Social commerce is the use of social interaction and participatory digital media to support commercial activities, especially the discovery, evaluation, purchase, and promotion of goods and services. It combines elements of electronic commerce with communication among consumers, sellers, creators, and communities. Examples include purchasing products presented during a livestream, ordering through a messaging group, joining a group-buying offer, and buying an item tagged in a short video.
The concept is broader than selling through a social-media application. Social relationships, public recommendations, community participation, or creator content must play a meaningful role in the commercial process. Some definitions include transactions initiated on social platforms but completed on an external website; narrower definitions require an integrated purchasing function. This variation affects both research findings and estimates of market size.
Social commerce matters because it changes how demand is generated and how commercial trust is established. Rather than beginning with a deliberate product search, a purchase may emerge from entertainment, conversation, or identification with a community. Its development connects retailing with advertising, the creator economy, platform economics, and questions concerning consumer protection and personal data.
Scope and distinguishing characteristics
Conventional online retail commonly organizes shopping around a searchable catalogue: a consumer identifies a need, compares alternatives, and places an order. Social commerce introduces additional routes into that process. A demonstration may reveal an unfamiliar product, a friend’s recommendation may reduce uncertainty, or a community discussion may establish what counts as a desirable purchase. Search and social discovery coexist, but their relative importance differs across platforms and product categories.
The commercial sequence is therefore often embedded in a broader customer journey. A buyer might encounter a cooking tool in a video, read comments about its durability, ask the presenter a question, and complete checkout without leaving the application. Alternatively, the same content may direct the buyer to a merchant’s website. The first arrangement is often called closed-loop commerce; the second relies on a handoff between services. Neither arrangement alone establishes how influential the social interaction was.
Social commerce overlaps with influencer marketing, but the two are not synonymous. Influencer marketing concerns promotion through people with an audience and may pursue awareness rather than immediate sales. Social commerce can operate without a prominent influencer, as in neighborhood purchasing groups or customer-to-customer recommendations. Conversely, a sponsored celebrity post with no meaningful shopping connection may be better understood as advertising than as a complete social-commerce activity.
A similar distinction applies to an online marketplace. A marketplace matches buyers with sellers and provides transaction infrastructure; it becomes socially organized when interaction, shared content, or community relationships substantially shape exchange. Customer reviews are an important borderline case. Broad definitions treat reviews and ratings as social-commerce features, while narrower definitions reserve the term for more interactive environments, such as social networks, livestreams, and messaging communities. An ordinary product listing does not become strongly social merely because it can be shared.
Social commerce also differs from mobile commerce, which identifies the device or access channel rather than the organization of exchange. Much social commerce occurs on smartphones, but mobile banking and a retailer’s conventional shopping application are not necessarily social commerce. Likewise, omnichannel retailing concerns coordination across stores, websites, applications, and other channels. Social commerce can form one part of that system, connecting a digital recommendation to an in-store purchase, collection point, or later service interaction.
Historical development and regional trajectories
Social influence over purchasing predates digital networks. Personal recommendations, direct selling, buying clubs, and televised shopping all provided earlier ways of connecting interpersonal influence or demonstrations with transactions. The distinctive contribution of the internet was to make recommendations persistent, searchable, measurable, and distributable beyond immediate acquaintances. Early discussion forums, auction feedback systems, and product-review sites established several of the mechanisms later associated with social commerce.
The term became prominent in the mid-2000s as commercial websites adopted participatory features associated with Web 2.0. Yahoo’s Shoposphere initiative and Pick Lists, introduced in 2005, are frequently associated with its early business use. These tools allowed users to assemble and share product selections rather than encounter only merchant-generated catalogues. Academic usage subsequently expanded to include ratings, recommendations, online communities, and social-network-based exchange. Consequently, accounts of the field’s origins vary according to whether they emphasize the terminology, the underlying practices, or integrated checkout.
The spread of smartphones during the 2010s brought cameras, notifications, messaging, and payments into the same device. This reduced the distance between encountering content and acting on it. WeChat, launched in 2011, developed into a communication environment supporting commercial services; its mini programs, introduced in 2017, allowed businesses to provide application-like functions within the platform. Such arrangements made it possible for conversations, membership programs, product information, and purchasing functions to remain closely connected.
China became an important setting for large-scale experimentation with several distinct models. Pinduoduo, founded in 2015, made team purchasing and socially circulated offers central to its growth model. Taobao Live, launched in 2016, connected real-time broadcasting with an established retail marketplace. Douyin and Xiaohongshu developed different combinations of video, recommendations, community content, and shopping. These systems should not be treated as interchangeable: one may emphasize transaction coordination, another entertainment-led discovery, and another product evaluation and lifestyle discussion.
