizing fight of programmatic in under two years: every exchange has a named platform, the data specialists got absorbed by credit bureaus and holdcos, retail media and CTV fused into curated marketplaces, and now the agencies are building marketplaces of their own — Stagwell's, announced this week, assembles bespoke supply with AI agents fed by The Trade Desk's own quality data. This is the audience economy reorganizing itself on the supply side. The question is not whether the packaging adds value. It is who signs the package — and what happens to a shelf of pre-cut deals when buyers gain an agent that can assemble supply per brief.
An AdExchanger column just named the shift the whole industry is feeling: from an audience economy to a context economy. It is right, and it is the direction a season of this site's writing has been pointing. But 'understand the moment and act on it instantly' is one governance layer away from being a faster surveillance economy — and that layer is the whole ballgame.
Amazon opened AMC to all sponsored-ads advertisers in Sept 2025. What's free, what still costs, what the 100-user threshold hides, and what expires Dec 2026.
People search 'CDP vs. reverse ETL' like it's a product bake-off. It's a category error — one is a suite, the other is a pipe. The real question is where the customer truth should live, and once you answer that, the comparison resolves itself. A decision table, three scenarios, and the trap on each side.
The same campaign, 54 defensible iROAS values, 83% flipping sign on methodology alone. Eight audit questions — and the failure mode each one catches.
AdCP's sell side is a deal machine beside the wrapper, not inside it — the architecture, the deal loop, five publisher decisions, the caveats nobody prints.
The W3C Attribution API debate, read as a constitutional question. Mozilla's Martin Thomson says the quiet part precisely: the test of any privacy technology is not the information flows — it is who decides. The browser is bidding to become the counterparty that grants measurement its standing, and every operator concern in the spec follows from that one move.
A framework for the warehouse-native generation of customer data: what 'no data egress' actually buys, why reverse ETL became the last mile, how multi-cloud clean rooms split the work, and where the lock-in really went.
AdTech platforms now own more capability than ever. The winners will be the ones that turn capability into one buyable story through commercial architecture.
AR/VR, Web3, regulation, and generational power shifts. The five-year outlook for the ad industry.
The strategic read on App Tracking Transparency — Apple's privacy posture as competitive intelligence.
Ten transformative trends — shoppable TV, retail media × streaming, data-driven decisioning — for 2024 planning.
The post-cookie playbook — zero-party data plus clean rooms as the durable identity stack.
Secure multi-party data environments — the substrate for targeting, attribution, and personalization without raw PII.
Decentralization, authenticity, transparency, and privacy — how Web3 primitives reshape brand-consumer trust.
Purpose, People, Platform, Process, Performance — Paul Roetzer's five-P framework applied to programmatic.
How proactively-shared customer data fuels personalization, product roadmap, and durable trust.
When hundreds of LinkedIn job titles carry the word, it is time to define it. Addressability is the capacity for an entity to be targeted and found: the $100B+ of post-cookie open-web spend that depends on it, the three-letter acronyms that run it, and the four addressability paths and which ones survive 2023.
The four pillars of the marketing loop — Targeting, Delivery, Measurement, Optimization — rebuilt for the post-cookie era.
Five practical tips — DMPs, third-party data, segmentation, look-alikes, data-driven action — for profitable consumer relationships.
Decoding two over-used terms — and why the DMP is the central nervous system of an interoperable mar-tech stack.
How DMPs power PMP packages, custom audiences, private data marketplaces, and the four scenarios that matter.
Why the DMP is the foundation for multi-touch attribution and the connective tissue across marketing channels.
Deterministic vs probabilistic matching. The eight questions to ask every cross-device vendor before signing.
4 research papers · 7 playbooks · 1 appearance
What buyers can receive, keep, use again and rebuild from nine data-collaboration products
4 essays on this path
The thinking on this page shows up in the vertical playbooks below — abstracted from real client engagements.
From Data to Impact: How Martech Drives Real Business Growth
Watch the panelAsk about the essays, frameworks, playbooks, or how to work with Evgeny.