When AI Gets It Wrong: A Real World Lesson in Generative Engine Optimization
JoAnna Dettmann | Partner

Last week, members of our team were in Bend, Oregon for a retreat. While at a well known local restaurant, a situation unfolded that perfectly illustrates the very real business risks of inaccurate AI generated information.

A customer approached the bar around 5:15 pm and asked for the happy hour menu. The bartender informed him that happy hour had ended at 5:00 pm. The customer immediately disagreed and showed the bartender what Google’s AI Overview returned when searching “Pine Tavern happy hour.”

The AI result confidently stated that happy hour ran Monday through Friday from 3:00 to 6:00 pm.

The problem is that this information was wrong.

The restaurant had already updated its actual happy hour hours to end at 5:00 pm, and its Google Business Profile reflected the correct time. However, the AI result surfaced outdated information first and positioned it as factual. As can happen, the situation escalated, the guest was frustrated, the staff was put in an uncomfortable position, and a real customer experience was negatively impacted because of incorrect AI data.

After digging into the source, we discovered the root of the issue. An old happy hour PDF was still hosted on the restaurant’s website. Even though the business had updated their Google listing, the AI engine pulled outdated information from the website and presented it as truth.

This is exactly how Generative Engine Optimization becomes a revenue protection issue.

Search behavior is changing rapidly. People are no longer just scrolling through search results. They are reading AI summaries and treating them like absolute truth. If your brand’s content is inconsistent across your website, PDFs, menus, landing pages, and business listings, AI will still answer for you using whatever data it can find.

And it does not check if it is current.

This creates serious risks:

  • Customer frustration and public disputes
  • Lost trust
  • Lost revenue
  • Incorrect business information being broadcast at scale
  • Staff caught in the middle of AI misinformation

Traditional SEO alone does not solve this problem. Ranking well does not ensure that AI engines pull the correct version of your data. GEO focuses on controlling what generative systems index, summarize, and present as factual answers.

Generative Engine Optimization includes:

  • Auditing all indexable documents and assets
  • Removing or updating outdated PDFs and menus
  • Structuring content so AI can interpret it correctly
  • Aligning website data with Google Business Profiles and local listings
  • Ensuring one consistent source of truth across all machine readable surfaces

This is no longer theoretical. This is happening in real time, at real businesses, with real customers.

If AI is now speaking on behalf of your brand, you need to be certain it is speaking accurately.

JoAnna Dettmann
JoAnna Dettmann leads operations at Elasticity, where she streamlines systems, improves cross-team coordination, and supports the agency’s day-to-day momentum. Before joining Elasticity, she founded and operated several successful businesses across marketing, fitness, consulting, and travel, giving her a grounded perspective on how teams function and how great work gets delivered. She is committed to maintaining strong employee morale and creating an environment where people feel supported, valued, and able to do their best work.
Privacy Policy | All rights reserved. © 2026 ELASTICITY

Best Software Development Company