Monetizing Intelligence: Advertising and Subscription Design on LLM Platforms

Yuhang Ji; Fasheng Xu; Tianjun Feng

econ.THOct 23, 2026

Monetizing Intelligence: Advertising and Subscription Design on LLM Platforms

Authors: Yuhang Ji, Fasheng Xu, Tianjun Feng

Organizations: School of Management, Fudan University · School of Business, University of Connecticut

Abstract

Large language model (LLM) platforms introducing advertising must decide how much capability ad revenue should fund and what consumers should pay to avoid ads. We study a monopolistic platform that jointly chooses costly response quality, advertising intensity, and subscription prices when advertising reduces user value but better service raises advertising returns. Three findings follow. First, when partly ad-supported and ad-free plans coexist, advertising can finance useful service for consumers who receive none under pure subscription. Lower effective quality in middle plans weakens the alternatives available to premium users, allowing higher prices for unchanged ad-free service. Some fully ad-supported users gain utility, while partly ad-supported and ad-free users lose: broader access can coexist with utility polarization. Second, an intermediate advertising cap can maximize consumer surplus by preserving some revenue that finances service while limiting degradation. Third, cheaper compute can raise quality and reduce ad intensity, yet sufficiently large cost reductions can increase aggregate advertising revenue above its initial level because better service raises the value of the remaining advertised interactions. An extension shows how advertising-funded upgrades can bring plans to the same maximum capability and advertising exposure, eliminating the basis for different prices. These results make menu design central to evaluating LLM advertising: the consequences extend beyond users who see ads to the alternatives and prices faced by those who do not.

Figures & tables

Panel of Figure 1 — Figure 1: The Impact of Ad Degradation ( $\delta$ ) on Platform Allocations ( $\bar{r}=0.1$ , $\lambda=0.2$ )

Panel of Figure 2 — Figure 2: The Impact of Ad Profitability ( $\bar{r}$ ) on Platform Allocations and Pricing ( $\lambda=0.2$ , $\delta=0.6$ )

Panel of Figure 3 — Figure 3: Quality and Utility: Hybrid Monetization versus Pure Subscription ( $\lambda=0.2$ , $\bar{r}=0.15$ , $\delta=0.6$ )

Panel of Figure 4 — Figure 4: Advertising Caps and Consumer Surplus ( $\lambda=0.5$ , $\bar{r}=0.18$ ; Panel (a): $\delta=0.9$ )

Appendix figures & tables5 assets

Supplementary material from the paper’s appendix.

Appendix

Ad Degradation ( $\delta$ )	Quality-Driven Ad Premium ( $\lambda$ )	Base Ad Rev ( $\bar{r}$ )	User Segment ( $\theta$ )	Ad Load ( $\alpha^{*}$ )	Quality ( $q^{*}$ )
$(0,\frac{1}{2}]$	$(0,\delta)$	$(0,\delta-\lambda)$	$\left[0,\frac{1}{2}-\frac{\lambda}{2(1-\delta)}\right)$	1	0
$(0,\frac{1}{2}]$	$(0,\delta)$		$\left[\frac{1}{2}-\frac{\lambda}{2(1-\delta)},\frac{1}{2}+\frac{\lambda}{2\delta}\right)$	1	$(2\theta-1)(1-\delta)+\lambda$
$(\frac{1}{2},1)$	$(0,1-\delta)$		$\left[\frac{1}{2}+\frac{\lambda}{2\delta},\theta_{2}\right)$	$\alpha^{int}$	$q^{int}$
$(\frac{1}{2},1)$	$(0,1-\delta)$		$\left[\theta_{2},1\right]$	0	$2\theta-1$
$(\frac{1}{2},1)$	$[1-\delta,\delta)$	$(0,\delta-\lambda)$	$\left[0,\frac{1}{2}+\frac{\lambda}{2\delta}\right)$	1	$(2\theta-1)(1-\delta)+\lambda$
			$\left[\frac{1}{2}+\frac{\lambda}{2\delta},\theta_{2}\right)$	$\alpha^{int}$	$q^{int}$

