Ask Not What Your Customers Think of You, But What Their AI Thinks of You
How to thrive when AI systems, not customers, decide who makes the shortlist
If you were a film director in the 90s, you'd spend opening weekend with your stomach in knots, waiting for Roger Ebert and Gene Siskel's verdict on your latest work. Those iconic "two thumbs up" weren't just reviews—they were career-makers. Studios would hold their breath as the critics delivered their judgment, knowing that a positive review could drive box office success while a negative one might doom a film before word-of-mouth could spread. Directors would craft scenes with these critics in mind, producers would schedule strategic screenings, and marketing teams stood ready to plaster "Two Thumbs Up!" across every poster and TV spot if fortune smiled.
We've seen this power of gatekeepers evolve over time. From individual critics like Siskel and Ebert, we moved to aggregators like Rotten Tomatoes, democratizing the critical voice somewhat but still maintaining that filtering function.
But now, a new critic has entered the theater, one that doesn't publish reviews or give thumbs up. This critic whispers directly in your customer's ear, suggesting what they might enjoy before they've even asked. This critic is AI, and it's rapidly becoming the most influential reviewer in history—not just for movies, but for everything.
Economist Tyler Cowen recently made this point explicit in an interview with Dwarkesh Patel: "The last book I wrote is I'm happy if humans read it, but mostly I wrote it for the AIs. I wanted them to know I appreciate them. My next book, I'm writing even more for the AIs."
Cowen isn't being facetious—he's acknowledging a shift in who we need to impress. Just as restaurants once lived or died by the New York Times review, and films rose or fell on Roger Ebert's thumbs, tomorrow's businesses will succeed or fail based on what AI systems think of them. The critic's pen has been replaced by the algorithm's recommendation, but the power dynamic remains the same—except now, these new critics are universal, omnipresent, and increasingly trusted.
Welcome to the era of the AI Perception Index, where what matters most isn't what your customers think of you, but what their AI thinks of you.
From SEO to APO: Optimizing for AI Perception
For decades, businesses have obsessed over search engine optimization (SEO). They've analyzed keywords, built backlink strategies, and restructured websites—all to appear higher in Google's search results. SEO was fundamentally about rankings—climbing up the list for specific keywords, fighting for those coveted first-page positions. An entire industry emerged around understanding and influencing the algorithms that determined visibility.
Now we're entering the age of AI Perception Optimization (APO). The stakes are higher and the game is different. While SEO was about rankings—like fighting for better placement in the Yellow Pages—APO is about making a personalized shortlist, more akin to the New York Times "36 Hours in Dublin" guide. When the Times publishes their guide, they don't include every restaurant and attraction—they select a handful they deem worthy.
When a user searches Google, they see a universe of options—pages of results they can browse through, with the comforting "Next Page" button promising more possibilities. When they ask an AI assistant, they often receive a single answer or a highly curated shortlist tailored specifically to them. This is AI's promise: removing the overwhelm of endless options by delivering a manageable, personalized selection. But this promise comes with a curse for businesses—there's no "See More" link, no pagination, no way to browse beyond what the AI initially offers. The difference between being on that shortlist and missing it isn't a slight drop in traffic—it's potentially existential. If you're not recommended, you effectively don't exist.
In this article, we'll explore this new reality through the eyes of Mountain Pine Resort, a fictional Colorado ski destination wondering how to adapt to this AI-mediated world.
Travel: The Perfect Testing Ground for AI's New Gatekeeping Role
The travel industry stands at the forefront of this shift from search to AI-mediated discovery. According to Statista, 40% of travelers now use AI for planning—up from zero just three years ago. This rapid adoption isn't surprising when you look at the problem AI solves.
Consider a family from the Texas Triangle—the metropolitan region encompassing Dallas, Houston, San Antonio, and Austin, home to 20 million people. When planning a spring break ski trip, they face a classic choice overload: dozens of resorts across multiple states, each with their own terrain, amenities, and ski schools. Traditionally, this meant hours spent across numerous websites, comparing options and making trade-offs.
With AI, they simply describe their situation: "We have two children who've never skied before. We'd prefer to drive rather than fly. And we'd like some non-skiing activities." The AI doesn't just search—it evaluates, delivering not an endless list but a curated shortlist of perhaps three resorts that best match their specific needs.
The benefits for consumers are clear and measurable:
Better personalization: 68% of travelers say AI-based itineraries feel more tailored to their preferences (Knowridge.com)
Time efficiency: 72% of travelers spend significantly less time planning with AI tools (Knowridge.com)
Reduced stress: 85% of AI travel tool users find chat interfaces more engaging and less stressful than traditional search (PR Newswire)
AI functions essentially as a chief of staff or personal travel agent—taking the infinite list of options and creating customized shortlists. It's like having those "If you only have 2 days in Paris" travel guides, but personalized for your specific circumstances.
For the Texas family, this means receiving recommendations for perhaps Taos Ski Valley (closer driving distance), Keystone (dedicated family learning area), or Copper Mountain (terrain ideal for beginners)—each matched to their specific needs.
This is a win for consumers. But for Mountain Pine Resort and thousands of other travel businesses, it creates an urgent question: how do you ensure you make the shortlist when a family from Texas asks about spring break ski destinations? When the AI becomes the gatekeeper, what determines whether you get past the velvet rope?
The AI Perception Framework: Understanding Your Standing with the New Gatekeepers
If Mountain Pine Resort's marketing team faced the challenge of AI perception, they would start with a familiar tool: the brand perception framework. Companies have used these frameworks for years to track how different customer segments view their offerings across key attributes. This proven approach can be repurposed for the AI era.
The key adaptation is measuring what AI systems think about your business for each customer segment, not just what customers think directly. It's adding a new layer to the traditional model—one that sits between you and potential customers. After all, AI is now a gatekeeper whose perception matters on behalf of your target personas.
