The Conversion Premium
5 independent data sources have measured the conversion behavior of AI-referred traffic. They used different methodologies, at different scales, in different time periods. They reached the same conclusion.
Adobe Digital Insights
Adobe Analytics is used by the majority of the top 100 U.S. internet retailers. Adobe's holiday reporting draws on more than 1 trillion observed U.S. retail site visits across 100 million SKUs and 18 product categories.1
During the 2025 holiday season, traffic referred from AI platforms converted at a rate 31% higher than all other traffic sources.2 In July 2024, AI referral traffic was 43% less likely to convert than traditional sources. By February 2025, that gap had closed to 9%. By the 2025 holiday season, AI referrals led all channels in conversion rate.3
On peak shopping days, AI-referred visitors converted at rates 54% higher on Thanksgiving and 38% higher on Black Friday. Revenue per visit from AI referrals was up 254% year-to-date. AI-referred visitors were 33% less likely to bounce, an improvement of 14 percentage points since the start of 2025.4
Seer Interactive
Seer Interactive's June 2025 data shows platform-specific conversion rates: ChatGPT referrals at 15.9%, Perplexity at 10.5%, Claude at 5%, Gemini at 3%. Google's organic conversion rate over the same period: 1.76%.5
Go Fish Digital
Go Fish Digital's agency case study documented AI referral leads converting at 25x the rate of leads from traditional search, with results achieved in under 90 days.6
Salesforce
Salesforce tracked 1.5 billion global shoppers during the 2025 holiday season. Shoppers referred from AI-powered search channels converted 9x more often than those arriving through social media.7 AI and agents influenced 20% of all retail sales during the holiday season, fueling $262 billion in global revenue.8
Similarweb
AI search referrals to transactional sites converted at approximately 7%, compared to approximately 5% from Google referrals.9 By the end of 2025, AI platforms generated 2 billion referral visits, a 778% year-over-year increase.10
The Size of the Value Pool
The conversion premium exists. The next question is how much commerce flows through it.
Holiday 2025 saw $257.8 billion in U.S. online spending and $1.29 trillion globally.11 AI traffic to retail sites was up 693% year over year during the holiday period. Growth extends across verticals: retail up 693%, travel up 539%, financial services up 266%, technology up 120%.12
Researchers from Carnegie Mellon and Stanford estimated in August 2025 that AI search generates $97 billion in annual consumer surplus in the United States. The average user would require $98 per month to give up access.13
Bain found that 80% of consumers rely on zero-click results for at least 40% of their searches, reducing organic web traffic by an estimated 15% to 25%.14 Retailers that deployed AI agents saw 6.2% year-over-year sales growth versus 3.9% for those that did not.15
Gartner surveyed 377 U.S. consumers in January 2026. 31% consider more product options because of AI overviews, while 7% consider fewer. 31% spend more time searching because of AI summaries. Over 2/3 continue past Google's AI Overview.16
What Each Position Is Worth
The conversion premium is proven. The value pool is measured. The remaining question is what a specific position in a ChatGPT response is worth in dollars. That calculation requires 3 inputs, all of which are now available from published research.
Profound's November 2025 algorithm analysis found that ChatGPT surfaces 3 to 4 brands per response to purchase-intent queries.17 This creates a finite set of positions: brands either appear in positions 1 through 4, or they do not appear at all.
Backlinko's study of 4 million search keywords produced click-through rate curves for every position on a results page.18 Google's raw CTR for positions 1 through 4 is approximately 39%, 19%, 10%, and 7%. Normalized to a 4-position universe, those shares become 52%, 25%, 14%, and 9%. In the absence of AI-native click-through data, this curve is used as a proxy for how attention distributes across ChatGPT's positions. It is an inference from Google SERP behavior, not a direct measurement of ChatGPT behavior, and it is applied conservatively: ChatGPT's conversational format presents brands sequentially within a single answer rather than across a scannable page, which concentrates attention at top positions at least as tightly as Google's curve. These shares therefore represent a conservative floor and will be refined as AI-native data emerges.
Seer Interactive measured ChatGPT's conversion rate at 15.9%, based on 11,000 AI-referred sessions.19 That rate is 9x Google's organic conversion rate of 1.76%.
Omnisend's 2025 analysis of 150,000 e-commerce brands found the average U.S. order value rose to $182, a 22% year-over-year increase.20
With those data points, the per-position value becomes calculable. The following example uses industry benchmarks.
