The Future of Websites in the AI Era: Why Trust May Matter More Than Personalization

Introduction

Anda Gansca, Co-Founder & CEO of Knotch, brings a sharp perspective to one of the more uncomfortable questions facing marketers today: what happens to the website when AI changes how people discover, evaluate, and trust brands? As the leader of an AI-powered content intelligence company that's raised over $40 million and works with Fortune 1000 companies, Gansca spends her days helping marketers co ect digital content to real business outcomes. In a candid conversation, she and Benjamin Shapiro dig into the tension between trust, traffic, and personalization—and what marketers still believe about their websites that simply won't hold up three years from now.

Rethinking Traffic as a Success Metric

One of the most provocative takeaways is that traffic—long a foundational website metric—may matter far less in the AI era. As AI-driven discovery and answer engines intercept user journeys, marketers can expect fewer raw visits to their digital properties. That shift forces a hard question: if traffic shrinks, what becomes the meaningful measure of website success? The answer increasingly points away from volume and toward influence and outcomes.

Why This Matters for Marketers

For marketing leaders who still build reporting decks around pageviews and sessions, this is a wake-up call. When fewer people arrive at your site but the ones who do carry higher intent, the goal shifts from attracting traffic to converting trust into action. Knotch's own positioning reflects this—capturing data on not just what content is seen, but how it influences conversions across the path to purchase.

The Personalization Paradox

Gansca initially makes the case for personalization, arguing that it drives trust, which in turn drives word-of-mouth and traffic. But she quickly complicates her own answer, warning that personalization can "jump the shark" and slide into the creepy factor. "We might have experiences that are so personalized they get into the creepy factor," she notes—a reminder that more tailoring isn't automatically better. The conversation draws a direct line back to the early days of marketing automation, when tools like Marketo enabled marketers to stuff every available data point into emails, producing experiences that felt invasive rather than helpful.

The LLM Training Data Problem

A critical insight emerges around how AI handles personalization today. Gansca points out that we're no longer relying on humans to personalize—we're asking large language models to do it better. But Shapiro raises a sharp counterpoint: "Their training data is the mediocre job we were doing before." If the humans doing personalization only ever did a "mid job," the AI learning from that work inherits the same mediocrity. The implication is significant—throwing an LLM at personalization doesn't automatically elevate quality, because the model is calibrated on the very work it's meant to improve upon.

Why Trust May Be the Real Differentiator

The conversation lands on a clear strategic priority: trust. Shapiro makes the case that when a customer already trusts a brand, the message doesn't need to be as precisely tailored, because the relationship equity is already there. In a world where personalization risks crossing into creepiness and AI-driven personalization inherits flawed human habits, trust becomes the more durable asset. The human element still matters—getting tone and tenor right requires judgment that training data alone can't replicate.

Conclusion

The throughline of this discussion is that marketers need to stop treating their websites the way they did three years ago. Traffic is becoming a less reliable measure of success, personalization is a double-edged sword that can erode the relationships it's meant to build, and AI-driven personalization carries the baggage of mediocre human training data. The strategic bet that holds up best is trust—the relationship equity that lets brands earn attention without manipulating it. For marketing leaders, the action is clear: invest in building genuine trust through your digital properties, measure influence over volume, and apply human judgment where AI alone falls short. Until next time, focus on keeping your customers happy.