Step 1: Define Your Key Personas and Their Critical Attributes
Mountain Pine would identify their most important customer segments and the specific attributes each segment prioritizes when making decisions. For families with children, these attributes include kid-friendly slopes, quality of ski schools, family accommodations, and non-ski activities—each with different priority weightings. Other segments like expert skiers or luxury travelers would naturally have different priorities.
As shown in Figure 1, the Texas Triangle Family has specific priorities with clear weightings. Understanding these attribute priorities matters because they directly determine whether Mountain Pine will make the AI's shortlist. To be recommended to families, Mountain Pine must rank highly on the attributes families care about most.
Step 2: Map the AI Landscape
A crucial step Mountain Pine would need to take is mapping which AI systems matter most to their potential customers. This isn't a single-AI world:
ChatGPT currently holds the largest market share for general queries and also powers travel industry AI bots like TripAdvisor and Expedia
Gemini, Claude, and Perplexity each have growing user bases with different strengths
Newer models like DeepSeek are constantly entering the field
Mountain Pine would need to test their perception across this entire ecosystem, weighing each AI's importance based on market share and relevance to travel planning.
Step 3: Measure AI Perception on Key Attributes
With personas and their critical attributes defined, Mountain Pine would measure how each AI perceives them on each specific attribute that matters to each persona. The key is measuring perception not just overall, but for every attribute that influences the AI's recommendation decision.
Figure 2 reveals how Mountain Pine might rank on each specific attribute when evaluated for Texas families. Since attributes like kid-friendly slopes and ski school quality are high priorities for this persona, poor rankings on these specific attributes would explain why Mountain Pine isn't making the shortlist.
Tracking these attribute-specific measurements over time gives Mountain Pine actionable insights. By establishing a baseline for each attribute and monitoring monthly, they can spot:
How website updates affect perception of specific attributes
Which attributes improve or decline after AI model updates
How competitors' rankings on key attributes change over time
Which attributes need immediate attention to improve shortlist chances
When Mountain Pine notices a drop in how Claude ranks their ski school quality, they can investigate what changed—whether it's their content, competitor improvements, or model updates.
For Mountain Pine Resort and thousands of other businesses, this framework provides a systematic way to understand how AI systems perceive them. The next question becomes: once you know where you stand, what do you do about it?
From Insights to Action: Addressing AI Perception Gaps
Once Mountain Pine has measured how AI systems perceive them across key attributes, they'll discover gaps between their desired and actual perception. These gaps fall into two categories:
Reality Gaps: When AI accurately perceives actual shortcomings in Mountain Pine's offerings. These can only be addressed through product or service changes.
Perception Gaps: When Mountain Pine's actual offerings are strong, but AI systems don't recognize this. These require communication and content strategy adjustments.
The Mountain Pine team might discover a significant perception gap around their ski school. Despite investing $2 million in a new children's ski school facility with specialized instructors, AI assistants still describe their ski school as "primarily focused on advanced skiers" when Texas families ask about ski lessons for beginners.
This perception gap doesn't require Mountain Pine to rebuild their ski school—it requires them to bridge the information gap between reality and AI perception. They might create dedicated landing pages profiling their children's ski instructors, complete with their backgrounds teaching young skiers. They could actively encourage families who've had positive experiences to leave specific reviews mentioning the ski school. They might develop content answering common questions parents have about teaching children to ski.
In addressing these perception gaps, Mountain Pine isn't just fixing a marketing problem—they're gaining valuable insight into what their customers truly value.
The Critic's Gift: Finding Connection Through AI
As Mountain Pine works to align AI perception with reality, a surprising truth emerges: what initially appears as a barrier between business and customer may actually bring them closer than ever before.
What begins as a story of separation often reveals itself as one of connection. When Siskel and Ebert first wielded their thumbs, filmmakers saw only gatekeepers standing between them and their audience. Yet the wisest directors discovered something more valuable—a window into how their work was experienced, a voice that articulated what audiences felt but couldn't always express.
In the digital age, this clarity faded. The chorus of online reviews created noise without signal, volume without direction. Businesses found themselves adrift in a sea of contradictory feedback, unable to discern which voices truly mattered or why.
Now, as AI steps between businesses and customers, we face what appears to be another layer of separation. Mountain Pine Resort might wonder how they can connect with Texas families when an AI speaks on their behalf. Yet in this apparent distance lies an unexpected intimacy.
Unlike the silent algorithms of search engines or the chaotic sprawl of online reviews, AI testing reveals patterns that would otherwise remain hidden. By systematically measuring how different AI systems rank Mountain Pine across key attributes for each customer segment, they gain a competitive intelligence goldmine. They can see exactly where they stand relative to competitors on the attributes that matter most to Texas families.
This comparative view reveals insights that traditional market research might miss. Mountain Pine might discover that while they've invested millions in high-speed chairlifts, their competitors who rank higher for families have invested in features like hot chocolate stations along beginner trails and colorful trail markers designed for children.
This isn't disintermediation—it's illumination. The AI doesn't obscure customer needs; it clarifies them with precision. It reveals patterns of preferences that businesses can understand and address.
These insights become the North Star for product development and customer experience. The companies that thrive in this new landscape won't be those lamenting the AI gatekeeper, but those embracing it as the most articulate focus group they've ever encountered. They'll build experiences that respond not to guesswork but to precisely understood customer needs.
What first appeared as a barrier reveals itself as a bridge. In the age of AI gatekeepers, we may find ourselves not further from our customers, but closer than we've ever been—listening not just to what they say directly to us, but to what they whisper to their AI assistants when dreaming of their next adventure.
The screenshots above are from the beta of our AI Perception Tracker. Interested in learning more? Sign up here.
