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 2026-10-08

Oct 8
9 min read

This week, the market moved from designing a presence on the agentic shelf to monetizing and operationalizing the entire decision interface. A week ago, the key question was whether a brand would be selected by an agent and whether its infrastructure could complete the transaction. Today, the question is who will fund, program, and measure that decision. Amazon, Google, and OpenAI are transforming advertising from a message that redirects users to a website into a conversational layer for recommendations, qualification, and purchase. At the same time, they are automating media planning, creative production, budgeting, and analytics. For growth and RevOps leaders, this means an urgent shift in the management model: campaigns will increasingly cease to be a collection of manually configured keywords, segments, and formats, and instead become a system of constraints, permissions, data, and economic goals handed over to agents. In this environment, the advantage no longer comes from AI implementation alone, but from the ability to govern autonomy through data quality, brand consistency, content rights, margin-protection rules, and incremental revenue measurement.


TL;DR


  • Advertising is becoming an interface. Google’s conversational formats, Amazon Branded Conversations, and Sponsored Agents can take over the roles of a landing page, product advisor, lead form, and part of the sales process.

  • Budgeting is shifting from campaign-structure management to demand steering. Rather than manually defining keywords and segments, teams will set profitability thresholds, SKU priorities, exclusions, brand rules, and autonomy boundaries.

  • Agentic commerce is not a universal way to reduce friction. It creates the greatest value where the cognitive cost of making a decision exceeds the cost of executing the transaction, especially in travel, finance, electronics, and complex B2B.

  • Data governance is becoming a revenue function. AI does not fix fragmented data; it simply turns observable proxies into seemingly certain budget decisions faster.

  • GEO is moving from an experiment to a media product, but it still primarily sells exposure and citation share. Without incrementality measurement and a connection to the pipeline, it can easily become another category of vanity metrics.

  • Owned audience is evolving into an identity and rights management system. Newsletters, registrations, apps, communities, and content-access controls support monetization, first-party signal collection, and negotiations with AI platforms at the same time.


Summary of changes


The most important change from last week is the entry of agents into the paid distribution system. Previously, we analyzed the agentic shelf as a limited shortlist of products generated by a model. Now, advertising itself is beginning to participate in creating that shortlist: it explains the choice, answers questions, qualifies the customer, builds the offer, and initiates the transaction. Google is testing sponsored responses, personalized explanations, and lead agents directly within the generative search experience. Amazon is combining Branded Conversations, Sponsored Services, full-funnel automation, and natural-language analytics into a single retail media operating system. OpenAI is developing a similar model through Sponsored Agents and integrations with CRM and commerce platforms. The line between organic AEO, ad creative, merchandising, sales, and customer service is starting to blur. For brands, this means that product data, inventory levels, pricing, delivery rules, knowledge bases, user consent, and CRM data are becoming the raw material from which the platform dynamically assembles a commercial response.

The second change is the shift from campaign automation to delegated capital allocation. In the demand-led budgeting model, the platform responds to detailed queries, sudden trends, and creator-driven demand, while the budget scales as long as the required return threshold is maintained. This fundamentally changes the work of marketing operations. A campaign is no longer a stable container for budget and targeting, but an ongoing decision-making process in which humans define the objective function and operating boundaries. The problem is that the agent optimizes only the reality represented in the data. If the CRM rewards low-cost MQLs, identity resolution is incomplete, and the influence of brand, dark social, and recommendations remains invisible, the system will fund the easiest channels to observe with increasing confidence. This creates a bad information doom loop: a flawed proxy steers the budget, the algorithm’s output seemingly confirms its validity, and the next iteration increases the deviation from real incrementality. AI readiness should therefore be measured by data governance maturity, data validation before records are written to the CRM, lead handoff quality, consistency in funnel definitions, and the ability to send SQL, opportunity, and closed-won signals back to platforms.

The third change concerns the economics of content. Last week, owned media primarily served as a defense against the erosion of referral traffic and as a source of proprietary facts for models. Now, they are also becoming a layer for rights management, identity, and monetization. Publishers are testing controlled crawler access, registrations, licensing, and commercial GEO products. YouTube, in turn, is developing membership and community mechanisms that turn platform audiences into relationships resembling an owned audience. At the same time, the dismissal of the Chegg and Penske lawsuits weakens the assumption that indexing constitutes an implied exchange of content for traffic. In practice, publishers and brands can no longer assume that visibility will automatically generate a session. They must consciously decide which part of their knowledge to make available for indexing and LLM seeding, which part to use as an incentive for registration, and which part to protect as a proprietary data asset.


