Headlines have sounded the same refrain for years: most AI projects fail. Research from MIT Sloan Management Review and others paints a sobering picture of large investments, ambitious initiatives and too few tangible results.
But that story, however, is incomplete. The real problem isn’t that AI is failing, its’ that many organizations are aiming AI at the wrong targets. The real opportunity lies elsewhere, where AI directly connects with customers and creates measurable value.
New data from The State of AI Agents & No-Code Global Report 2025 — a survey of 564 global business and technology decision-makers — reveals that AI succeeds where it touches the customer. Enterprises that focus AI on sales, marketing and service functions are already realizing measurable ROI, stronger customer loyalty and faster growth. The evidence suggests that: AI works best when deployed where business value is most visible.
The pivot point: Customer-facing functions are leading the way
When executives were asked where they’re most likely to deploy AI agents, one group of functions stood out: sales, marketing and service. Nearly half, 46 percent of respondents said they’re highly likely to deploy AI agents in these areas.
Why here? Because these are the frontlines of growth. Every interaction with a prospect or customer carries an immediate impact on revenue, loyalty and experience. Unlike internal efficiency projects, which often focus on incremental efficiency gains, customer-facing initiatives provide clear, measurable outcomes that can be evaluated quickly.
We’re already seeing examples across industries:
- In sales, AI helps reps prioritize leads, personalize outreach and accelerate deal velocity.
- In marketing, AI agents drive hyper-personalized engagement, campaign optimization and smarter segmentation.
- In service, they assist human agents to deliver faster, more empathetic experiences.
Notably, when decision-makers were asked to name a single task or workflow they’d like to see fully handled by AI agents in the next 12 months, “customer” was the most frequent word in their answers. It’s not just a technical priority — it’s a strategic one.
From cost reduction to revenue acceleration
For many years, automation strategies focused primarily on cost reduction. Organizations used technology to streamline back-end operations, remove manual tasks and improve efficiency. Those gains are real, but they are incremental.
Frontline AI deployment, by contrast, is about accelerating revenue. When applied to customer engagement, AI can boost conversion rates, shorten deal cycles, deliver personalized marketing experiences and resolve service issues faster.
This shift, from the back office to the front line, is where AI’s potential becomes fully visible. Because outcomes are easier to measure, these initiatives tend to attract executive sponsorship and sustained investment. Boards are more willing to fund projects that clearly demonstrate financial impact, and AI at the customer interface does exactly that.
Augmentation, not replacement
Perhaps the biggest misconception about AI is that it exists to replace humans. The survey results tell a different story. Only 11 percent of leaders expect AI to reduce headcount. A striking 84 percent say AI will augment existing teams, create growth opportunities or even generate new roles.
This “AI + human” hybrid workforce model is emerging as the most powerful use case: AI handles repetitive, low-value work, freeing people to focus on creativity, relationship-building and strategic decisions. Instead of displacing the workforce, AI elevates it.
As Wendy Albers, SVP of member experience at Generations Federal Credit Union, puts it, “AI capabilities are not to be feared but, rather, can be viewed as valuable tools that augment the work of humans and free us to perform the more important work of connection and creativity.”
This isn’t just cultural insight. It’s a business advantage. Companies that align AI with their people strategy, rather than setting the two at odds, are unlocking sustained performance improvements and higher levels of engagement.
Re-imagine, don’t simply replicate
Too often, companies approach AI projects by automating existing workflows rather than rethinking the experience altogether. This is a critical strategic error. Automating a bad or inefficient process doesn’t make it better; it just makes the inefficiency run faster.
Customer-facing processes are especially vulnerable to this trap. Legacy workflows are often built around internal systems and silos, not around customer needs. AI provides an opportunity to step back and redesign these journeys from the ground up: to imagine how an agent can handle interactions in real time, personalize experiences dynamically and reduce friction at every touchpoint.
The organizations seeing the strongest ROI are the ones asking: “If we were to design this process today with AI and agents in mind, what would it look like?” That mindset shift — from replication to reinvention — is where transformation truly happens.
Think cross-functionally across customer journeys
For decades, CRM vendors have promised end-to-end customer journeys and “Customer 360.” Most enterprises still struggle to deliver this because their systems don’t talk to each other. Data is siloed, workflows are fragmented and achieving journey orchestration requires complex, slow integrations.
AI agents offer a new path. Because they can operate across systems and functions, agents can bridge departmental boundaries — connecting sales, marketing, service and operations into cohesive journeys without forcing a massive re-platforming effort.
For example, an agent could seamlessly hand off a prospect from marketing to sales while tracking engagement history and context in real time. Or it could assist both service reps and product teams simultaneously, ensuring customer issues are resolved while insights flow back to engineering.
This cross-functional orchestration is one of the most underappreciated advantages of agentic AI: it finally delivers on the promise of connected experiences without waiting years for technology consolidation.
Empowering the business and accelerating results
One of the most persistent challenges in customer-facing teams is that they often lack the technical expertise to build or deploy new tools. They depend on IT backlogs, vendor roadmaps or costly third-party projects, which slow down innovation and stifle experimentation.
This is exactly why AI and no-code are such a powerful combination. No-code platforms lower the barrier to entry, letting business users directly design, deploy and refine AI-powered workflows. This puts control back in the hands of the teams closest to the customer; the people who understand the nuances of engagement, timing and personalization.
According to the report, 67 percent of organizations have already adopted no-code either enterprise-wide or within key departments. These platforms remove complexity, allowing business users to deploy and iterate AI agents under proper governance. The results are tangible: faster time to value, lower reliance on scarce technical talent and stronger alignment between business goals and AI implementation.
This democratization of AI development means that customer-facing teams don’t have to wait for IT roadmaps to innovate. They can build and refine AI-powered workflows directly while maintaining compliance and control. As AI matures, this ability to move at the speed of the business may become one of its most decisive competitive advantages
Start where the ROI is obvious
The AI “failure” narrative misses a critical truth. It’s not that AI can’t deliver value. It’s that many enterprises are starting in the wrong places with low-impact use cases, pilot purgatory and little business sponsorship.
Organizations that begin with customer-facing functions, where impact is visible, measurable and fast, prove that AI can and does work.
AI isn’t failing. It’s waiting to succeed in the right place. The companies that align their AI strategies with customer impact and frontline execution will set the pace for the next era of growth.
This article is published as part of the Foundry Expert Contributor Network.
Want to join?
Read More from This Article: AI isn’t failing — You’re just looking in the wrong places
Source: News