Appendix

Table 1: Optimal Equilibrium Structure under Market Regularity

	Zero-quality ( $\alpha^{}=1,\,q^{}=0$ )	Low ( $\alpha^{*}=1$ )	Medium ( $\alpha^{int}\in(0,1)$ )	High ( $\alpha^{*}=0$ )
Quality $q^{*}$ as $\delta\uparrow$ (Prop. 5.1 )	upgraded a	$\uparrow$ if $\theta<\tfrac{1}{2}$ $\downarrow$ if $\theta>\tfrac{1}{2}$	$\downarrow$ if $\alpha^{int}>\tfrac{1}{2}$ $\uparrow$ if $\alpha^{int}<\tfrac{1}{2}$	unchanged
Price $T^{*}$ as $\bar{r}\uparrow$ (Prop. 5.2 )	unchanged	unchanged	$\downarrow$	$\uparrow$ b
Utility vs. pure subscription (Prop. 5.3 )	unchanged	$+$ c	$-$	$-$

Appendix

Table 2: Comparative Statics across Market Segments (Propositions 5.1 – 5.3 )

Panel of Figure 5 — Figure 5: Advertising Caps: The Underlying Mechanism ( $\lambda=0.5$ , $\bar{r}=0.18$ , $\delta=0.9$ )

Number of Positive-Quality Tiers	Quality-Driven Ad Premium ( $\lambda$ Interval)	Type $L$ ( $v_{L}=-4/3$ )	Type $M$ ( $v_{M}=1/3$ )	Type $H$ ( $v_{H}=1$ )
2	$\left(0,4/15\right]$	$\alpha_{L}^{}=1$ , $q_{L}^{}=0$ (Zero-Quality)	$\alpha_{M}^{}=\alpha^{int}_{M}$ $q_{M}^{}=q^{int}_{M}$	$\alpha_{H}^{}=0$ $q_{H}^{}=v_{H}$
3	$\left(4/15,3/5\right]$	$\alpha_{L}^{}=1$ $q_{L}^{}=\tilde{q}_{L}(1)$	$\alpha_{M}^{}=1$ $q_{M}^{}=\tilde{q}_{M}(1)$	$\alpha_{H}^{}=0,\ q_{H}^{}=1$
	$\left(3/5,4/5\right)$			$\alpha_{H}^{}=\alpha_{H}^{int},\ q_{H}^{}=q_{H}^{int}$
	$\left[4/5,14/15\right)$			$\alpha_{H}^{}=1,\ q_{H}^{}=1$
2	$\left[14/15,19/15\right)$	$\alpha_{L}^{}=1$ $q_{L}^{}=\tilde{q}_{L}(1)$	$\alpha_{\mathrm{Pool}}^{}=1$ , $q_{\mathrm{Pool}}^{}=\bar{q}$ (Top Pooling)
1	$\left[19/15,+\infty\right)$	$\alpha_{\mathrm{Pool}}^{}=1$ , $q_{\mathrm{Pool}}^{}=\bar{q}$ (Universal Pooling)

Appendix

Table 3: Phase Transitions in Positive-Quality Service Tiers (

\delta=0.8

\bar{r}=0.2

\theta_{M}=2/3

\bar{q}=1

)

