The Displacement of Legacy Authority: A 125-Hour AI Share of Voice Case Study
In AI search, domain authority is no longer the safety net it once was.
This case study documents what happened when a boutique brand, Cassie Clark Marketing, updated a single piece of content specifically for AI search using the FSA Framework (Freshness, Structure, and Authority). Within hours, legacy SEO publishers like Search Engine Journal lost their place in AI-generated answers.
The content update did not include a backlink or paid promotion. The only change was how the content was structured, updated, and positioned for large language models.
Within 96 hours, AI Share of Voice (AI SoV) rose from a baseline of 26.67% to a peak of 72.7%. More importantly, that visibility held. Over the following days, the legacy publishers that previously dominated the prompt dropped to 0% visibility while the updated content remained the primary source.
This isn’t a story about beating big brands and “winning the Internet.” Instead, it’s a look at how AI systems actually choose what to cite—and why authority in generative search is something brands can earn continuously, not something they inherit.
What is AI Share of Voice (AI SoV)?
AI Share of Voice (AI SoV) measures how often a brand appears inside AI-generated answers compared to other sources when a specific prompt is asked.
AI search engines assemble answers, meaning that visibility is distributed across a small set of sources the model decides to trust enough to cite. This differs from traditional SEO, where rankings and clicks are tracked.
At its simplest, the formula looks like this:
AI SoV = (Number of times a brand is cited ÷ Total citations across all brands) × 100
If five brands are cited in response to a prompt, AI SoV shows who gets how much of the answer—and who gets left out entirely.
AI SoV behaves differently from keyword rankings. When one brand gains visibility, another usually loses it. In practice, this creates a winner-take-most dynamic. The primary source often captures the majority of the attention, while secondary sources receive fractional mentions.
In this study, AI SoV was tracked over time to see whether early gains held or if the system shifted citations as it re-evaluated sources.
The Methodology: A Controlled Experiment
To keep the experiment focused, only one variable was introduced.
The prompt: “I need to do a baseline AI visibility audit. Help.”
This prompt was chosen intentionally. It reflects a real, early-stage research query. And it’s exactly the kind of question buyers ask before they know which tools, frameworks, or vendors to trust.
The variable: Only one page was updated: Baseline Audit for AI Search Visibility (Template + KPI Definitions). No other content was edited or promoted
The tracking conditions: All prompts were run in logged-out, private browsing sessions across both desktop and mobile to eliminate personalization and device-based bias.
The measurement window: AI SoV was tracked continuously from December 17 to December 22, capturing more than 125 hours of system behavior, including multiple re-ranking cycles.
The Data: The Three Phases of Displacement
AI Share of Voice doesn’t move all at once. It shifts in stages as the system reassesses which sources belong in an answer. Over the 125-hour tracking window, four distinct phases emerged.
Phase 1: The Baseline
Before any updates were made, Cassie Clark Marketing’s AI SoV sat at 26.67%, roughly tied with competitors like TheRankMasters. Citations were distributed across several sites, including legacy publishers, as the system pulled from familiar but loosely aligned sources.
Phase 2: The Rapid Re-Ranking
Within two hours of updating the page, Perplexity began reassessing sources. Cassie Clark Marketing’s AI SoV jumped to 45.45%, signaling that the updated content had been re-evaluated and elevated.
Phase 3: Consolidation
By December 20, it appeared that Perplexity had finished rebalancing. Cassie Clark Marketing’s AI SoV peaked at 72.7%, and the updated content was locked in as the primary source.
At the same time, legacy publishers—including Search Engine Journal—were quietly nudged to 0% visibility for the tracked prompt.
Phase 4: Retention
The final signal came after the spike. By December 22, Cassie Clark Marketing’s AI SoV stabilized at 67%, holding steady across multiple re-ranking cycles. This retention shows the system didn’t just experiment with a new source—it learned it as a canonical source.
The “Purple Line” Analysis
AI search visibility is zero-sum. As Cassie Clark Marketing’s AI SoV climbed to 72.7%, other sources lost visibility.
During both the peak and retention phases, Search Engine Journal dropped out of the citation set entirely. Once the system selected a primary source, it consolidated around it rather than distributing visibility across competitors.
