top of page
AI processor with icons for security, growth, and marketing

7 eye-catchers in display advertising, i.e. how to create graphics that will catch the customer’s eye

Week in Digital Marketing 2026-10-01

Oct 1
8 min read

This week, the market moved from valuing visibility without clicks to competing for algorithmic selection and transaction execution capabilities. A week ago, the central challenge was linking AI exposure to revenue and taking ownership of first-party relationships as the funnel shortened. The current shift is more fundamental: being cited is no longer sufficient proof of advantage. A model may know a brand and mention it in its answers while still recommending a competitor. An agent, meanwhile, can compare offers, build a cart, and initiate a purchase without the user ever visiting the seller’s website.

A new layer of competition is emerging — the agentic shelf — where three to five machine-selected products replace the traditional results page, digital shelf, and much of the conventional funnel. For growth, commerce, and RevOps leaders, this requires an urgent reprioritization. Advantage no longer comes from content production alone, or even from a high citation share, but from the ability to provide diverse proof of value, current product data, secure transaction infrastructure, real-time data activation, and independent measurement of outcomes that platforms cannot attribute to themselves.


TL;DR


  • GEO is shifting from measuring presence to measuring preference. Recommendation share, primary choice rate, and selection context matter more than citation volume alone.

  • AI is becoming not only a discovery interface but also a funnel operator. Agentic commerce moves control over offer selection, merchandising, and the moment of purchase ahead of the seller’s website.

  • Product truth is becoming new marketing infrastructure. Catalog completeness, inventory status, delivery terms, variants, and data consistency directly affect whether a product makes it onto the agentic shelf.

  • Reputation operates as a distributed data model. Reviews, communities, research, media, and comparison sites not only increase visibility but also influence whether AI considers a brand the right choice in a specific situation.

  • Owned media are evolving from a retention channel into a source of proprietary facts, relationships, and signals. Trade publications, newsletters, and original research should build a public record for the brand that generic content cannot replace.

  • Agentic marketing raises the importance of activation speed and governance quality. Organizations that can safely shorten the time from signal to action, while holding automation accountable through incrementality and revenue recorded in backend systems, will gain an edge.


Summary of Changes


The most important change is the separation of visibility, credibility, and selection. Until now, the market has focused on whether a brand appears in AI answers, which sources cite it, and the sentiment surrounding its mentions. New data shows, however, that a brand can be well represented informationally and still lose commercially.

In LEOPRD’s analysis, brands overlooked in favor of competitors sometimes had more citations than brands selected as the primary recommendation. The difference was not the volume of evidence but its content: concrete validation of product capabilities, experience quality, and fit for the user’s situation. This means GEO teams must stop optimizing exclusively for citation share and build a recommendation analysis system. It should monitor which prompts result in the brand being selected, overlooked, or presented as risky.

LLM seeding cannot consist of massively increasing the number of mentions. It should be used to build a consistent and diverse public record of the brand across reviews, research, media, communities, and comparisons, because most of the material used by models sits outside owned channels.

The second shift concerns the move from the answer layer to the execution layer. AI-assisted commerce, Universal Commerce Protocol, Sponsored Agents, assistants from Kroger, Giant Eagle, and DoorDash, and Amazon’s conflict with Meta Muse show that platforms no longer want merely to answer shopping questions. They want to control the moment when intent turns into a shortlist, cart, offer, and payment.

Competition is shifting from the traditional SERP and retail media to the agentic shelf, where the model reduces thousands of products to a handful of options. The role of commerce infrastructure is changing as well. The website remains important for experience and conversion, but it is no longer the only entry point. The catalog, product feed, API, availability system, substitution policy, payment mechanism, and order fulfillment history are also becoming marketing surfaces. Growth is therefore no longer solely about traffic acquisition. It is becoming a question of data interoperability and controlled openness to external agents.

The third shift is taking place in marketing operations. As discovery and part of the persuasion process happen outside the observable click path, and agents can act on behalf of users and marketers, speed from signal to activation becomes an operational advantage.

A real-time CDP should not be assessed based on vendor claims, but on measurable event-to-profile and profile-to-destination latency. At the same time, campaign automation is evolving from executing rules to bounded autonomy — autonomy constrained by clearly defined rules, budgets, and permissions. The agent observes, decides, acts, and learns, but only within designated boundaries.