Elsewhere, social commerce developed through a mixture of marketplace listings, messaging-based selling, creator promotions, and platform checkout experiments. TikTok Shop launched in the United States in September 2023, illustrating the extension of video-led commerce into another major retail market. The COVID-19 pandemic accelerated many forms of online selling and remote customer interaction, but adoption remained uneven. Payment access, delivery reliability, retail institutions, platform availability, and consumer habits all influenced local outcomes. The history is therefore not a single progression toward one universal model; it consists of overlapping adaptations to different commercial and technological environments.
Commercial formats and participation
One common format is the shoppable feed: photographs, short videos, or other posts contain product tags or links that connect directly to an offer. The content may be produced by a brand, a retailer, an independent creator, or a customer. The commercial mechanism depends on placing the product within a use context—for example, clothing shown as part of an outfit rather than as an isolated catalogue image. Product information and checkout then convert interest into a possible transaction.
Live commerce combines real-time video with purchasing functions. A presenter can demonstrate a product, answer questions, compare variants, and react to audience comments while viewers place orders. This resembles television shopping but adds interactive feedback and individual account-based tracking. The format is particularly suited to products whose characteristics can be demonstrated visually, such as cosmetics, clothing, kitchen equipment, and food. It is less capable of resolving attributes that require prolonged use, independent inspection, or specialist testing.
Group buying coordinates multiple customers around a shared offer. In a simple arrangement, a lower price becomes available when enough people join before a deadline. In more operationally intensive arrangements, an organizer aggregates orders and coordinates delivery or collection. Potential savings may arise from demand aggregation, reduced selling costs, or consolidated logistics, but advertised discounts can also be financed through marketing subsidies. A group-buying mechanism does not by itself demonstrate lower production costs or a sustainable commercial model.
Messaging-based commerce places transactions inside individual conversations or groups. Sellers may distribute catalogues, answer questions, confirm availability, and arrange payment through services such as WhatsApp or WeChat. This format can support small businesses that lack a sophisticated storefront, but it may leave order records, return procedures, and transaction protections fragmented across messages and payment systems. Community selling also includes neighborhood groups, hobby communities, and secondhand exchanges in which participants’ continuing relationships influence behavior.
The division of work differs across these formats. Merchants generally control sourcing, pricing, and fulfillment; creators supply content and audience access; platforms provide discovery and transaction infrastructure; and customers contribute attention, reviews, referrals, and sometimes promotional content. Affiliate marketing connects several of these roles through commissions on attributable sales. A multi-channel network or specialist agency may coordinate creators, negotiate commercial arrangements, and manage production. These intermediaries can reduce coordination costs, but their presence also makes it important to identify who is responsible for product claims, customer service, and contractual obligations.
Trust, social influence, and recommendation systems
Social commerce addresses a central problem of remote exchange: buyers cannot fully inspect products or directly observe sellers’ reliability before purchasing. This creates information asymmetry. Reviews, demonstrations, purchase histories, and community reputation can provide additional evidence, reducing uncertainty about quality, suitability, and delivery. Their value depends on whether the evidence is authentic, relevant, and representative. A convincing demonstration may establish that a blender crushes ice without establishing how long its motor will last.
Social proof occurs when people use others’ behavior or judgments as evidence about an appropriate choice. Ratings, visible purchases, comments, and popularity indicators can therefore affect demand independently of detailed product information. The inference may be reasonable when other users possess relevant experience, but it can also reproduce errors. A highly visible product may receive more attention because it already appears popular, allowing early advantages to reinforce themselves even when product quality is uncertain.
Creator influence involves additional mechanisms. Audiences may perceive expertise, similarity, or consistency in a presenter’s recommendations. A parasocial relationship—a one-sided sense of familiarity with a media figure—can make a recommendation feel more personal than a conventional advertisement. These relationships are not necessarily deceptive, but commercial incentives can complicate them. Payment, free products, commissions, and long-term brand contracts may influence what is presented and how alternatives are described.
Recommendation systems determine which content receives visibility. Using machine learning, platforms can estimate the probability of viewing, engagement, clicking, or purchasing from signals such as previous activity and content characteristics. The commercial result depends on the objective being optimized. A system trained to maximize viewing time may favor entertaining demonstrations, while one optimized for completed orders may distribute different content. New users and products present a cold-start problem because little behavioral evidence is available.
