How does online shopping affect offline price sensitivity?

arXiv:2506.15103 · econ.GN, q-fin.EC · Submitted 2025-06-18 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Economics. The day in economics: theory, mechanisms and the evidence beneath them.

Ruth: We're looking at the paper: "How does online shopping affect offline price sensitivity?".

Warren: Here's the gist of the paper. The rapid rise of e-commerce has fundamentally altered consumer behavior, leading researchers to investigate how online shopping adoption influences subsequent offline purchasing decisions.

Ruth: This study specifically examines whether consumers who adopt a retailer’s online channels exhibit increased price sensitivity in their later offline purchases with that same retailer.

Warren: Utilizing transaction-level data from a large Brazilian pet supplies retailer that operates both online and physical stores, the research contrasts "adopters"—customers who transitioned from an offline-only shopping pattern to incorporating online shopping—with "non-adopters" who remained exclusively offline.

Ruth: The methodology employed involves estimating a discrete choice logit model incorporating individual-level heterogeneity, followed by applying a staggered difference-in-differences (DID) approach to derive the Average Treatment Effect on the Treated (ATT).

Title and authors: Ruth: Moving on to how this study addresses its own limitations and what improvements the authors are suggesting for their work on "How does online shopping affect offline price sensitivity?".

Warren: I want to hear about those improvements, because if the identification method is weak, the whole result might not hold up when we try to apply it elsewhere.

Iris: The authors suggest using a staggered difference-in-differences framework that compares adopters with non-adopters before and after online adoption to formally estimate the average treatment effect on the treated.

Ruth: That’s a better way to get that causal estimate, especially when dealing with different groups of people adopting online channels at different times.

Warren: They also talk about using a minorization-maximization algorithm to estimate their discrete choice logit models because standard maximum likelihood estimation can be too slow when you have a lot of customer and brand fixed effects.

Iris: They mentioned that they should look at how switching costs explain the differences across categories, focusing on the relative increase in offline price elasticity between those groups.

Ruth: That’s interesting because it suggests that low switching costs are important—the way we switch products online seems to dictate how much our offline prices matter more.

Warren: It means that for items where you can easily switch from one brand to another online, the impact of price changes in the physical store is probably even stronger.

Iris: They also point out that they should address endogeneity in prices by using cost as an instrument combined with a control function approach to deal with price interaction terms in their demand estimation.

Ruth: That sounds like a lot of technical work to fix, but it’s necessary if you want to make these findings more robust for real-world policy or pricing decisions.

The paper's summary: Warren: So, summarizing what the paper actually found in "How does online shopping affect offline price sensitivity?", we saw that the main finding is that offline price sensitivity increases after online adoption in three out of four product categories.

Ruth: That's consistent with the initial observation that consumers become more sensitive to prices when they start shopping online, especially for things like dog hygiene items.

Iris: The paper specifically highlights how this effect varies by category; dog hygiene had a seventy-two point seven percent increase, while cat hygiene had a thirty-four point seven percent increase.

Warren: They also found that the effect isn't uniform; dog food showed a smaller increase of eight point four percent, and there was no effect found for cat food in this analysis.

Ruth: It seems the paper’s central message is that online adoption is a key driver here, suggesting managers need to account for this price sensitivity when setting prices in physical stores.

Iris: The researchers used data from a large Brazilian pet supplies retailer that sold through both channels between January one two thousand nineteen and June thirty two thousand twenty-one to generate these findings <ref:2506.15103#pg3>.

Warren: So, the paper uses transaction-level data to show this relationship between online adoption and offline price sensitivity.

Ruth: And the core mechanism they suggest is that exposure to lower online prices makes in-store prices seem inflated by comparison, prompting greater scrutiny of different brands.

The paper's improvements: Ruth: So, wrapping up "How does online shopping affect offline price sensitivity?", the paper confirms that consumers become more price-sensitive in their offline shopping trips when they start shopping online.

Warren: And the key numbers show this effect is significant in three out of four product categories, with dog hygiene and cat hygiene showing the largest increases.

Iris: The implication for business is clear: pricing strategies for physical stores need to account for this shift driven by online behavior.

Ruth: We have to remember that these results are based on a large Brazilian pet supplies retailer, and the study itself has some limitations regarding what the method doesn't capture.

Warren: It’s a solid piece of empirical work showing how digital channels influence traditional retail behavior through price perception in physical spaces.

Iris: It sets up future work by suggesting more advanced causal inference methods to better isolate this effect across different adoption groups.

Ruth: That’s it for this paper, "How does online shopping affect offline price sensitivity?". We've seen how online adoption directly links to a change in offline price sensitivity.