Patterns of change


The market is moving from the economics of delegated choice to the economics of delegated management. An agent does not merely select a product for the consumer. Increasingly, it also selects audiences, creative, channels, budgets, and activation timing for the advertiser. Systems on both sides of the transaction are therefore optimizing according to different objective functions. The customer’s agent seeks fit, price, and risk reduction. The platform’s agent maximizes interface utility and advertising revenue. The marketer’s agent pursues ROAS, pipeline, or margin objectives. Competitive advantage will depend on who designs the rules governing interactions between these systems most precisely.

Marketing is moving toward a model in which execution becomes cheaper while orchestration becomes more valuable. Variant production, media buying, and reporting are becoming platform functions, but positioning, original data, governance, experiment design, and accountability for outcomes remain within the organization. This increases the importance of marketing engineering: a capability that combines strategy, RevOps, data, automation, creative, and measurement architecture. Organizations that fail to organize this layer will not achieve lasting productivity growth through AI. They will only gain greater production scale, more automatically generated decisions, and faster error accumulation.


Topic clusters


Advertising as an interface and agent-driven demand steering



RevOps, data governance, and the automatic-proxy trap



Owned audience, access control, and content rights



Trust, discovery, and brand visibility in AI


  • How to get B2B brand back in the budget Brand is nominally returning to B2B strategy, but it still loses budget to performance marketing because the latter is easier to report on. Visibility in LLMs strengthens the business case for consistent positioning, proof of value, and distinctive assets, because models assess the clarity of the entire brand entity.

  • If Consumers Trust Friends More Than Brands, What Can Communicators Do? Private recommendations and group conversations cannot be targeted directly, but their likelihood can be increased through experience quality, reviews, and credible proof. Earned media, communities, and customer advocacy build a validation layer used by both people and AI models.

  • Social search and AI visibility: 2026 report data Discovery is fragmenting across specialized interfaces: social media inspires, search engines support research, Reddit provides social validation, and AI compares options and creates shortlists. This requires a shared social, SEO, and GEO system built around customer questions, reviews, product data, and a single content map.

  • Media Briefing: Publishers are turning GEO from an experiment into a business Publishers are turning editorial authority and AI visibility into a paid extension of branded content. Commercial demand is growing faster than measurement standards, which is why citations and share of answers remain temporary proxies without a clear connection to sales.

  • Generative Engine Optimization: Growth Strategies and Metrics For the AI Era Ahrefs frames GEO as an extension of SEO encompassing earned media, reviews, communities, and a multi-platform brand presence. The most durable strategy combines clear content structure with external mentions and measurement of AI share of voice, citations, and traffic quality.


GEO as a media product and a content-asset portfolio


  • Fighting traffic declines: Focus on owned channels, GEO Real-time reporting, verified local facts, and community trust are harder to copy than generic explanatory content. GEO should direct users into a recirculation system of registrations, newsletters, events, and subscriptions, rather than serve solely to win citations.

  • Why Existing Content May Be More Valuable Than What Is Published Next Click erosion undermines the equation in which producing more content automatically increases traffic. The archive should be treated as a portfolio of visibility assets: selectively refresh evergreen content, remove cannibalization, and assess conversion potential.

  • Since 2024: GEO and AEO Search Data, Two-Year Shift | Web Tonic™ The overview points to growth in zero-click searches, broader reach for AI Overviews, and continued CTR erosion despite AI referral traffic still accounting for a small share. Reporting should separate rankings, citations, share of answers, and conversion quality rather than combine data derived from different methodologies.

  • Explaining GEO The analysis organizes the mechanics of retrieval, query fan-out, answer synthesis, and measurement, while also showing the limited quality of evidence behind many GEO tactics. Crawler accessibility, source authority, original information, and controlled experiments remain safer investments.

  • Google Wins Dismissal of Lawsuits About AI Search Impact on Website Traffic The dismissal of the Chegg and Penske lawsuits challenges the assumption that indexing creates an implied exchange of content for traffic. Organizations dependent on referral-traffic monetization must build their own audience and conversion value, as legal protection for the historical click economy remains limited.

hello@1000i.pl

+48 503 04 1000

bottom of page