Milena Traikovich is a prominent figure in the demand generation space, specializing in the delicate art of turning complex data into high-quality leads. With a career built on the front lines of performance optimization and marketing analytics, she has spent years helping enterprises navigate the turbulent waters of digital transformation. Her deep understanding of how CRM architecture interacts with emerging technologies makes her a vital voice in the current debate surrounding the actual utility of agentic AI. As businesses grapple with the gap between high-level executive promises and the gritty reality of fragmented data systems, Milena offers a grounded perspective on why the industry is seeing a significant disconnect between AI hype and practical adoption.
The following discussion explores the current challenges facing major tech platforms as they pivot toward autonomous agents. We examine the critical role of data readiness, the financial implications of slow adoption rates that have wiped billions in market value, and the shift in priorities for marketing leaders. Rather than focusing on the glamour of AI, this conversation dives into the structural foundations required to make these tools work and the evolving strategies of CIOs who are increasingly skeptical of unproven “prime time” solutions.
Many enterprises struggle with fragmented CRM records and disconnected systems when trying to implement high-level AI; how does this data debt specifically undermine the promise of autonomous agents?
The reality on the ground is that AI agents are only as intelligent as the data they are fed, and right now, most company data is a mess. When we look at the adoption of platforms like Agentforce, seeing that only 34% of customers have actually integrated it tells a very specific story of frustration and technical hurdles. Marketers were promised a world where autonomous agents would handle customer service and sales tasks with ease, but instead, they are finding themselves stuck in the “muted” response phase because their data isn’t structured for autonomy. We are seeing cases where teams spend as much time manually organizing and cleaning their records as they do actually interacting with the AI, which completely negates the efficiency gains they were promised. If your customer information is inconsistent or locked in disconnected silos, an AI agent cannot make the split-second, informed decisions necessary to qualify a lead or launch a campaign effectively.
With the market seeing a staggering loss of more than $200 billion in value due to skepticism over AI strategies, is the hesitation more about the technology’s maturity or a lack of organizational preparation?
It is a bit of both, but the financial whiplash we are seeing—with shares falling more than 50% from their December 2024 peak—suggests that investors are losing patience with the “proof-of-concept” stage. Analysts are pointing out that only about 23,000 out of 150,000 customers are actually using these new agentic tools, which signals a massive gap between the CEO’s “all in” vision and what the average business can actually execute. There is a palpable sense of anxiety in the air when major firms like KeyBanc and Bernstein issue rare, simultaneous downgrades because the product just doesn’t feel ready for prime time yet. For a marketing leader, this means the pressure is on to prove that these investments aren’t just expensive experiments; however, until the product matures and the data foundation is solid, many organizations are choosing to pull back. It’s hard to justify increasing spend when the median increase in some sectors is a mere 3%, while other CIOs are actively looking to deprioritize these platforms in their upcoming IT budgets.
Marketing teams are often told that agents are the next major evolution in enterprise software, yet many are stuck in the proof-of-concept phase; what are the specific operational barriers preventing enterprise-wide rollouts?
The shift from a small-scale pilot to an enterprise-wide rollout is where most companies are hitting a brick wall, largely because the operational foundation just isn’t there to support it. When you look at the research, it’s clear that many deployments are stalled because the agents can’t handle the complexity of real-world, messy CRM records that haven’t been touched in years. We’re seeing a significant trend where CIOs are more likely to reduce spending over the next 12 months rather than increase it, simply because they aren’t seeing the ROI from these early projects. The acquisition of companies like Informatica shows that even the biggest tech giants realize they need better data-management and governance tools before they can sell more AI. Without the ability to automatically pull customer data from external sources and ensure it is clean and connected, an autonomous agent is essentially a high-tech engine without any fuel, and that realization is causing a lot of teams to hit the brakes on their wider implementation plans.
For a marketer looking at the current landscape, how should they shift their priorities between buying the newest AI software and fixing their existing data governance?
The smartest move a marketer can make right now is to ignore the shiny object and focus on the plumbing. The companies that are moving the fastest aren’t necessarily the ones with the biggest budgets for new software; they are the ones that already built a data foundation that allows AI to deliver meaningful results. We are seeing a shift where the priority is moving away from just “automating campaign execution” and toward improving the underlying data quality and integration. If you try to automate lead qualification or customer personalization on top of a broken system, you’re just going to scale your mistakes and alienate your customers faster. The takeaway from the current market volatility is that you will see much greater returns from spending the next six months on data governance than you will from trying to deploy a “prime time” AI agent that isn’t actually ready for your specific business environment.
As we see some analysts upgrading stock while others warn of a downturn, how can businesses differentiate between hype-driven marketing and genuine growth engines in the AI space?
You have to look past the public dismissals of “bad calls” and look at the internal metrics that actually drive growth, such as whether the tool is actually solving a customer friction point or just adding another layer of complexity. While some firms like Guggenheim and Monness see a meaningful upside and point to the fact that some companies are increasing their spending, the broader market is still signaling caution. A genuine growth engine is something that doesn’t require a manual “data-cleaning” marathon every time you want to launch a new feature. For marketers, the “growth” isn’t in the AI itself, but in the 1:1 experiences at scale that the AI is supposed to enable; if you can’t get to those experiences because you’re stuck in a proof-of-concept loop, then it’s not a growth engine for you yet. We need to be very critical of any tool that claims to be the “fastest-growing product in history” if the majority of the customer base is still struggling to get the basic data to sync up.
What is your forecast for the future of agentic AI in the enterprise marketing sector?
My forecast is that we are heading into a “great consolidation” period where the focus shifts entirely from AI features to data-centric infrastructure. Over the next year, I expect we will see a massive surge in demand for data integration and governance specialists as companies realize that their $200 billion dreams of autonomous marketing are currently being held back by messy spreadsheets and fragmented records. We will likely see a widening gap between the 23,000 early adopters who have the resources to fix their data and the rest of the 150,000 customers who might stay on the sidelines. The true evolution won’t be the agents themselves, but the move toward “clean-data-as-a-service” models that allow these agents to finally operate without constant human intervention. Once the industry moves past this initial “prime time” friction, the companies that spent this year fixing their foundations will be the ones that finally see the autonomous automation they were promised.
