This blogpost is a port of the whitepaper True Ventures Fellowship Project: AI Audit at Zephyr Cloud from Nitin Kanchi, Miles Kennedy, James Duong
As part of the True Ventures Fellowship, our team at Zephyr Cloud (Miles, Nitin, and James) conducted a detailed audit of our AI driven workflows.
While we discovered from the start that Zephyr’s processes are already heavily optimized with AI integration across code review, content generation, productivity, and management, we identified a critical marketing problem after talking with our CEO, Zack Chapple:
the company lacks optimization for inbound traffic coming from generative AI platforms such as ChatGPT, Perplexity.ai, Claude, and others.
This is an important shift, as these platforms increasingly serve as the starting point for user information queries competing with, and in some cases replacing, traditional search engines like Google.
The Problem: From SEO to GEO
Traditional Search Engine Optimization (SEO) no longer captures the full picture on how users discover information online. SEO alone does not account for how Generative AI engines source, synthesize, and present information to end users. Generative AI Platforms are changing information discovery by:
- Generating natural-language answers rather than returning link lists
- Aggregating multiple sources into coherent responses
- Bypassing traditional search ranking with their own content selection algorithms
For Zephyr Cloud, the challenges were:
- Measurement Gap - No current method to track traffic coming from AI platforms
- Optimization Gap - Content not structured for AI engine prioritization
Understanding Generative Engine Optimization (GEO)
Our research led us to the concept of Generative Engine Optimization (GEO), a marketing channel designed to improve visibility in the output of generative AI engines.
GEO is conceptually similar to SEO but focuses on the factors that influence whether and how generative AI engines reference or cite a piece of content.
This involves understanding what sources are used by an AI and why they chose to use them.
Academic Foundation
A key reference in this area is the paper “GEO: Generative Engine Optimization” by Pranjal Aggarwal, Vishvak Murahari, Tanmay Rajpurohit, Ashwin Kalyan, Karthik Narasimhan, and Ameet Deshpande (2023). In this study, the authors:
- Proposed a creator-centric optimization framework
- Introduced GEO-bench, a benchmark dataset of 10,000 queries to evaluate generative engine performance.
- Demonstrated that certain targeted content optimizations can improve visibility in Perplexity.ai’s responses by up to 37%
- Show improvements of up to 40% across generative engines in general
(arXiv link).
Why GEO Matters Now
Generative engines are becoming a key discovery mechanism for information, and their influence is only projected to grow. For a company like Zephyr Cloud, which invests heavily in thought leadership through blogs, case studies, and whitepapers, failing to optimize for these systems could mean missing out on significant audience reach and hindering the customer’s journey through the marketing funnel.
Specifically, GEO is critical for moving prospects from awareness to consideration, the stage where written marketing materials drive conversion. Without GEO optimization, we’re failing to capture new business.
Real-World Impact
Consider this scenario: A user queries Perplexity.ai about “emerging trends in cloud optimization.” Even if Zephyr Cloud has a detailed blog post on this topic, the engine may not cite it if the post’s structure, metadata, or language does not align with GEO best practices.
However, with targeted adjustments, Zephyr could significantly increase citation odds by:
- Including relevant numerical data
- Citing credible sources
- Structuring content in concise sections
Proposed Solution
We propose implementing an AI-powered GEO diagnostic tool into Zephyr Cloud’s content creation workflow. Here’s how it would work:
- Content Creation - The marketing team drafts a blog post or article as usual.
- AI-GEO Analysis - The content is passed through our AI tool, which analyzes factors known to improve generative engine visibility, such as:
- Inclusion of relevant statistics or numerical data.
- Clear subheadings and concise sectioning.
- Use of direct, unambiguous language.
- Proper citation of credible sources.
- Optimization Recommendations - The tool outputs a report with suggested edits, highlighting which sections could be improved to increase GEO score.
- Testing and Iteration - Optimized content is published, and inbound traffic from generative AI platforms is tracked over time to measure improvements.
This cycle would allow Zephyr Cloud to continuously refine its content for generative visibility adapting in real time as engines evolve.
Challenges and Considerations
While GEO presents significant opportunities, it also involves certain challenges:
- Trial and Error - Many generative engines, especially emerging ones, are opaque in how they select sources. Optimization often requires iterative experimentation.
- Platform Variability - Techniques that work well for Perplexity.ai may not translate directly to ChatGPT or other engines.
- Evolving Algorithms - As with SEO, GEO is not static. Best practices may change rapidly, requiring ongoing monitoring and updates.
Expected Impact
By implementing a systematic GEO approach, Zephyr Cloud can expect:
- Increased AI-driven traffic - Better positioning within AI-generated responses.
- Higher brand authority - Being cited by trusted AI engines reinforces brand credibility.
- Future-proof marketing - Staying ahead of the curve as generative search adoption accelerates.
In the long run, this approach positions Zephyr not only to adapt to changes in how audiences discover content but also to lead in an emerging field that is likely to become standard practice for content marketing in the AI era.
In Practice
Below you will find a demonstration of the AI optimization site as well as the link so you may try it yourself!
Full access to AI features requires Claude API key