Why keyword rankings don't work for AI search
In traditional SEO, you measure success with keyword rankings. You're position 3 for 'best dentist in Austin' and position 7 for 'Invisalign Austin TX.' Those positions are relatively stable, easy to track, and meaningful.
AI search doesn't work this way. There are no fixed positions. AI generates a fresh answer for every query, pulling from different sources each time. The same question asked twice might produce different recommendations. About 50% of sources cited in AI answers change month-to-month.
This means the entire measurement framework from traditional SEO — keyword rankings, SERP position, rank tracking — doesn't translate to AI search. You need new metrics. The two most important ones are Share of Voice and Citation Share.
Share of Voice: the big picture metric
Share of Voice (SOV) measures the percentage of total visibility your brand earns in a market or channel compared to competitors. It's not a new concept — marketers have tracked SOV in paid media and PR for decades. But in AI search, it takes on new importance.
The formula is simple: (Your brand's mentions or visibility / Total mentions or visibility of all brands) x 100.
If AI tools mention 50 different dental practices across 100 queries about dentists in your city, and your practice appears 12 times, your SOV is 12%. If your top competitor appears 18 times, their SOV is 18%. You know exactly where you stand.
Why SOV matters more than ever
Research from Nielsen has shown that brands with Share of Voice higher than their market share are more likely to grow revenue over time. This principle, established in traditional media, applies even more strongly in AI search.
In the zero-click era — where roughly 60% of Google searches end without a click to any website, according to SparkToro — your brand being named in an AI answer may be the only impression you get. There's no click-through to measure. There's no impression in the traditional sense. There's either a mention or silence.
SOV captures this new reality. It tells you how much of the AI conversation in your market includes your brand, regardless of whether it leads to a click or a website visit.
SOV across different channels
Share of Voice isn't just an AI metric. It's measured across multiple channels, and understanding the full picture matters:
- —Paid media: Your ad impressions relative to total ad impressions in your market
- —Organic search: Your keyword visibility relative to competitors in traditional SERPs
- —Social media: Your mentions, hashtags, and engagement relative to competitors
- —Public relations: Your editorial coverage and news mentions relative to competitors
- —AI answer engines: Your citations and mentions in ChatGPT, Perplexity, Gemini, and AI Overviews
The AI search channel is the newest — and for many businesses, the one with the most room to grow. While most businesses actively manage their paid media and organic search SOV, almost nobody is tracking their AI search SOV yet.
Citation Share: the content performance metric
While Share of Voice measures brand visibility, Citation Share measures how often your specific content is cited relative to all citations in a defined set.
The formula: (Number of citations attributed to your content / Total citations in the defined set) x 100.
Citation Share is more granular than SOV. It tells you not just whether AI mentions your brand, but whether it actually links to your content. This distinction matters because citations (links) are a stronger signal than mentions (name drops) alone.
Types of Citation Share
Citation Share can be measured in several ways, each useful for different purposes:
- —Brand citation share: How often your brand is cited relative to all brands in your category
- —Domain citation share: How often your website is cited relative to all domains in a topic
- —Topic citation share: Your visibility for a specific topic compared to all entities cited for that topic
- —Platform citation share: Your citation rate on a specific AI platform (ChatGPT vs. Perplexity vs. Gemini)
For local businesses, brand and domain citation share are the most actionable. They tell you how much of the AI answer space you're capturing in your market.
Citation Share vs. co-citation
One distinction worth understanding: Citation Share measures your proportion of total citations. Co-citation refers to two entities being cited together by a third source. If AI frequently cites both you and a competitor in the same answer, that's a co-citation pattern — and it tells you something about how AI sees your competitive relationship.
How to measure your AI visibility
Here's a practical approach to measuring both SOV and Citation Share for your business:
Step 1: Define your query set
Identify 20-30 queries your customers actually ask. Include a mix of general ('best dentist near me'), specific ('Invisalign provider in Austin'), and condition-based ('dentist for tooth pain Austin'). These become your measurement benchmark.
Step 2: Run queries across platforms
Test each query on ChatGPT, Perplexity, Gemini, and Google. Record two things for each response: (1) whether your brand was mentioned by name, and (2) whether your website was cited with a link.
Step 3: Calculate your metrics
SOV = responses mentioning your brand / total responses across all queries. If you appear in 15 out of 100 total query responses, your SOV is 15%.
Citation Share = citations to your domain / total citations across all responses. If there are 300 total citations in your 100 queries and 20 go to your site, your Citation Share is 6.7%.
Step 4: Track over time
Repeat monthly with the same query set. Look for trends, not single data points. A month where your SOV drops from 15% to 12% might be normal drift. Three consecutive months of decline signals a real problem.
What the data tells us about AI visibility
Our research reveals several patterns that make SOV and Citation Share essential metrics:
Only 30% consistency: Only about 30% of brands maintain visibility between consecutive AI responses. This means any single measurement is unreliable — you need to measure across multiple queries and over time to get an accurate picture.
57% recovery rate: About 57% of brands that disappear from AI answers eventually resurface. This is why monthly tracking matters — a temporary disappearance isn't the same as losing visibility.
40% durability advantage: Brands earning both mentions and citations are 40% more likely to recover after disappearing. This is why tracking both SOV (mentions) and Citation Share (links) gives you a complete picture.
50% monthly churn: About half of the sources cited in AI answers change every month. Your Citation Share will naturally fluctuate. The question is whether your trend line is up, down, or stable.
Using these metrics to make decisions
Numbers are only useful if they inform action. Here's how to use SOV and Citation Share practically:
Low SOV, low Citation Share: AI barely knows you exist. Priority: foundational work on website structure, dedicated service pages, and business information consistency.
Low SOV, moderate Citation Share: AI links to your content sometimes, but doesn't mention your brand by name. Priority: build third-party mentions — community presence, review profiles, industry mentions.
Moderate SOV, low Citation Share: AI mentions you but doesn't link to your content. Priority: improve your on-site content structure so AI has pages worth citing.
High SOV, high Citation Share: You're in great shape. Priority: maintain freshness, monitor for competitive substitution, and expand to cover new query topics.
Not sure where you stand? Our free AI Visibility Checker gives you a starting point in about 30 seconds. For a complete measurement system, our ongoing monitoring tracks SOV and Citation Share across all major AI platforms monthly.
Frequently Asked Questions
What's the difference between Share of Voice and Citation Share?
Share of Voice measures your brand's total visibility compared to competitors. Citation Share measures how often your specific content is cited relative to all citations. SOV is about brand presence; Citation Share is about content performance.
How do I measure my Share of Voice in AI search?
Identify 20-30 customer queries. Run them across ChatGPT, Perplexity, Gemini, and Google AI Overviews. Record mentions and citations. Your SOV is the percentage of responses where your brand appears.
What's a good Share of Voice percentage?
In a local market with 20-30 competitors, 15-25% would be strong. The key is your SOV relative to competitors — being mentioned twice as often as your nearest competitor is more meaningful than any absolute percentage.
Can I improve Share of Voice without a huge budget?
Yes. AI recommendations correlate with content structure and information quality, not ad spend. Independent businesses consistently outperform chains. The most impactful improvements are effort-intensive but don't require massive budgets.
How often should I measure AI Share of Voice?
Monthly for most businesses. About 50% of cited sources change month-to-month, so regular measurement catches trends early. Weekly checks on 10-15 core queries can spot competitive substitution faster.
Want to see how AI search sees your business?
Get a free AI visibility audit and find out where you stand.