Why This Matters for Series A/B & Enterprise Brands
For Series A/B and enterprise brands, this case study highlights a shift that many teams have not yet fully accounted for.
1. Legacy authority is more fragile than it looks.
High domain authority still matters, but it’s no longer a shield. In AI search, LLMs do not appear to refer to reputation alone. If content is stale, loosely structured, or misaligned with how the model assembles answers, even well-known publishers can be bypassed.
This case shows that legacy brands aren’t being “penalized,” but rather outcompeted by content that’s easier for AI systems to interpret and reuse.
2. The speed of change has accelerated dramatically.
Traditional SEO updates play out over weeks or months. In contrast, AI visibility can shift in a matter of hours.
Once a model detects clearer structure, fresher context, or stronger authority signals, it can rapidly rebalance which sources it selects. That means visibility risk—and opportunity—now exists on a much shorter timeline than most teams are used to managing.
3. AI search is winner-take-most.
In AI-generated answers, being cited once isn’t enough. The “Primary Source” typically captures the majority of the answer, while secondary sources receive only partial mentions or disappear altogether. From a buyer’s perspective, that primary source becomes the default frame of reference.
In practice, this means AI search visibility boils down to whether your brand is present and trusted at the moment decisions are being formed.
What This Experiment Suggests
If a boutique agency can displace Search Engine Journal in under 96 hours, it raises an important question about where your brand stands in AI-generated answers today. And whether that visibility would hold up if the system were re-evaluated tomorrow.
I’m opening AI visibility audits for teams that want a clear and honest read on where they stand. Book your AI Visibility Audit.
If we haven’t met yet….
Hi, I’m Cassie, a fractional content strategist for startups and enterprise brands who focuses on AI search optimization.
I build content programs that connect strategy with execution: clear positioning, systems, and content that actually drive revenue. If you’re ready to stop guessing and start growing, here’s how we can work together.
Want more insights like this? Subscribe to The Visibility Report, where I break down how AI engines interpret authority — and how you can show up in the results.
What is AI Share of Voice (AI SoV)?
AI Share of Voice (AI SoV) measures how often your brand appears inside AI-generated answers compared to other sources for a specific prompt. It reflects citation distribution inside answer engines, not rankings in a list of links.
How do you calculate AI Share of Voice?
A simple way to calculate AI SoV is: (brand citations ÷ total citations across all brands for the prompt) × 100. The goal is to track how citation share shifts over time as the model re-ranks and consolidates sources.
Is AI SoV the same as “rank” in ChatGPT or Perplexity?
No. AI SoV is not a static ranking position. AI answers are assembled, and citation visibility is distributed across a small set of sources. That distribution can shift as the system re-evaluates which sources to include.
Why can AI visibility be “zero-sum”?
Because AI answers rarely expand to include unlimited sources. When one brand becomes the primary source and takes a larger share of citations, other sources often lose share—or drop out of the citation set entirely.
What does it mean to “displace” a legacy publisher?
Displacement means the AI system stops selecting a previously cited source for the prompt and replaces it with a different primary source. It doesn’t necessarily mean the legacy site is “bad”—it means the system found another source that better fits how it wants to build the answer.
Does domain authority still matter for AI search visibility?
Traditional authority signals still matter, but they aren’t a guarantee. In generative search, freshness, structure, and clear relevance can outweigh legacy authority—especially when the model is selecting a single primary source.
How quickly can AI Share of Voice change?
In this case study, major shifts happened within hours, and peak visibility occurred within 96 hours. AI systems can re-rank quickly as they reassess sources and consolidate around a primary answer.
What should teams track besides a single peak result?
Retention matters more than a spike. Track whether visibility holds over multiple days and re-rank cycles. Sustained selection is a stronger signal than a short-lived jump in citations.
Can this method be replicated for other brands and prompts?
The exact results will vary by prompt, category, and content ecosystem. The repeatable takeaway is the approach: establish a baseline, change one variable, and track citation distribution over time to understand what the system consistently rewards.
What’s the next step if my brand has low AI visibility?
Start with a baseline AI visibility audit: identify the prompts that matter, document where you appear (and don’t), and map visibility gaps to specific content and authority signals. From there, prioritize updates using a framework like Freshness, Structure, and Authority.