How developers can build AI shopping catalogs with Bing APIs

SerpApi published a comprehensive tutorial on December 9, 2025, demonstrating how developers can construct AI-powered shopping catalogs by combining Bing Copilot API with Bing Shopping API capabilities. The technical guide, authored by SerpApi engineer Sonika Arora, addresses holiday season e-commerce challenges through automated gift discovery and price comparison systems that transform generative AI suggestions into actionable purchase opportunities.

The tutorial targets e-commerce builders confronting seasonal competition spikes and rising advertising costs during peak shopping periods. According to the published documentation, the implementation uses SerpApi's Bing Copilot API to generate high-quality gift ideas based on customer personas, then employs the Bing Shopping API to locate optimal pricing and purchase links for suggested items. This two-stage process automates what traditionally requires manual research across multiple platforms.

The technical architecture streamlines web scraping complexities that typically burden custom implementations. "SerpApi manages the intricacies of scraping and returns structured JSON results," according to the tutorial. The platform handles proxy management and CAPTCHA resolution automatically, eliminating infrastructure concerns that developers would otherwise need to address through tools like BeautifulSoup and Requests. The service maintains parser compatibility as Bing experiments with layout modifications and interface changes, reducing ongoing maintenance requirements for development teams.

Implementation requires minimal setup. Developers install the google-search-results Python library—a naming convention that extends beyond Google APIs to support all SerpApi services. The platform provides 250 free monthly search credits for initial exploration and testing. Authentication occurs through API keys retrieved from user accounts, with optional environment variable configuration recommended for security purposes.

The Bing Copilot integration represents a departure from traditional search engine behavior. Standard search engines return ranked lists of articles requiring manual review. Bing Copilot generates direct answers with structured recommendations when prompted appropriately. The tutorial demonstrates this capability through a sample query: "Christmas gift ideas for a tech savvy teenager in 2025 who likes smart home gadgets and gaming. Provide 5 shopping item ideas in bullet points numbered 1-5."

The function structure follows straightforward patterns. The tutorial provides code implementing generate_ideas_with_bing_copilot, which accepts query strings and returns JSON-formatted results containing text blocks with list-type responses. The copilot engine processes natural language requests and produces numbered recommendations including items like "Nintendo Switch 2" alongside descriptive context about each suggestion.

Product discovery transitions from conceptual recommendations to concrete purchasing options through the Bing Shopping API integration. The tutorial's get_shopping_results_with_bing_shopping function accepts product names extracted from Copilot responses and returns structured shopping data including titles, external purchase links, prices, and seller information. This separation of concerns enables developers to maintain distinct API calls for ideation and commerce fulfillment.

The complete catalog generation system combines both API interactions into unified workflows. The tutorial's implementation defines customer personas, calls the Copilot API for gift suggestions, creates CSV files with appropriate headers, extracts gift ideas through regular expression parsing that handles Bing's formatting patterns with dashes and colons, searches shopping results for each extracted idea, and writes consolidated product information to exportable files.

Regular expression processing addresses Bing Copilot's tendency to format recommendations with punctuation separators. The code uses re.split(r'[–:]', snippet, maxsplit=1)[0].strip() to isolate core product names from descriptive suffixes, ensuring clean query strings for subsequent shopping searches. This parsing step proves critical for maintaining search quality when transitioning between AI-generated content and structured e-commerce data.

The resulting CSV output provides complete buying catalogs with direct purchase links. According to the tutorial documentation, the final file contains columns for product titles, external URLs, pricing information, and seller identities. This structured format enables immediate integration with front-end shopping interfaces or backend inventory systems without additional data transformation.

The tutorial arrives as AI shopping adoption accelerates across consumer segments. Research released November 25, 2025, showed 85 percent of UK consumers planning to use AI for Christmas shopping would trust autonomous agents to place orders and execute payments. McKinsey projections estimate agentic commerce could orchestrate between $900 billion and $1 trillion in U.S. B2C retail revenue by 2030, with global figures reaching $3 trillion to $5 trillion.

However, skepticism persists regarding commercial viability despite technological capabilities. Independent analyst Andrew Lipsman published analysis on October 6, 2025, identifying eight structural challenges facing autonomous shopping systems. Amazon and Shopify collectively control more than 50 percent of U.S. e-commerce and both platforms block AI agents to protect retail media businesses and maintain discovery ownership.

