Generative AI and the efficiency crisis: the new watershed moment in marketing.

In-depth Analysis and Contextualization

In recent days, three movements have intertwined and made it clear that marketing has entered a new cycle: the consolidation of generative AI in the daily lives of Brazilians, the efficiency crisis of paid media, and the radical reconfiguration of the discovery journey, previously dominated by traditional search. The feeling that it’s possible to continue doing more of the same with a little automation has fallen apart; the game has moved to a different board.

On one hand, virtually every Brazilian CMO has already decided to double down on artificial intelligence. According to the study The Reality of Marketing in Brazil 2025, 97.9% of leaders intend to increase the use of AI in their strategies this year, and 95.4% already see a direct ROI gain in AI-supported initiatives, with 72.3% reporting a significant improvement in the quality of deliverables[1]. At the same time, the ICT Households 2025 survey shows that generative AI is already part of the routine of 32% of Brazilians, which means that consumers have also changed the way they seek information, compare products and make decisions[5].

This meeting of supply and demand mediated by algorithms reshapes the very logic of digital presence. Recent reports indicate that features such as AI summaries within search engines reduce clicks to external websites by up to 34.5% in certain types of searches[1]. In parallel, Brazilian digital marketing faces rising advertising costs and falling conversions, shifting the focus from mass acquisition to retention and operational efficiency[2]. It is in this tension that the new power map between brands, platforms and consumers is being drawn.

Data, Cases and Examples

In the field of performance, the scenario is clear: the old playbook has become too expensive. Data compiled by ecosystem players shows that the average CPM on Facebook reached R$ 51 in September 2024, while the average conversion rate of Brazilian e-commerce fell to 1.92%[2]. Even so, the sector is expected to generate R$ 224.7 billion in revenue in 2025, a 10% increase according to ABComm[2]. Growth remains possible; growth with margin has become the central problem.

To protect profitability, the market is accelerating the adoption of data-driven loyalty tools and intelligent automation. Companies like Smartbis report increased demand for relationship programs driven by real purchasing behavior, while AI-powered messaging platforms integrated with WhatsApp are orchestrating personalized journeys on a highly engaging channel[2]. Metrics such as LTV, repurchase rate, and engagement are replacing likes and impressions as investment compasses.

When looking at demand generation, the playing field also shifts. Social commerce and user-generated content are no longer secondary players. The case of Granado’s sulfur soap, which jumped from 100,000 to 1.2 million units sold driven by UGC, consolidates the logic in which narrative, community, and social proof convert as much as, or even more than, a good paid media creative[3]. In digital retail, so-called double dates like 7.7 and 9.9 have become pillars of the calendar, with platforms distributing hundreds of millions in coupons, massive live streams, and influencers activated for events that, in some cases, already rival Black Friday itself[3].

Even so, the date remains strong in the imagination of the Brazilian consumer. The country is now the second that searched the term Black Friday most on Google in the last year, with 387 thousand searches, behind only France[4]. ABComm projects R$ 13.34 billion in e-commerce sales during the Black Friday period in 2025, compared to R$ 11.63 billion in 2024[4]. What changes is the form: instead of a single peak, retail will experience multiple peak periods throughout the year, requiring fine integration between CRM, media, pricing and logistics operations[3].

Consequences, Opportunities and Trends

The common denominator of these movements is simple and unsettling: marketing has ceased to be a game of reach and frequency and has become a game of relevance, data, and orchestration across multiple algorithmic contexts. As AI summaries capture some of the attention that previously flowed to websites, SEO needs to be rethought not only to rank on search engine results pages, but to be a source of trust for the AI models themselves[1]. Content depth, thematic authority, and quality signals now interact with the still opaque criteria of generative algorithms.

For brands, this opens up a set of strategic dilemmas. Investing heavily in paid media in an environment of inflated CPM and falling conversion tends to generate growth without margin if customer retention and lifetime value are not prioritized[2]. On the other hand, relying exclusively on third-party platforms – whether marketplaces, super apps or AI assistants – increases the risk of brand disintermediation. Building proprietary databases, active communities and data ecosystems becomes a competitive imperative.

From a professional standpoint, the profile of the marketer from 2025 onwards is shifting towards a hybrid of product strategist, applied data scientist, and journey architect. There is a growing demand for skills in data analysis, LTV modeling, the use of predictive AI, the design of loyalty programs, and native multiplatform social commerce content [2][3]. In a scenario where almost everyone will use AI to produce more, the differentiating factor is no longer just the tool, but the ability to formulate better questions, interpret weak signals, and connect business data to creative decisions. Those who understand this shift will not only survive the efficiency crisis but will also be able to lead the next wave of growth.

References

[1] Retail Center – AI reshapes marketing strategies and digital presence in Brazil
[2] NB News – Brazilian digital marketing faces efficiency crisis
[3] Marketing World – Top 10 news stories that had the most impact in 2025
[4] Meio & Mensagem – Black Friday 2025: Brazil is second in the search for discounts
[5] LCP Agency – ICT Households 2025 and the adoption of generative AI in Brazil