The 2025 edition of Google Marketing Live Has highlighted a discreet but decisive transformation of Google’s advertising ecosystem. Through the integration of the Gemini 1.5 Pro model in Google Ads tools, the Mountain View firm accelerates the automation of campaigns, by deep reorganizing the relationship between data, creation and performance.
The progressive integration of the Gemini 1.5 Pro model transforms the way in which advertising campaigns are created, activated, optimized and measured in Google Ads. At the crossroads of generative artificial intelligence, dynamic personalization and multimodal research, Google redefines the entire marketing tunnel, from intention to conversion.
Technological unification around Gemini
The Gemini 1.5 Pro model works as a transverse base capable of processing texts, images, videos and vocal signals. It is based on a contextual window extended to a million tokens, authorizing the analysis of complex corpus (product descriptions, purchasing historical, creative content, visual constraints) to produce recommendations, generate visuals, or adapt formats to different channels.
This common infrastructure makes it possible to converge creation, targeting, diffusion and measurement in a unified logic, controlled by AI.
Advertising auction managed by the margin
The Max Performance Platform, so far focused on volume optimization (clicks, visits, conversions), now incorporates financial logics. Advertisers can transmit the unit margin By product, which allows the algorithm to prioritize the most profitable articles, not only the most clicked.
This development slides the advertising strategy towards an oriented approach profitabilityand brings marketing, finance and data closer to the management of campaigns.
Automated creation, but in accordance with the brand
Generative AI does not only intervene in activation. It is also mobilized in the production of visual content. The brands can submit their graphic charter, their typographies, color palettes or iconographic references so that the engine produces variations in accordance with the brand’s codes.
This process makes it possible to generate, on a large scale, contextualized visual assets, personalized and compatible with the creative requirements of advertisers, without systematic recourse to studios.
An advertising integrated into generative research
With the launch ofAi Overviews In the search engine, advertisements no longer appear only at the top or bottom of the results, but can be integrated directly into the response blocks produced by the AI. Unlike the first SGE type experiments, which excluded the announcements, the new approach combines them with the content generated.
This integration logic aims to extend the user route while maintaining a contextual, discreet, but commercially effective contextual advertising environment.
Immersive shopping, 3D formats and virtual try-on
Google also enriches its shopping formats with immersive features. It becomes possible to visualize a garment on different types of morphologies without multiplying the photo shoots. The system adapts the images according to the size of the user, with a realistic rendering generated automatically.
In addition, product images can be transformed into interactive 3D formats from simple photos, paving the way for visualization experiences enriched on mobile as well as on desktop.
A new role for agencies
This new model redistributes roles in the value chain. The agencies are no longer content to manage media plans. They are led to intervene on three inseparable dimensions: creation,, activationAnd the measure. AI requires a logic of convergence, in which visual content, distribution and performance must be piloted in synergy.
These change also impact the economic model of agencies, which evolve from an indexed remuneration on media purchase volumes to a hybrid logic combining fees and value sharing.
Three technical and organizational pillars
The transformation is based on three additional foundations:
- Data quality (stock, inventory, timetables, price)
- Technical integration via API (Merchant Center, Maps, CRM)
- The rise in human skillswith an increased requirement for training, transversality and governance.
The role of marketing departments is redefined. The orchestration of the IA levers requires breaking the internal silos, collaborating with CIOs, data teams, and anticipating the regulatory and ethical issues linked to large -scale algorithmic activation.
Google’s advertising system is taking a new strategic step.
The deployment of Gemini in the Google advertising environment is not limited to a functional improvement, it is a switch to a system entirely managed by data, margin and creative automation. The AI does not replace the marketing professions, but redefine its contours, tools, and arbitrations. For brands as for their partners, the issue is now to fit into this new technical order without losing their strategic coherence, their ability to adapt and their singularity.
