This paper examines how a publisher should set reserve prices for real-time bidding (RTB) auctions when selling display advertising impressions through ad exchanges, a $595 billion and growing market. Conducting a field experiment to induce exogenous variation in reserve prices at a major publisher, we find that setting reserve prices increases the publisher’s revenues by 35% compared to using a zero reserve price. We also find empirical evidence that advertisers face a minimum impression constraint to ensure sufficient advertising reach.
Based on this insight, we develop a structural model of advertiser bidding behavior that incorporates impression constraints, allowing us to infer overall demand for advertising as a function of reserve prices. We then use this demand model to solve the publisher’s pricing problem. Accounting for minimum impression constraints in setting reserve prices yields a profit increase of 9 percentage points over a solution that does not incorporate the constraint. In a final field experiment, we validate our model’s predictions by showing that ad revenues tend to be highest near the model-predicted optimal reserve prices.

Platform-based digital ecosystems form the backbone of our interactions with the Internet. Over the past decade, they have witnessed significant growth, both in terms of industry footprint and academic research. Yet, the challenges associated with their operations, governance, and regulation continue to evolve. Our paper aims to aid researchers in better understanding these developments, such that they can contribute more effectively to active debates and open questions in this space. We structure our paper in two parts. First, we lay out emerging challenges in research topics related to platforms, both from an internal (platform design) and external (platform regulation) perspective. Then, we complement this by presenting challenges inherent in conducting empirical research in platform markets, both with and without the collaboration of the platforms themselves. Our insights highlight the importance of multidisciplinary and multi-method approaches to the study of digital ecosystems to get a full grasp of the value they create for firms and consumers.

Firms solicit online reviews to enhance sales and reduce mismatches, yet it is unclear whether solicitations generate reviews where buyers benefit most, since they target reviewers while value accrues through buyers. Using rich individual-level panel data on transactions, reviews, and returns from a large apparel e-commerce retailer, we examine alignment between reviewer response and buyer value. Leveraging natural variation in solicitations, we estimate a two-stage framework: Stage 1 measures how solicitations affect review generation across informational states, and Stage 2 quantifies how these induced reviews influence subsequent sales and returns. Additional reviews increase sales and reduce return rates by up to 0.25 percentage points for products with zero-to-low prior reviews. However, solicitations face a first-review barrier, failing to elicit reviews in these high-value states, and thereby creating a reviewer-buyer misalignment. Counterfactual simulations indicate that raising first-review likelihood by two percentage points increases annual net revenue by $65 per product, highlighting the value of targeting zero-review products.

Retail media, a fast-growing channel for digital advertising, surpassed $55 billion of ad spend in 2024. A common retail media format involves position auctions, in which advertisers bid for higher placements on a retailer’s product listing page. Advertiser bids are combined with a retailer-set quality score to determine the allocation of sponsored slots and the resulting payments. Quality scores boost certain advertisers’ positions and reduce their per-click price. Unlike search engine advertising, retail media position auctions can monetize sales commissions as well as clicks. This paper develops a quality score approach to effectively balance these monetization options.
To connect quality scores to retail revenues, the paper develops a structural model linking advertiser bids and revenues to the retailer’s quality score choices coupled with a machine learning model of consumer behavior. These models are estimated using auction-advertiser level data from a quality score experiment conducted at a mid-size US based retail marketplace. Policy simulations show that a quality score approach that balances clicks and commissions improves retailer profits by 7% and advertiser surplus by 42% over click-based approaches typically used by retailers, leading to a win-win outcome for both.

This paper summarizes the display advertising literature, organizing the content by the agents in the display advertising ecosystem, and proposes new research directions. In doing so, we take an interdisciplinary view, drawing connections among diverse streams of theoretical and empirical research in information systems, marketing, economics, operations, and computer science. By providing an integrated view of the display advertising ecosystem, we hope to bring attention to the outstanding research opportunities in this economically consequential and rapidly growing market.

This paper considers the monetization of online marketplaces. These platforms trade-off fees from advertising with commissions from product sales. While featuring advertised products can make search less efficient (lowering transaction commissions), it incentivizes sellers to compete for better placements via advertising (increasing advertising fees). We consider this trade-off by modeling both sides of the platform. On the demand side, we develop a joint model of browsing (impressions), clicking, and purchase. On the supply side, we consider sellers’ valuations and advertising competition under various fee structures (CPM, CPC, CPA) and ranking algorithms.
Using buyer, seller, and platform data from an online marketplace where advertising dollars affect the order of seller items listed, we explore various product ranking and ad pricing mechanisms. We find that sorting items below the fifth position by expected sales revenue while conducting a CPC auction in the top 5 positions yields the greatest improvement in profits (181%) because this approach balances the highest valuations from advertising in the top positions with the transaction revenues in the lower positions.

In 2019, the $60 billion display advertising market was shaken when Google transitioned from a second-price to first-price auction. Many reported rationales underpinned the change, including the increased transparency of the auction (an advertiser pays what they bid rather than relying on the auction partner to report the closing price), the facilitation of bidding across partners, greater allocative efficiency, higher revenues for publishers, and criticisms of Google’s so-called last look algorithm which purportedly enabled its exchange users to outbid competitors from competing exchanges. Consistent with industry views, Despotakis et al. (2021) theorize that the change to the first-price auction, precipitated by header bidding, leads to higher clearing prices.
Despite these beliefs, little empirical evidence exists to ascertain how the change affected auction display outcomes. Bidding became more complex as advertisers turned to various bid shading tools from different vendors to manage the complex task of optimizing their bids. Left unanswered is whether this change did lead to higher bid advertiser CPMs and greater publisher revenue.
Using data collected during the duration of the switch, we find that i) advertisers adjust bids quickly to changes in format, ii) the FPA does not increase allocative efficiency, iii) advertisers bid too low, and iv) publisher revenues suffer under FPA. We consider two possible explanations for these deviations from theoretical predictions: a) advertisers have difficulty computing optimal bids and b) advertisers do not fully account for competitive bidding.

A publisher in a display ad market sells its advertising inventory both via ad exchange auctions and via direct sales. Ad exchanges (e.g., DoubleClick Ad Exchange) are centralized marketplaces where publishers supply and advertisers demand display advertising impressions in real-time bidding (RTB) auctions. The direct sale involves the advance sale of a bundle of ad impressions directly to the advertiser at a fixed price. The important research questions are then i) how ad inventories should be priced (fixed price in the direct sale and the reserve price in the exchange channel), ii) how ad inventories should be allocated across distribution channels (direct or exchanges first), and iii) how much information should be revealed to advertisers about specific ad impressions.
Based on a two-stage game theory model, we show that the equilibrium channel choice depends on the coefficient of variation (CV) of the valuation distribution. Advertisers with higher CV prefer the exchange over the direct channel because the value of information revealed in the ad exchange (e.g., cookie) is high and they have a higher option value of waiting. We are extending this theoretical framework to an empirical context with the novel dataset collected both for the direct and the exchange channels from a large, premium publisher, ranked within U.S. top 10.

E-commerce platforms, including online marketplaces, often send out SMS messages for advertising and promotions. Using a large-scale dataset involving natural experiments, I find that SMS messages increase the average purchase amount significantly (i.e., short-term direct responses); but at the same time, they also increase consumers’ opting-out of marketing communications. I also find that less than 1% of consumers opt back in after making these opt-out decisions. Therefore, I examine the optimal SMS targeting strategy to maximize the long-term gain by considering both the rate of opt-out decisions and the effects on short-term direct responses.