Field Notes/Pricing Technical

Price elasticity modeling for marketplace sellers — why Shopee, Lazada, and TikTok Shop pricing decisions need it

A practical guide from DataGlass Labs Research on the most under-used lever in Southeast Asian e-commerce: knowing how customers actually respond to price changes.

May 7, 202614 min readBhum Soonjun · DataGlass Labs Research
SEA platform GMV
$157.6B

2025 Momentum Works estimate

Top-three control
98.8%

Shopee, Lazada, TikTok Shop/Tokopedia

Retail pricing upside
5-10%

Margin growth range reported by McKinsey

Cross-SKU cannibalization
22%

Average uplift borrowed from sibling pack sizes

Pricing

The pricing question every seller secretly fears

Every Shopee, Lazada, and TikTok Shop seller has had the same conversation with themselves at 11 pm the night before a major sale event. Should I drop the price by 10% or 20%? Should I run a Bundle Deal at "buy 2, save 15%" or "buy 3, save 25%"? Will dropping price by 5% give me 50% more units, or just shave 5% off my margin and change nothing?

The honest answer, in almost every seller's actual workflow, is nobody knows. The discount level gets picked by intuition, by what worked last 11.11, or by mirroring whatever the seller next door is doing. Then everyone watches the dashboard and hopes.

This is the gap that price elasticity modeling closes. It replaces guessing how customers will respond with estimating it from your own historical data, the platform's promotional history, and a small amount of deliberate experimentation. It is not a magic spreadsheet — it is a piece of analytical infrastructure. Built correctly, it is one of the highest-leverage investments a marketplace seller can make in pricing strategy [9][10].

This post explains, without math, what elasticity modeling actually does, why current seller workflows leave so much margin on the table, and how the lever applies concretely on Shopee, Lazada, and TikTok Shop.

What "elasticity" really means, in plain language

Strip away the textbook and price elasticity is one question: if I change my price, how much does demand change?

A high-elasticity product is price-sensitive: a 10% discount might pull 30% more units. Fast-moving consumer categories — beauty, fashion accessories, snack food — usually live here.

A low-elasticity product is price-insensitive: dropping the price by 10% might only pull 5% more units, which means the discount is mostly destroying margin. Specialty items, branded goods with loyal customers, and replenishment essentials usually live here.

Two products in the same shop, in the same category, can have very different elasticities. The seller's job is to act on that difference. The seller who discounts the elastic product hard and the inelastic product gently makes more money than the seller who discounts everything by the same 15%. That is the entire opportunity.

The catch is that elasticity is not visible to the naked eye. It does not appear on the Shopee Seller Centre dashboard. It is not in Lazada Business Advisor [25]. It is not in TikTok Shop Seller Center's Data Compass [24]. It has to be estimated — and estimating it well requires more than a rolling average of past sales.

What marketplace sellers are doing today (and why it underperforms)

In the absence of real elasticity estimates, sellers fall back on four heuristics. All are common; none works as well as people assume.

HeuristicWhat it looks likeWhy it underperforms
Copy-the-competitor pricingFind the bestseller in the niche, undercut by a few percentYou inherit a price set for their margin structure, inventory position, and ad spend — and race them to the bottom on yours
Flat-discount-everything during sale eventsUniform percent-off applied across the catalog for 9.9, 11.11, Lazada Birthday Sale, TikTok Mega SaleSubsidizes inelastic shoppers who would have bought at full price, and under-discounts the elastic SKUs that need a real cut to convert
Reference-price erosionRoutine weekly or weekend discountsTrains customers never to pay full price; the reference price drifts down and the seller's only remaining lever is deeper, more frequent cuts
Bundle Deal by gut feel"Buy 3, save 15%" because 3 felt rightMisses the optimal threshold (sometimes 2, sometimes 5) and the optimal discount (sometimes 8%, sometimes 22%) — leaving meaningful contribution unclaimed

The aggregate effect of these four heuristics is what we call the decision-intelligence gap: a structural margin loss that compounds quietly across thousands of small pricing choices. In internal DataGlass benchmarking across Southeast Asian marketplace cohorts, 2025-2026, this gap often separates single-digit-margin shops from healthier mid-teens-margin shops selling similar goods on the same platform.

What elasticity modeling actually gives you

A working elasticity model is not a single number per product. It is a conditional demand curve — a relationship that says at this price, with this inventory, on this day of the week, in this category, with this much ad pressure, the expected units sold is X, with a confidence interval. It produces three things sellers cannot get from a dashboard.

