Why Most Brands Still Get AI Wrong – Insights from a 7-Year-Old Prediction

In 2018, I completed my honors thesis on Artificial Intelligence in Marketing at Barrett, The Honors College at Arizona State University. At the time, AI was still viewed as a fringe technology in marketing - something exciting, but not yet fully understood or adopted. In that thesis, I argued that while AI had the potential to reshape marketing entirely, most brands would struggle to implement it correctly unless they rethought their structures, strategies, and mindsets.

Seven years later, in 2025, much of what I predicted has played out. AI has become one of the most talked-about tools in business, yet the same fundamental mistakes persist. Brands still confuse AI adoption with AI transformation. They chase shiny tools instead of building sustainable systems. And they often overlook the long-term implications of what AI means for consumers, marketers, and the very nature of business itself.

This article revisits the key insights from my thesis, showing how the challenges I outlined then are the same ones brands still grapple with today and what it will take to finally get AI right.

The Hype vs. Reality Gap

In 2017, surveys showed that 98% of marketers expected AI to benefit their work, but only 20% had actually implemented it at scale. I described this as the “hype vs. reality gap.” The enthusiasm was there, but execution lagged far behind.

Fast forward to today, and while adoption has increased dramatically, the pattern hasn’t changed. Many companies have bolted AI tools onto their existing processes (chatbots, predictive analytics dashboards, programmatic ad platforms) without rethinking how AI fundamentally changes the way marketing should operate.

The result is what we see everywhere: superficial personalization, endless retargeting, and dashboards that promise “intelligence” but deliver little more than glorified automation.

What Brands Still Get Wrong

1. Superficial Personalization

AI was supposed to usher in true one-to-one marketing: hyper-personalized experiences that made consumers feel understood. But most brands still rely on cookie-based retargeting and surface-level segmentation. Back in 2018, I warned that cookies were a weak foundation for personalization and would eventually collapse under privacy concerns. That prediction came true with GDPR, CCPA, and the broader privacy-first movement. Yet many brands haven’t adapted.

2. Data Silos

In my thesis, I outlined how internal silos cripple AI implementation. Without unified data across teams and channels, AI cannot deliver the deep insights marketers expect. Seven years later, this remains one of the biggest barriers. Brands collect more data than ever, but much of it sits in fragmented systems, making “AI-driven insights” shallow at best.

3. Blind Trust in Vendors

Another recurring mistake is treating AI platforms as black boxes. Too often, marketers trust the outputs without questioning how they’re generated or whether they align with strategic goals. In 2018, I cautioned against overspending on premature AI implementations driven by hype. Today, brands still waste millions chasing promises of automation without demanding transparency, accountability, or ROI proof.

4. Ignoring Consumer Trust

Perhaps the biggest oversight is consumer perception. My research showed that while marketers were overwhelmingly optimistic about AI, consumers were far more cautious. That disconnect is still visible today in the rise of AI-related distrust - concerns about privacy, job displacement, and algorithmic bias. Brands that fail to address these concerns directly risk eroding trust.

Where the Predictions Came True

Despite the challenges, many of the benefits I highlighted in 2018 have proven accurate:

Hyper-Personalization

AI enables brands to tailor experiences at scale, when executed correctly.

Efficient Spending

‍Automation of repetitive tasks reduces waste and increases ROI.

Deeper Insights

‍Machine learning allows marketers to identify patterns in customer behavior that humans could never detect.

Scalable Experiences

‍From chatbots to dynamic creative, AI allows brands to deliver consistent value across channels.

I also predicted pitfalls that have materialized: malicious use of AI (deepfakes, fraud), job displacement debates, and technical limitations that make AI less reliable than the hype suggests.

Case Studies Then and Now

In my thesis, I examined brands experimenting with AI at the time. Harley Davidson NY, for instance, used an AI platform called Albert to increase leads by nearly 3,000% in just three months. OYO Hotels used AI-driven personalization to boost bookings by 5x. These early successes showed AI’s potential, but they also revealed a pattern: AI worked best when paired with human oversight, unified data, and a willingness to restructure marketing processes.

Today, the most successful brands with AI share that same trait. They treat AI not as a campaign add-on but as an operating system for how marketing functions. Companies that fail to make this leap remain stuck in the cycle of hype, underperformance, and consumer frustration.

How Brands Can Finally Get AI Right

If brands want to stop “getting AI wrong,” they need to shift their approach:

1. Think System, Not Tool: AI should be built into the operating fabric of marketing, not just plugged into campaigns.
2. Break Down Silos: Unified, cross-channel data is the foundation for meaningful AI-driven insights.
3. Demand Transparency: Don’t accept black-box outputs. Require vendors to show how models work and how ROI is measured.
4. Augment, Don’t Replace: AI should amplify human creativity and empathy, not attempt to erase it.
5. Center Consumer Trust: Be transparent, respect privacy, and design AI experiences that add value without intrusion.

Looking Ahead

In 2018, I called the coming era the “AI Marketing Era” - a shift as fundamental as the introduction of computers. That era is now here. But the brands that thrive will not be the ones with the flashiest AI demos. They’ll be the ones that reimagine marketing itself through the lens of intelligence, ethics, and consumer trust.

Seven years ago, I predicted that AI would change the very nature of marketing. That prediction proved true. What remains to be seen is whether brands will adapt - or whether they’ll keep making the same mistakes for another seven years.

‍