Conclusion: Ruth: So we’ve been looking at "How does online shopping affect offline price sensitivity?". The main point is that when consumers start shopping online, they actually become more sensitive to prices when they go into a physical store.

Warren: That’s what the study showed, right? It means that for many people, the act of checking prices online changes how much attention they pay to those same prices later in person.

Iris: The numbers are interesting because it varies by product type; dog hygiene sees a big jump, but cat food doesn't show any effect in this analysis.

Ruth: Exactly. It’s not a blanket rule for every item across the board, which is important for anyone trying to set prices in a brick-and-mortar setting.

Warren: The study uses transaction data from a Brazilian pet supplies retailer to prove this connection between online adoption and offline price elasticity.

Iris: I noticed they did some heavy lifting on the identification method, using a staggered difference-in-differences approach to get that average treatment effect on the treated.

Ruth: That’s smart. It moves beyond just seeing a correlation and tries to isolate what online shopping *causes* in those offline purchases.

Warren: But they also flag some technical hurdles, like the need for more advanced methods to handle high-dimensional data when estimating their choice logit models.

Iris: Yeah, the authors admit that standard estimation can be computationally difficult with so many customer and brand effects involved.

Ruth: It’s a lot of work to get these kinds of causal estimates right, but it gives managers a clearer picture about how to price things in physical stores after digital adoption starts.

Warren: So, the takeaway is that online channels aren't just for convenience; they fundamentally change how much people react to price changes in other places.

Iris: It suggests that when you set prices physically, you need to think about the online shopping behavior of your customers.

Ruth: Right. And next time we look at papers, we should keep an eye on how these adoption effects interact with things like switching costs or even cost as an instrument in price modeling.

Warren: It’s a deep dive into consumer behavior that definitely has some practical implications for retailers everywhere.

University of Washington

econ.GN, q-fin.EC

Submitted: 2025-06-18

Updated: 2026-11-12

Comments: 105 pages

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 97/100

The gist: The rapid rise of e-commerce has fundamentally altered consumer behavior, leading researchers to investigate how online shopping adoption influences subsequent offline purchasing decisions.

Key concepts

Adopters vs. Non-adopters
This compares two groups: customers who switched from only buying in physical stores to also using the retailer's online channels (adopters), and those who stayed exclusively offline (non-adopters). This comparison is used to see if the online shopping behavior causes a change in their later offline purchasing habits.
Offline Price Sensitivity
This measures how much a consumer changes their buying behavior—specifically, how sensitive they are to price changes—when making a purchase in a physical store. The study found this sensitivity increases significantly for adopters after they start shopping online.
Discrete Choice Logit Model
This is a statistical tool used to predict consumer choices, such as which product or store a customer will choose. The researchers used this model to estimate how price and other factors influence the decision of whether a customer buys something offline.
Staggered Difference-in-Differences (DID)
This is a causal method that compares changes in outcomes over time between groups that experienced different treatments at different times. Here, it helps researchers isolate the specific effect of adopting online shopping on later offline price sensitivity.

Terminology

Summary

The rapid rise of e-commerce has fundamentally altered consumer behavior, leading researchers to investigate how online shopping adoption influences subsequent offline purchasing decisions. p. 0

Where the paper says it

Shirsho Biswas, University of Washington* Hema Yoganarasimhan, University of Washington Haonan Zhang, University of Washington October 3, 2025 Abstract The rapid rise of e-commerce has transformed consumer behavior, prompting questions about how online adoption influences offline shopping. We examine whether consumers who adopt a retailer’s online shopping channels become more price-sensitive in their subsequent offline purchases with that retailer.

This study specifically examines whether consumers who adopt a retailer’s online channels exhibit increased price sensitivity in their later offline purchases with that same retailer. p. 0
Where the paper says it

Shirsho Biswas, University of Washington* Hema Yoganarasimhan, University of Washington Haonan Zhang, University of Washington October 3, 2025 Abstract The rapid rise of e-commerce has transformed consumer behavior, prompting questions about how online adoption influences offline shopping. We examine whether consumers who adopt a retailer’s online shopping channels become more price-sensitive in their subsequent offline purchases with that retailer.

Utilizing transaction-level data from a large Brazilian pet supplies retailer that operates both online and physical stores, the research contrasts adopters—customers who transitioned from an offline-only shopping pattern to incorporating online shopping—with non-adopters who remained exclusively offline. The methodology employed involves estimating a discrete choice logit model incorporating individual-level heterogeneity, followed by applying a staggered difference-in-differences (DID) approach to derive the Average Treatment Effect on the Treated (ATT). p. 0
Where the paper says it

We estimate a discrete choice logit model with individual-level heterogeneity, based on an algorithm that can handle both high-dimensional fixed effects and price endogeneity. We then apply a staggered difference-in-differences approach to the estimated price elasticities and obtain the Average Treatment Effect on the Treated (ATT).