  • Per-SKU sensitivity classification. The model tells you which SKUs are price-sensitive and which are not. That alone changes how you allocate discounts during a sale event.
  • Response-curve shape, not just the slope at one point. The optimal discount is rarely 10% more than last time; it depends on how the curve bends around your current price. Modern systems estimate this from historical promotion variation, partial pooling across similar SKUs, and deliberate price tests [17][18].
  • Calibrated uncertainty. The model gives you confidence bands around the elasticity estimate. A high-confidence elastic estimate lets you discount aggressively; a low-confidence estimate tells you to run a small holdout test before committing the whole catalog.

Three use cases on Shopee, Lazada, and TikTok Shop

Use case 1 — Shopee Bundle Deal optimization

Shopee's Bundle Deal is a quantity-conditioned discount: a customer pays full price for one unit but unlocks a percent-off when they reach a threshold [5]. This is second-degree price discrimination — elastic, multi-unit shoppers self-select into the discount; inelastic single-unit shoppers do not [15][16].

What elasticity modeling does here: picks the threshold (2? 3? 5?) and the discount percent (8%? 15%? 22%?) that maximize expected profit subject to your margin floor. In DataGlass pilot work, modeled-optimal Bundle Deal configurations have recovered single-digit to low-double-digit contribution lift on SKUs where basket structure and price variation are clean enough to identify the threshold effect. We treat the lift as category-specific rather than universal; the gain concentrates where the elasticity curve is steepest just above the threshold.

Use case 2 — Lazada FlexiCombo and voucher-stacking strategy

Lazada's promotional surface is wider — FlexiCombo, Sponsored Discovery vouchers, store-wide vouchers, platform-wide vouchers — and the stacking rules interact in non-obvious ways [6].

What elasticity modeling does here: lets a seller decide where the voucher budget actually buys conversion versus where it just subsidizes loyal customers who would have bought anyway. The right voucher strategy on Lazada is rarely "broadcast a 15% voucher to all followers"; it is usually "send a 22% voucher to the elastic segment of the follower base, and a 5% retention voucher to the inelastic segment." That segmentation is exactly what an elasticity model produces.

Use case 3 — TikTok Shop flash-sale and live-stream pricing

TikTok Shop is structurally different — discovery is creator- and live-stream-driven, conversion windows are minutes rather than days, and the pricing decision is often made in real time during a live session.

What elasticity modeling does here: the principle is the same — match discount depth to demand sensitivity — but the time scale shifts to hours, not weeks. Elasticity models adapted for TikTok Shop emphasize the short-window, high-velocity case: how aggressively to drop price during a live, when to trigger flash-deal stacks, and how comment-driven boosts interact with discount depth [21][22][24].

What changes when you actually know your elasticities

Three things happen, in order, when a seller migrates from gut-feel pricing to elasticity-aware pricing.

Month 1. The discount mix flattens out. The catalog stops getting uniform 15% cuts and starts getting a long-tailed distribution where the top-elastic SKUs see deeper cuts and the inelastic SKUs see almost none. Net margin moves up despite no change in topline revenue.

Month 2. Sale-event configuration changes. 9.9 and 11.11 stop being a "discount everything by the campaign minimum" event and start being a targeted promotional plan. The same advertising spend pulls more orders because the discounts are landing on items that actually convert.

Month 3. The seller stops running unprofitable promotions altogether. The elasticity model identifies which historical promotional decisions destroyed margin and which created it, and the calendar adjusts. This is when the cumulative gain becomes visible. Published dynamic-pricing programs anchor the realistic company-level envelope at 5-10% margin growth, with higher pilot gains possible in promotion-heavy catalogs [9][10].

That is the prize. It is not glamorous. It is not viral. But it is the most reliable upgrade in the marketplace-seller's analytical playbook, and on Shopee, Lazada, and TikTok Shop — where every basis point of margin matters because the platform is taking its share — it is the lever that separates the shops that compound from the shops that survive.

Frequently asked questions

Key takeaways

  • Price elasticity is a per-SKU number that determines whether a discount creates margin or destroys it.
  • The four common seller heuristics — competitor copying, flat discounting, weekly reruns, and gut-feel Bundle Deal configuration — all systematically misallocate discount budget.
  • Elasticity modeling produces three things a dashboard cannot: per-SKU sensitivity, full curve shape, and calibrated uncertainty.
  • On Shopee, the highest-leverage application is Bundle Deal optimization. On Lazada, voucher segmentation. On TikTok Shop, live-stream and flash-sale depth.
  • A realistic public benchmark is 5-10% margin growth from dynamic-pricing programs, with higher pilot gains possible when a catalog has enough promotion history and price variation [9][10].

About DataGlass. DataGlass Labs Research builds decision-intelligence infrastructure for marketplace sellers across Southeast Asia. To explore a pilot or integration, visit DataGlass contact or email teams@dataglasslabs.com.

Cite as: DataGlass Labs Research, "Price elasticity modeling for marketplace sellers — why Shopee, Lazada, and TikTok Shop pricing decisions need it," DataGlass Labs Research blog, 7 May 2026.

