Notes from the CMO Summit: AI Adoption Is Accelerating, But the Marketing Budget Still Runs on Excel

Geoff helps enterprise organizations modernize marketing, planning, and revenue operations through AI-enabled transformation strategies. With 29 years of experience across enterprise technology, SaaS, and data-driven marketing, Geoff specializes in connecting strategy, operations, and technology to help clients improve decision-making, accelerate execution, and better measure business impact. 

Argano recently attended the CMO Summit San Francisco, hosted by CMO Alliance. The event is deliberately positioned in contrast to large-scale marketing conferences such as Adobe Summit, which draws upward of 30,000 attendees in a brand-event format. By comparison, CMO Summit convened roughly 100 attendees; 95% of them were senior marketing leadership, representing more than 75 companies, predominantly B2B organizations and technology firms.  

The format favors structured peer discussion over vendor exhibition. Our conclusion is that this conference was valuable because unlike others, there was less emphasis on vendor pitches and more on direct exchange aong marketing leaders. Because of that, we gained amazing insights into the core challenges facing marketing executives today related to many areas, including the wave of AI crashing over them. 

Agenda Structure Reflected the Industry's Core Tension 

The agenda was organized deliberately. Morning sessions convened both B2B and B2C leaders around shared pressures: growth, AI adoption, ROI, and evolving customer behavior. Afternoon sessions split into dedicated tracks, reflecting how differently those pressures are being addressed in practice: B2B sessions focused on account-based marketing, personalization, longer buying cycles, and navigating unpredictable markets; B2C sessions addressed customer loyalty, experience, retention, and brand relevance. 

The event’s stated theme — “attention is bought, but trust is earned” — proved to be an accurate descriptor of the substantive discussion across sessions, particularly those concerning AI governance and brand integrity. 

Two Dominant Themes: AI and Operational Discipline 

The summit centered on two themes that were in constant tension: artificial intelligence, and the operational discipline required to run a marketing organization. Attendance was split nearly evenly between the two tracks, with substantial cross-participation, indicating that marketing leaders are not treating these as separate concerns but as an integrated set of decisions. 

The central question raised throughout the event: how much operational authority should be extended to AI systems, and how much should remain with existing processes and personnel. Organizations that over-index on AI-driven autonomy risk losing visibility into whether marketing activity is translating into measurable business outcomes. The prevailing view among attendees was that AI functions best as an accelerant to existing operational structures rather than a replacement for them. 

Brand Governance Remains a Human-Judgment Problem 

A recurring point across sessions: automated systems evaluating creative assets can assess technical and brand-guideline compliance, but they do not reliably identify cultural or historical context that a human reviewer would catch.  An automated review process might register a compliant, on-brief asset without recognizing the broader context that made the imagery objectionable. For CMOs, this underscores that AI governance is not solely a question of output qualitybut rather it is a question of where final approval judgment authority resides in not just content but also decision making in general. 

This connects to a related theme in the agenda: as buyers increasingly discover and evaluate brands through AI-mediated search and answer engines, brand representation is being interpreted by systems the CMO does not directly control. Governance is no longer limited to internal creative review; it extends to how AI intermediaries represent the brand before a prospect ever engages directly. 

Budget Management Remains Largely Unautomated 

A consistent finding across conversations: every CMO interviewed continues to manage marketing budgets in Excel. This is a known and acknowledged limitation, and organizations report pressure to modernize while also determininghow AI should be incorporated into that process. 

An informal survey conducted during these conversations asked how comfortable organizations would be allocating marketing budget directly through an LLM-based tool. The response was uniformly negative. When asked to identify the source of that hesitation, respondents cited distrust of AI vendors, the immaturity of the underlying technology, and uncertainty about whether model providers themselves maintain full control over their AI technology. This pattern heldconsistently among enterprise organizations. Mid-market organizations showed greater willingness to adopt LLM-based budgeting tools more quickly, often driven by competitive pressure to move faster than larger incumbents. This is a divergence worth monitoring as the market matures. 

The distinction driving this caution: budget allocation is a deterministic function. It is factual, not probabilistic. Decisions about how that budget is deployed, and forecasting the outcomes of that deployment, are appropriately probabilistic and represent where AI-driven tools can deliver additional value. Conflating the two, i.e., treating budget management itself as a probabilistic exercise, is where organizational trust breaks down. 

The Prevailing Model: Augmentation, Not Replacement 

The most concrete articulation of this principle came from a conversation with a senior marketing executive at a major B2B organization ahead of the summit: core systems of record remain in place, with an agentic layer introduced on top, focused specifically on reducing friction between process steps rather than assuming ownership of underlying data. This reflects an emerging first-phase model for AI adoption. For example, agentic systems operating on top of the existing ERP and SaaS technology stack rather than displacing its core components, can lead to greater efficiencies, especially for processes that cut across multiple systems and teams. 

Implications for CMOs 

The central conclusion from the summit: marketing operational infrastructure has not kept pace with the speed of AI adoption pressure and closing that gap without compromising brand integrity or financial governance represents the primary challenge facing CMOs today. This is not an argument for slowing AI adoption. Instead, it is an argument for being deliberate about where AI-driven process improvement and insight adds value while ensuring the necessity of human decision making and oversight.