Ad Degradation ( $\delta$ )	Quality-Driven Ad Premium ( $\lambda$ )	Base Ad Rev ( $\bar{r}$ )	User Segment ( $\theta$ )	Ad Load ( $\alpha^{*}$ )	Quality ( $q^{*}$ )
$(0,1)$	$[0,\min\{\delta,1-\delta\})$	$[\delta-\lambda,+\infty)$	$\left[0,\frac{1}{2}-\frac{\lambda}{2(1-\delta)}\right)$	1	0
			$\left[\frac{1}{2}-\frac{\lambda}{2(1-\delta)},\frac{1}{2}+\frac{\lambda}{2\delta}\right)$	1	$(2\theta-1)(1-\delta)+\lambda$
			$\left[\frac{1}{2}+\frac{\lambda}{2\delta},1\right]$	$\alpha^{int}$	$q^{int}$
$\left(0,\frac{1}{2}\right)$	$[\delta,1-\delta)$	$[0,+\infty)$	$\left[0,\frac{1}{2}-\frac{\lambda}{2(1-\delta)}\right)$	1	0
			$\left[\frac{1}{2}-\frac{\lambda}{2(1-\delta)},\frac{1}{2}+\frac{1-\lambda}{2(1-\delta)}\right)$	1	$(2\theta-1)(1-\delta)+\lambda$
			$\left[\frac{1}{2}+\frac{1-\lambda}{2(1-\delta)},1\right]$	1	$\bar{q}=1$

Appendix

Table 4: Optimal Equilibrium Structure outside the Regular Parameter Region

Online Advertising

Online advertising studies how digital ads are targeted, ranked, priced, measured, and delivered while balancing budgets, business goals, and user interests.

See all 88 papers

LLM Advertisement based on Neuron Auctions
May 8, 2026Peiran Yun, Wenxin Xu, Jiayuan Liu +4Mechanism Design Online Advertising
Tsinghua University · Carnegie Mellon University · Harvard University +1
Evaluating and Pricing Advertisements in AI-Generated Responses
Jul 30, 2026John L. Turner-Smith, Zimeng Huang, Yuhan Fu +2Mechanism Design Online Advertising
Carnegie Mellon University, Pittsburgh, PA, USA · College of AI, Tsinghua University, Beijing, China
NaiAD: Initiate Data-Driven Research for LLM Advertising
May 11, 2026Yihang Zhang, Zimeng Huang, Ren Zhai +2LLM Evaluation Online Advertising
Tsinghua University Beijing, China · College of AI Tsinghua University Beijing, China · Department of Literature, Arts and Communication Anhui International Studies University Anhui, China +1

May 8, 2026J/K move · Enter open · S save

LLM Advertisement based on Neuron Auctions

Peiran Yun, Wenxin Xu, Jiayuan Liu, Yihang Zhang+3

Tsinghua University · Carnegie Mellon University · Harvard University+1

Mechanism Design Online Advertising

Open paper AskPDF arXiv

As Large Language Models (LLMs) transition into conversational agents, generative advertising emerges as a crucial monetization strategy. However, embedding advertisements within unstructured LLM outputs introduces a critical trilemma: balancing advertiser payoffs, platform revenue, and user experience. Existing methods, such as prompt injection or rigid position slots, disrupt semantic coherence and lack a parametric framework for independent control, rendering rigorous mechanism design intractable. To bridge this gap, we introduce Neuron Auctions, a novel paradigm that shifts the auction object from the surface text space to the LLM's internal representations. Leveraging mechanistic interpretability, we identify brand-specific feed-forward network (FFN) neurons and demonstrate that competing brands activate within approximately orthogonal subspaces. This near-perfect independence allows us to define continuous, disentangled intervention budgets (specifically, neuron counts and amplification factors) as auctionable commodities. Building on this computational carrier, we design a continuous menu-based auction mechanism that naturally guarantees strategy-proofness and optimizes revenue for the platform. By explicitly incorporating a user utility penalty into the platform's optimization objective, our framework dynamically prices out overly aggressive interventions. Extensive experiments demonstrate that Neuron Auctions effectively preserve natural discourse quality while achieving an optimal alignment between commercial incentives and user satisfaction.

Abstract

Online Advertising

Explore similar work

LLM Advertisement based on Neuron Auctions

Evaluating and Pricing Advertisements in AI-Generated Responses

NaiAD: Initiate Data-Driven Research for LLM Advertising