Worked Example
Consider a brand with 100,000 monthly organic website visitors and the U.S. average order value of $182.20
Estimate AI-referred visitors
100,000 monthly organic visitors × 3% AI referral rate = 3,000 AI visitors per month
The 3% figure is AiRR's 2026 baseline assumption, reflecting Conductor's late-2025 measurement of AI-originated traffic at approximately 1% of total visits and its ~1% month-over-month growth trajectory. Brands with Google Analytics referral data should use their actual number.21
Calculate the brand's TAAIR
3,000 AI visitors × 15.9% ChatGPT CVR × $182 AOV = $86,814 per month
$86,814 × 12 = $1,041,768 per year
This is the brand's TAAIR: the total annual AI-influenced revenue generated by its AI visitors. Position share determines how much of it the brand captures.
Apply the position share curve
| Position | AI Share | Monthly Value | Annual Value |
|---|---|---|---|
| Position 1 | 52% | $45,143 | $541,719 |
| Position 2 | 25% | $21,704 | $260,442 |
| Position 3 | 14% | $12,154 | $145,848 |
| Position 4 | 9% | $7,813 | $93,759 |
| Not Appearing | 0% | $0 | $0 |
A brand at Position 3 captures an estimated $145,848 per year from AI-influenced traffic on that prompt. A brand at Position 1 captures $541,719. The gap between Position 3 and Position 1 is $395,872 in annual revenue.
A brand that does not appear at all captures $0. For this example brand, the difference between not appearing and owning Position 1 is $541,719 per year on a single prompt.
Sensitivity by Average Order Value
The worked example uses $182, the U.S. e-commerce average. Brands in higher or lower AOV categories should recalculate with their own inputs. The table below shows how Position 1 and Position 4 annual values scale at the same 100,000-visitor baseline across a representative AOV range.
| Average Order Value | Position 1 Annual | Position 4 Annual |
|---|---|---|
| $100 | $297,648 | $51,516 |
| $182 (example) | $541,719 | $93,759 |
| $500 | $1,488,240 | $257,580 |
The Formula
The worked example above uses industry benchmarks. Any brand can calculate its own position value using 2 formulas and 3 variables.
TAAIR is the total annual revenue flowing through AI search for a given brand. It is the full value pool: the sum of all AI-influenced revenue generated by the brand's AI visitors, distributed across positions 1 through 4. A brand at Position 1 captures 52% of its TAAIR. A brand at Position 3 captures 14%. A brand that does not appear captures 0%. The shares sum to 100% because TAAIR represents the entire pool.
To calculate the revenue gap between any 2 positions, subtract:
Revenue Gap = TAAIR × (S₁ − S₂)A brand with 200,000 monthly visitors and a $220 AOV:
TAAIR = 200,000 × 3% × 15.9% × $220 × 12 = $2,518,560
Revenue at Position 3: $2,518,560 × 14% = $352,598
Revenue at Position 1: $2,518,560 × 52% = $1,309,651
Revenue Gap: $2,518,560 × (52% − 14%) = $957,053
What Position Share Means Across Multiple Prompts
The position share variable represents a brand's average visibility across its full portfolio of purchase-intent prompts. A brand that holds Position 1 on every relevant query captures 52% of its TAAIR. A brand that holds Position 1 on half its queries and Position 3 on the other half captures a blended share of 33%. The formula does not require counting individual prompts. It requires knowing where the brand sits on average.
A brand's TAAIR is fixed. It is determined by visitor volume and order value. What changes is the share the brand captures, and that share depends on its blended position across every prompt where a customer might find it or miss it. Measuring that blended position is the variable no brand currently has.
The ChatGPT Conversion Multiplier
The formula uses 15.9% as the ChatGPT conversion rate and 1.76% as Google's organic baseline. That produces a 9x multiplier for the average brand. But a brand can calculate its own multiplier by dividing 15.9% by its specific Google conversion rate.
15.9% ChatGPT CVR ÷ Your Google CVR = Your conversion multiplierA brand converting at 1% on Google has a 15.9x ChatGPT premium. A brand converting at 3% has a 5.3x premium. The lower a brand's organic conversion rate, the more disproportionately valuable its AI position becomes. The brands with the most to gain from AI search are the ones that convert worst on Google.