These mechanisms interact with behavioral economics. Countdown offers, limited-stock messages, and rapidly accumulating purchase indicators can affect decisions under bounded rationality, especially when attention is divided between entertainment and evaluation. Personalization may improve product matching while also making comparisons less visible. Research must distinguish genuine persuasion from selection: people who watch a creator may already prefer the products being promoted. Observed sales following exposure do not, by themselves, establish that the exposure caused those sales.
Platform economics and business models
Social-commerce platforms coordinate several interdependent groups: users, merchants, creators, advertisers, and service providers. Their value often depends on network effects. More relevant creators can attract audiences; larger audiences can attract merchants; and a wider range of products can make the platform more useful. These effects are not unlimited. Excessive advertising, repetitive promotions, poor-quality sellers, or intrusive commercial content can reduce user participation and weaken the same system.
Revenue may come from transaction commissions, advertising, payment processing, subscriptions, merchant software, or logistics services. A platform may initially subsidize one group to attract another—for example, offering low selling fees while charging for promotional visibility. Integrated advertising and commerce allow a merchant to pay both to reach customers and to complete their orders. The resulting business model differs from a simple retailer margin because the platform may never take ownership of the goods.
Transaction costs help explain both the appeal and the limits of integration. Embedded product information and checkout can reduce the effort required to move between discovery and purchase. Standardized payments, order records, and dispute processes can make unfamiliar sellers easier to use. However, reduced purchasing friction does not necessarily mean lower total costs. Commissions, advertising expenditure, creator payments, returns, and compliance work may offset savings in search or administration.
Creator and merchant incentives can also diverge. A commission based on initial order value may reward aggressive selling even when products are frequently returned. A merchant may prefer repeat customers and lower service costs, while a creator may be compensated mainly for immediate conversions. This is a principal–agent problem: the party making promotional decisions does not necessarily bear all their consequences. Contracts can condition compensation on settled transactions, verified claims, or return-adjusted sales, although these arrangements introduce monitoring and accounting complexity.
Platform concentration raises questions about market power and dependence. Merchants may become reliant on access to a particular audience, ranking system, or advertising interface. Accumulated followers and transaction histories can create switching costs, especially when they cannot be transferred elsewhere. At the same time, sellers and creators may use several platforms, reducing exclusive dependence. Competitive outcomes depend on interoperability, differentiated audiences, fulfillment capacity, and the availability of alternative sales channels. A large user base alone is insufficient to determine whether a platform can profitably impose higher fees or less favorable terms.
Transaction infrastructure, fulfillment, and unit economics
A social-commerce transaction requires systems that are largely invisible within the content interface. Product catalogues must connect descriptions, images, prices, variants, and availability to the correct offer. A stock-keeping unit helps distinguish items such as different sizes or colors, while standardized identifiers can support matching across merchants and logistics providers. Incorrect mappings can produce misleading recommendations, price discrepancies, or orders for unavailable variants.
Checkout connects the social interface to a payment system, address collection, fraud screening, and order confirmation. The platform, merchant, and payment provider may have different contractual responsibilities. A stored payment credential can make checkout faster, but it also requires controls against unauthorized use. Embedded checkout should therefore be distinguished from ownership of the entire transaction: the platform displaying the purchase button may not be the seller, payment institution, warehouse operator, or final dispute-resolution authority.
Demand generated by viral content can be difficult to forecast. A short promotion may produce a concentrated order spike rather than a stable flow, placing pressure on inventory, packaging capacity, and delivery networks. Understocking causes cancellations and delays; overstocking ties up working capital and creates markdown risk. Merchants must coordinate promotional plans with their supply chain, particularly when products have long replenishment times, seasonal demand, or limited shelf lives.
Sales volume is not the same as economic viability. A simplified order-level contribution margin can be written as
where is net merchandise revenue, product cost, fulfillment and payment costs, platform and creator commissions, attributable acquisition spending, and expected return, refund, and service costs. The categories must be defined consistently to prevent double counting. Taxes collected on behalf of government and cancelled orders should not be mistaken for revenue available to cover the merchant’s costs.
Returns are particularly important when purchasing is driven by demonstrations or rapid decisions. Clothing may not fit, product appearance may differ from the presentation, and buyers may reconsider an impulse purchase. The operational consequences include reverse shipping, inspection, repackaging, damaged stock, and customer-service labor. Cross-border transactions add customs procedures, currency conversion, delivery uncertainty, and potentially conflicting rules. Social discovery can be nearly instantaneous, while physical delivery and after-sales obligations remain constrained by geography, capacity, and product characteristics.