Take the next step

Explore a calibrated elasticity pilot.

DataGlass models per-SKU price response from order-line data, promotion history, and margin constraints, then turns it into Bundle Deal, FlexiCombo, voucher, and flash-sale recommendations for marketplace sellers.

Sources & further reading

  1. 01
    McKinsey & Company — Pricing in retail: Setting strategy

    Retail-pricing strategy foundation and the case for analytical price setting over flat rules.

    https://www.mckinsey.com/~/media/McKinsey/Industries/Retail/Our%20Insights/Pricing%20in%20retail%20Setting%20strategy/Pricing_in_retail_setting_strategy.pdf

  2. 02
    McKinsey & Company — Dynamic Pricing in e-Commerce

    Dynamic pricing practice framing for elasticity-aware e-commerce pricing systems.

    https://www.mckinsey.com/capabilities/growth-marketing-and-sales/how-we-help-clients/dynamic-pricing

  3. 03
    Mazumdar, Raj & Sinha — Reference Price Research: Review and Propositions

    Reference-price literature review used for the discount habituation and expected-price sections.

    https://journals.sagepub.com/doi/abs/10.1509/jmkg.2005.69.4.84

  4. 04
    Rajendran & Tellis — Contextual and Temporal Components of Reference Price

    Empirical reference-price model distinguishing temporal and contextual reference points.

    https://papers.ssrn.com/sol3/papers.cfm?abstract_id=907228

  5. 05
    Shopee Help Center — What is a Bundle Deal?

    Shopee Bundle Deal mechanics and example quantity-conditioned discount labels.

    https://help.shopee.sg/4/article/76670-%5BBundle-Deal%5D-What-is-a-Bundle-Deal

  6. 06
    Lazada Seller Center — Flexi Combo help center

    Lazada FlexiCombo seller-promotion documentation.

    https://sellercenter.lazada.sg/seller/helpcenter/flexi-combo-10977.html

  7. 07
    TechNode Global / Momentum Works — TikTok Shop SEA GMV doubled to $45.6B in 2025

    Current cited market-size source for TikTok Shop regional GMV and growth.

    https://technode.global/2026/02/11/tiktoks-southeast-asia-doubles-gmv-year-on-year-to-45-6b-in-2025/

  8. 08
    SellerCraft — TikTok Shop vs Shopee GMV Trends in Southeast Asia

    Marketplace GMV and average-order-value comparison context.

    https://sellercraft.co/tiktok-shop-vs-shopee-gmv-trends-in-southeast-asia-2023-2025-unpacking-the-e-commerce-showdown/

  9. 09
    McKinsey & Company — How retailers can drive profitable growth through dynamic pricing

    Source for the 2-5% sales growth and 5-10% margin increase range reported for mature dynamic-pricing programs.

    https://www.mckinsey.com/industries/retail/our-insights/how-retailers-can-drive-profitable-growth-through-dynamic-pricing

  10. 10
    McKinsey & Company — The dos and don'ts of dynamic pricing in retail

    Promotion and dynamic-pricing implementation caveats.

    https://www.mckinsey.com/capabilities/growth-marketing-and-sales/our-insights/the-dos-and-donts-of-dynamic-pricing-in-retail

  11. 11
    DealStreetAsia / Momentum Works — Southeast Asia e-commerce GMV hit $157.6B in 2025

    Source for 2025 SEA platform e-commerce GMV, 22.8% growth, and the 98.8% top-three platform concentration figure.

    https://www.dealstreetasia.com/stories/momentum-works-report-2025-479044

  12. 12
    Google, Temasek & Bain — e-Conomy SEA 2025

    Regional macro context for Southeast Asia digital commerce and retail-media pressure.

    https://www.bain.com/insights/e-conomy-sea-2025/

  13. 13
    Berry, Levinsohn & Pakes — Automobile Prices in Market Equilibrium

    Canonical instrumental-variables demand-estimation paper used for price-endogeneity framing.

    https://www.its.caltech.edu/~mshum/gradio/papers/BerryLevinsohnPakes1995.pdf

  14. 14
    Kahneman & Tversky — Prospect Theory: An Analysis of Decision under Risk

    Behavioral foundation for reference-dependent evaluation of price gains and losses.

    https://web.mit.edu/curhan/www/docs/Articles/15341_Readings/Behavioral_Decision_Theory/Kahneman_Tversky_1979_Prospect_theory.pdf

  15. 15
    Adams & Yellen — Commodity Bundling and the Burden of Monopoly

    Foundational bundling and second-degree price-discrimination analysis.

    http://neconomides.stern.nyu.edu/networks/Adams_Yellen_Commodity_Bundling.pdf

  16. 16
    McAfee, McMillan & Whinston — Multiproduct Monopoly, Commodity Bundling, and Correlation of Values