The Missing Variable
The math works. Every number in the TAAIR formula comes from a published, citable source. A brand can calculate its total AI-influenced revenue pool with 2 inputs from its own analytics: monthly organic visitors and average order value. It can then model what it captures at any position share.
The only variable a brand cannot supply is S. It cannot confirm what position it holds on a given prompt. It cannot track whether that position changed last week. It cannot compare its position across different prompt phrasings. It cannot benchmark its position against competitors. It cannot determine whether an optimization investment moved it from Position 4 to Position 2 or from Position 2 to Position 4. The TAAIR is calculable. The position share is not.
In February 2026, Profound closed a $96 million Series C at a $1 billion valuation. Total funding: $155 million across 4 rounds. Khosla Ventures, Kleiner Perkins, Sequoia, and Lightspeed all participated.22 Profound serves more than 700 enterprise customers, including approximately 10% of the Fortune 500.23
MIT's Project NANDA surveyed 153 senior leaders and interviewed 52 organizations. Their estimate: enterprises have collectively invested $30 to $40 billion in AI search. 95% of those organizations report zero measurable return on the P&L.24 70% of enterprise AI search budgets flow to sales and marketing functions.25
The market is spending to optimize AI positions it cannot measure, in a channel where the conversion premium is 9x, the value pool is $262 billion, and the TAAIR is calculable to the dollar. The only missing variable is the position share itself.
What a position is worth, and why nobody can measure it yet
This paper set out to answer a specific question: what is a position worth in AI search?
The answer is calculable. For a brand with 100,000 monthly organic visitors and the U.S. average order value of $182, Position 1 on a single ChatGPT prompt is worth $541,719 per year. Position 3 is worth $145,848. The gap between them is $395,872. Not appearing is worth $0.
Those numbers are modeled estimates, not live analytics. But every input comes from a published source. Seer Interactive measured the 15.9% conversion rate. Profound documented the 3-to-4 brand response structure. Backlinko produced the position share curve from 4 million keywords. Omnisend provided the $182 average order value from 150,000 brands. The 3% AI referral rate is AiRR's 2026 baseline, benchmarked against Conductor's late-2025 measurement of ~1% and its monthly growth trajectory. The position share curve is a proxy inferred from Google SERP behavior; real ChatGPT attention data does not yet exist at scale, and closing that gap is AiRR's core measurement mission going forward. Change any input and the TAAIR recalculates, but the structure holds. The formula in this paper lets any brand run its own numbers.
What concerns me is what happens when you multiply this across the real economy. The example uses 1 brand. An enterprise brand operates across dozens of product categories, hundreds of prompt variations, and multiple AI platforms. Its TAAIR is not $1 million. It is a figure that belongs on the balance sheet. And no one is calculating it.
The conversion data makes this urgent. 18 months ago, AI referral traffic was 43% less likely to convert than traditional sources. Today it is 31% more likely. Revenue per visit is up 254%. The channel that nobody was measuring 2 years ago now converts at 9x the rate of the channel that receives most of the marketing budget. Every month a brand operates without knowing its position share in AI search, it is making resource allocation decisions without knowing what share of its TAAIR it actually captures.
The MIT NANDA data gives this a price tag. $30 to $40 billion in enterprise AI investment, 95% reporting zero measurable return. These are not companies that failed to invest. They are companies that invested without knowing their TAAIR or their position share. 70% of those budgets went to sales and marketing. The functions that need this data the most are the ones spending the most without it.
The math to value a position exists. The TAAIR is calculable from 2 inputs a brand already has. The position share curve is sourced from published research. The missing variable is the position itself. Whoever solves that measurement problem does not just sell a product. They create the denominator that every AI marketing investment is evaluated against.
The most important number in marketing is what a position is worth in AI search. This paper provides the formula. The question is whether brands will calculate their TAAIR before or after their competitors do.
Calculate your actual TAAIR
The AiRR Score platform tells you your actual position share across every prompt that drives your business. Know your S. Calculate your real number.
Perlman, S. (2026). What a Position Is Worth in AI Search: The Economics of Visibility When There Is No Page Two. AiRR Score Research. AI Reach Rank Inc. Retrieved from airrscore.com/what-a-position-is-worth.html
Get the next paper first
New research publishes quarterly. One email per paper, no marketing noise.