The primary finding confirms that offline price sensitivity increases significantly after online adoption in three out of four product categories, with this effect being particularly pronounced in items characterized by low switching costs, such as pet hygiene products. The relative increase in offline price elasticity is heterogeneous across these categories: a substantial effect is observed for dog hygiene (72. p. 3

Where the paper says it

The relative increase in offline price elasticity, defined as the ATT relative to the pre-adoption average price elasticity of the treated group is heterogeneous across categories. We find a large effect in the dog and cat hygiene categories (where price elasticity increases by 72.

7%) and cat hygiene (34. p. 2
Where the paper says it

This allows us to track the offline purchases of online shopping adopters, both pre- and post-adoption, and also all purchases of those consumers who remained offline only. We focus on four major product categories: dog food, cat food, dog hygiene, and cat hygiene, which together account for more than 50% of the retailer’s total revenue.

7%), while the increase for dog food is significant but smaller (8. p. 3
Where the paper says it

7%, respectively). We also observe a significant but smaller increase in the dog food category (8.

4%), and no effect was found for cat food within this analysis. p. 6
Where the paper says it

2 Dataset Construction We now discuss how we construct the dataset used for our empirical analysis. Selection of Product Categories: We focus on the four top-selling non-medical categories: Dry Dog Food, Dry Cat Food, Dog Hygiene (i.

The paper establishes online adoption as a key driver of offline price sensitivity, offering crucial insights from a managerial perspective regarding pricing strategies, underscoring the necessity of accounting for this effect when setting prices in physical stores. The overall treatment effect estimates consistently suggest that consumers become more price-sensitive in their offline shopping trips when they start shopping online. Robustness checks were conducted to ensure the reliability of these findings, including tests for parallel pre-trends, alternative demand specifications using a binary online adoption variable, and matching techniques such as Inverse Propensity Weighting (IPW) and Propensity Score Matching (PSM), all of which yielded consistent results.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Enhanced Causal Inference for Treatment Timing: The AI system can implement a staggered difference-in-differences approach using a Callaway and Sant’Anna estimator to accurately calculate the Average Treatment Effect on the Treated (ATT) across different adoption cohorts, addressing biases inherent in two-way fixed effects.

  2. High-Dimensional Model Estimation: The system should incorporate a Minorization-Maximization algorithm to efficiently estimate discrete choice logit models with high-dimensional customer and brand fixed effects, which is computationally infeasible with standard maximum likelihood estimation.

  3. Robust Price Elasticity Measurement: The AI can derive customer-month level own-price elasticities using a logit model where price sensitivity is a function of the extent to which a consumer adopts online shopping, allowing for the quantification of the effect of interest on offline price elasticity.

  4. Addressing Price Endogeneity: The system must utilize cost as an instrument in conjunction with a control function approach to address endogeneity in prices and price interaction terms within the demand estimation literature.

  5. Counterfactual Analysis via Staggered DID: The system can employ the staggered difference-in-differences framework, comparing adopters with non-adopters before and after online adoption, to formally estimate the Average Treatment Effect on the Treated (ATT) of online shopping on offline price elasticity.

  6. Heterogeneity in Effects Modeling: The AI can analyze how switching costs explain heterogeneity across categories by examining the relative increase in offline price elasticity, noting that The relative increase in offline price elasticity, defined as the ATT relative to the pre-adoption average price elasticity of the treated group is heterogeneous across categories.

Abstract

The rapid rise of e-commerce has transformed consumer behavior, prompting questions about how online adoption influences offline shopping. We examine whether consumers who adopt a retailer's online shopping channels become more price-sensitive in their subsequent offline purchases with that retailer. Using transaction-level data from a large Brazilian pet supplies retailer operating both online and offline, we compare "adopters" -- customers who began shopping online after a period of offline-only purchasing -- with "non-adopters" who remained offline-only. We estimate a discrete choice logit model with individual-level heterogeneity, based on an algorithm that can handle both high-dimensional fixed effects and price endogeneity. We then apply a staggered difference-in-differences approach to the estimated price elasticities and obtain the Average Treatment Effect on the Treated (ATT). We find that offline price sensitivity increases significantly after online adoption in three out of four product categories. Our results underscore the importance of recognizing cross-channel effects in consumer behavior and contribute to the literature on pricing and multichannel retailing by identifying online adoption as a key driver of offline price sensitivity.

Related papers