    Multiproduct bundling result supporting the need to model joint reservation-value distributions.

    https://academic.oup.com/qje/article-abstract/104/2/371/1854649

  17. 17
    Bhuwalka et al. — Hierarchical Bayesian regression for demand uncertainty

    Illustrative partial-pooling evidence for uncertainty reduction in demand-style estimation.

    https://onlinelibrary.wiley.com/doi/10.1111/jiec.13339

  18. 18
    Elasticity-Based Demand Forecasting and Price Optimization for Online Retail

    Online-retail elasticity forecasting and price-optimization method reference.

    https://arxiv.org/pdf/2106.08274

  19. 19
    Chernozhukov et al. — Double/debiased machine learning for treatment and structural parameters

    Causal ML framework for estimating price effects from observational retail data while reducing confounding bias.

    https://arxiv.org/abs/1608.00060

  20. 20
    BCG — Profiting from Personalization

    Source for the 6-10% revenue-lift range from advanced personalization programs.

    https://www.bcg.com/publications/2017/retail-marketing-sales-profiting-personalization

  21. 21
    Liu et al. — Impulse buying behavior during livestreaming

    Live-commerce scarcity, price perception, and impulse-buying mechanism reference.

    https://pmc.ncbi.nlm.nih.gov/articles/PMC10979273/

  22. 22
    Frontiers in Psychology — Livestreaming e-commerce, flow experience, and time pressure

    Empirical live-commerce study used for the time-pressure and urge-to-buy mechanism.

    https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2022.1019024/full

  23. 23
    Van Heerde, Gijsbrechts & Pauwels — Brand-Pack Size Cannibalization Arising from Temporary Price Promotions

    Source for the finding that roughly 22% of promoted pack-size uplift is borrowed from sibling pack sizes of the same brand.

    https://www.sciencedirect.com/science/article/abs/pii/S002243591200005X

  24. 24
    TikTok Business Help Center — About ads metrics in Seller Center

    Current TikTok Shop Seller Center wording for Shop Ads and Data Compass reporting surfaces.

    https://ads.tiktok.com/help/article/about-ads-metrics-in-seller-center

  25. 25
    Lazada Group — Southeast Asian ecommerce sellers and AI adoption

    Lazada-sourced release describing Business Advisor as an AI-powered analytics tool for seller performance insights.

    https://en.prnasia.com/releases/apac/three-in-four-southeast-asian-ecommerce-sellers-require-additional-support-in-their-ai-adoption-lazada-report-reveals-484620.shtml

  26. 26
    Simon-Kucher — Global Pricing Study 2025

    Pricing-power and AI-pricing adoption benchmark used to contextualize margin-pressure and price-realization claims.

    https://www.simon-kucher.com/sites/default/files/media-document/2025-06/COR_GPS_2025_Brochure_Digital_Final.pdf

More from the archive

  1. January 4, 2026

    Dynamic pricing for marketplace sellers

    Discounting is easy. Profitable pricing is hard. A 30% volume lift on a 10% price cut routinely lowers total contribution profit — the math says volume must lift by ~33% just to break even, and most SKUs underperform that bar. A research note on the price-elasticity arithmetic, the inventory × demand four-quadrant framework, and the per-SKU pricing decision that survives campaign-window pressure.

  2. April 22, 2026

    How to increase profit on Shopee without just selling more

    The standard advice — chase ROAS, scale what works — is structurally biased toward overspend. Why platform ROAS misleads at scale, and the per-SKU break-even bar that replaces it.

  3. March 4, 2026

    How to plan Lazada campaigns around profit

    Pay Day, Mega Sale, 11.11, and the long tail of category windows make participation feel mandatory. The platform documents the eligibility tiers; the seller absorbs them. A research note on what each campaign actually costs, when participation pays, and the margin-first rubric that survives all three.

  4. February 26, 2026

    How to increase profit on TikTok Shop in 2026

    TikTok Shop is the only SEA marketplace with a stacked second commission — affiliate commission (10–25% via the Open Affiliate Plan) layered on top of platform commission. A 6.0 platform ROAS routinely becomes ~1.4 true ROAS once the full four-line cost stack is subtracted. A research note on the affiliate-stack arithmetic, live-stream pricing discipline, and the per-SKU framework that recovers margin without retreating from the platform.

  5. December 12, 2025

    E-commerce Decision Engine: How Marketplace Sellers Turn Data Into Profit Recommendations

    A dashboard tells you what happened. A decision engine tells you what to do next, ranks the options by projected profit lift, and surfaces the math behind every recommendation. A research note on the five-layer architecture that separates the two, why marketplace commerce now requires the latter, and where the operating model breaks.

Stop guessing. Start deploying.

Join the sellers using DataGlass to turn shop data into the next profit-maximizing action.