This requires a new division of responsibility. Machines can take over reversible micro-decisions, while people should retain control over capital allocation, brand policy, privacy, and decisions that are difficult to reverse. Measurement must keep pace with this change. Deterministic identification and platform attribution remain operationally useful, but they must not be confused with evidence of impact. Channels with weaker observability, such as CTV, premium media, or exposure in AI answers, should be assessed using control groups, holdout tests, brand lift, MMM, and commercial outcomes — rather than discounted simply because the existing analytics stack captures clicks more easily.


Patterns of Change


In recent weeks, the market has moved from the traffic economy to the brand representation economy, and then to the economy of delegated choice. Today’s visibility increasingly concerns two audiences: the human and the machine acting on their behalf. The human interprets narrative, experience, and trust. The machine compares structured facts, the distributed public record, availability, and the ability to execute the task.

This shifts optimization from a single channel to the entire business system. AEO, GEO, PR, e-commerce, retail media, CRM, logistics, and RevOps are becoming components of one selection architecture.

The market is moving toward a model in which the interface is increasingly easy to capture, while advantage accumulates in layers that are harder to copy: original data, reputation, customer knowledge, reliable product data, transaction permissions, fulfillment quality, and causal measurement.

Companies that continue to maximize publication volume, sessions, and attributed conversions will report more and more vanity metrics covering less and less of the real decision-making process. Winners will manage three contracts simultaneously: the informational contract with the model, the transactional contract with the agent, and the trust contract with the customer.


Thematic Groups


From citation to model selection


  • Generative engine optimization (GEO): How to win AI mentions This guide frames GEO around entity clarity, technical accessibility, standalone content fragments, and presence across multiple platforms. The key implication: monitor prompt-level share of voice, mention context, and sentiment alongside traditional SEO metrics.

  • Goodbye Clicks, Hello AI: Zero-Click Search Redefines Marketing Bain shows that zero-click search structurally constrains the non-branded discovery stage where consideration has traditionally been built. The response should be to measure AI impact and reach, develop topical authority, and invest in formats that work beyond the traditional SERP.

  • Management - AI search is changing visibility, but the fundamentals of search still matter AI expands the scope of brand discovery but does not eliminate the importance of relevance, experience quality, topical authority, and earned trust. The most resilient strategy combines SEO, PR, and experience-based content, rather than treating GEO as a standalone growth channel.

  • Which Third-Party Signals Matter Most for AI? Customer reviews, community recommendations, research citations, and presence in editorial content are rated more highly than industry awards. Operational reputation management and customer advocacy are becoming part of visibility infrastructure, rather than merely a communications function.

  • Study: Brands Are AI's First Choice Just 29% Of The Time, And More Citations Don't Help The study shows that presence and citation volume do not guarantee a recommendation. Brands need to measure their ability to turn visibility into preference and continuously manage the distributed public record around product capabilities, experience quality, and potential risk signals.


Agentic shelf and transaction infrastructure


  • Ecommerce Sales Strategy: A Phased Framework (2026) Unified commerce is presented as a shared data layer for DTC, retail, marketplaces, and agentic storefronts. High-quality traffic referred by AI increases the value of a consistent catalog, customer profile, and infrastructure that enables purchases to continue across channels.

  • Meta Muse: Amazon Blocked It, Brokers Lost $35bn Anyway Amazon’s conflict with Meta Muse reveals the battle for control over the intent layer, product advertising, and behavioral data. The likely direction is paid interfaces with controlled agent access, rather than unconditional openness or permanent blocking.

  • Automating The Massive Ad Pivot Sponsored Agents turn the chatbot from an advertising surface into a conversational interface for qualification, product exploration, and CRM activation. This requires incrementality controls and reconciliation of platform reporting with actual revenue and backend data.

  • Win AI search-powered shoppers in India Google brings AI Overviews, AI Mode, ads, Direct Offers, and Universal Commerce Protocol together in a single commercial ecosystem. Brands must ensure interoperability across product feeds, first-party signals, and measurement in journeys that span an increasing number of touchpoints.

  • How Kroger, Giant Eagle and DoorDash are building their agentic AI assistants Retailers are deploying agents that turn inspiration and context into carts, replenishment, and product substitutions. As a result, marketing, merchandising, e-commerce, loyalty, and operations must work from a shared data layer and a single customer experience model.


Owned audience and the content economy after referral erosion



Agentic RevOps, activation, and causal measurement


hello@1000i.pl

+48 503 04 1000

bottom of page