Measurement, attribution, and research methods
The size of social commerce is difficult to measure because definitions differ. A narrow estimate may count only purchases completed within social applications. A broader estimate may include any online sale influenced by social content, even if checkout occurs elsewhere. The latter category is substantially harder to observe and overlaps with other marketing channels. Comparisons across countries or years are unreliable unless they use compatible boundaries, populations, and treatment of cancellations and returns.
Gross merchandise value measures the value of transactions conducted through a service under a specified definition; it is not the same as platform revenue or seller profit. A marketplace may process a large value of goods while retaining only commissions and service charges. Reported totals may also differ in their handling of shipping, discounts, taxes, unpaid orders, or refunds. Because these conventions vary, a transaction-value figure requires a clear explanation of what it includes.
Operational indicators include impressions, viewing duration, engagement, product-page visits, conversion rates, average order value, repeat purchase, and return rates. Each answers a different question. A conversion rate calculated per viewer is not directly comparable with one calculated per product click or checkout visit. Engagement can reflect amusement, criticism, or curiosity rather than buying intent. For merchants, return-adjusted margin and repeat purchasing may be more informative than initial sales generated during a highly subsidized campaign.
Marketing attribution assigns credit for a purchase to preceding contacts. Last-click attribution gives credit to the final measured interaction, potentially overlooking earlier creator exposure. View-through attribution counts certain purchases after an advertisement was displayed, but may attribute orders that would have happened anyway. Cross-device activity, private messaging, tracking restrictions, and offline purchases make the path harder to reconstruct. Attribution rules organize observed activity; they do not automatically identify incremental effects.
Research on causation therefore relies on stronger designs. A randomized controlled trial can assign eligible users to different exposure conditions, although sharing between users may contaminate the comparison. Geographic or time-based experiments can estimate broader effects but require attention to seasonal demand and other changes. Observational studies may use statistical controls or natural experiments, yet remain sensitive to unmeasured differences between audiences. Recommendation feedback loops create an additional complication: a successful promotion changes later visibility, so immediate and longer-term effects may differ. Studies must specify whether they estimate responses among exposed viewers, changes in total merchant sales, or shifts in purchasing between competing sellers.
Consumer protection, governance, and unresolved questions
Social commerce places commercial claims inside environments that also contain personal expression, entertainment, and conversation. This makes the identification of advertising important. A viewer may interpret an enthusiastic recommendation differently when it is accompanied by a commission, sponsorship payment, or free product. In the United States, the Federal Trade Commission’s Endorsement Guides, revised in 2023, address endorsements and the disclosure of material connections. Their application depends on the circumstances of the communication rather than on whether a platform labels itself a shopping service.
Reviews and other reputation signals require separate safeguards. Fabricated testimonials, purchased positive reviews, undisclosed insider reviews, and manipulated engagement can undermine the evidence consumers use to assess products. The FTC’s rule on consumer reviews and testimonials took effect in October 2024 and addresses specified deceptive practices, including fake reviews and certain forms of review suppression. Detecting manipulation remains difficult because suspicious patterns can resemble genuine bursts of popularity, particularly during livestreams or coordinated promotions.
Product safety, counterfeit goods, and seller identification create responsibilities beyond advertising disclosure. Consumers need to know which entity sold an item, which claims formed part of the offer, and how complaints or recalls can be handled. The European Union’s Digital Services Act, generally applicable from February 2024, establishes obligations that vary with service type and size, including specific requirements for online marketplaces. It operates alongside consumer, product-safety, and other legislation rather than replacing them.
Personalization also raises data privacy and data governance questions. Social-commerce systems may combine browsing, viewing, purchasing, communication, and inferred interests. The combination can reveal more than any individual data source, especially when used to infer sensitive characteristics. Applicable duties depend on jurisdiction, the data involved, the purposes of processing, and the roles of the organizations concerned. Children’s participation adds distinct concerns involving advertising recognition, profiling, and age-appropriate design.
Several analytical questions remain open. Researchers continue to examine whether community mechanisms broaden market access for small sellers or deepen dependence on dominant intermediaries; whether live interaction improves product understanding or primarily accelerates decisions; and how promotional work should be allocated between paid creators and ordinary users. Generative artificial intelligence adds questions about synthetic presenters, automated recommendations, and the authenticity of endorsements. The relevant distinctions concern who generated a claim, what evidence supports it, whether commercial interests are disclosed, and which organization can be held responsible when a transaction fails.