Major platforms have deployed competing approaches to AI-powered shopping throughout 2025. Google launched agentic checkout capabilities on November 13, 2025, enabling automated purchases through price tracking and Google Pay integration. OpenAI introduced instant checkout features on September 29, 2025, through the Agentic Commerce Protocol developed with Stripe, enabling transactions from Etsy and Shopify merchants directly through ChatGPT conversations.

The SerpApi implementation differs from autonomous checkout systems by focusing on discovery and aggregation rather than transaction execution. The tutorial's architecture generates curated product recommendations with purchase links, leaving final purchasing decisions and payment processing to consumers and existing e-commerce platforms. This approach aligns with merchant preferences for maintaining customer relationship control while leveraging AI capabilities for product discovery.

The technical foundation relies on Bing's search infrastructure, which Microsoft has consolidated substantially throughout 2025. Microsoft retired traditional Bing Search and Custom Search APIs on August 11, 2025, directing developers toward Azure-integrated alternatives or specialized services like SerpApi. The retirement affected both free tier users and enterprise customers, eliminating new deployments while maintaining existing resources until the shutdown date.

Microsoft's search advertising business generated $13.9 billion during fiscal 2025, representing 21 percent year-over-year growth according to earnings reported July 30, 2025. The revenue stream accelerated throughout the fiscal year as AI capabilities integrated across Bing search and Edge browser platforms. Chief Financial Officer Amy Hood attributed growth to "volume and revenue per search across Edge and Bing" during the earnings call.

The SerpApi tutorial addresses several e-commerce pain points that intensify during holiday periods. Competition spikes force advertising cost increases across all platforms. Consumers exhibit bifurcated behavior patterns—either knowing exactly what they want but hunting for optimal deals, or lacking direction entirely and scrolling endlessly seeking inspiration. Traditional search engines require multiple queries and manual comparison across retailers to resolve either scenario.

The automated catalog generation reduces research friction by combining AI-powered recommendation engines with comprehensive price comparison in single workflows. A tech-savvy teenager persona generates suggestions spanning smart home gadgets and gaming products. Each suggestion immediately connects to available inventory with pricing transparency across multiple sellers. This compression of research time potentially increases conversion rates by reducing decision fatigue and abandonment.

The Python implementation demonstrates accessibility for developers with moderate technical expertise. The code samples avoid complex architectural patterns, focusing on sequential API calls and data transformation using standard libraries. CSV output enables immediate integration with existing systems without requiring specialized database configurations or API endpoint development.

SerpApi's infrastructure handles rate limiting and request management that would otherwise require careful implementation. The platform maintains parser compatibility as Bing modifies interface elements and search result structures. This abstraction layer enables developers to focus on business logic and user experience rather than infrastructure maintenance and scraper reliability.

The tutorial includes complete code repositories on GitHub, enabling developers to examine full implementations beyond tutorial excerpts. Sample playground links demonstrate API responses in interactive environments, allowing experimentation with query variations and result structures before committing to implementation. Documentation references provide detailed parameter specifications for both Copilot and Shopping APIs.

The timing coincides with broader industry movement toward API-driven commerce integrations. Apple introduced Advanced Commerce API on January 23, 2025, targeting applications handling extensive content catalogs and complex subscription models. Google launched Merchant API in August 2025, establishing it as the primary tool for programmatic Merchant Center access while setting August 18, 2026, as the shutdown date for Content API for Shopping.

The SerpApi approach enables developers to bypass direct API relationships with Microsoft, which has imposed increasingly complex requirements on platform participants. Microsoft Advertising now requires all third-party publishers to install Microsoft Clarity across placements, with non-compliant traffic filtered as nonbillable. These platform-specific requirements create operational complexity that intermediary services like SerpApi can absorb on behalf of developers.

The catalog generation pattern extends beyond holiday shopping applications. E-commerce platforms could implement similar architectures for personalized product recommendations based on user preferences, competitive intelligence systems comparing inventory and pricing across marketplaces, inventory planning tools identifying trending products through AI analysis, and customer service systems generating product suggestions based on natural language inquiries.

Performance characteristics depend on API response times and query complexity. The tutorial does not specify latency benchmarks, though structured JSON responses typically enable sub-second processing for individual queries. Catalog generation for multiple products requires sequential API calls, with total processing time scaling linearly with product count. Developers implementing production systems would need to consider rate limiting, caching strategies, and asynchronous processing for larger catalogs.

The CSV output format prioritizes simplicity over sophistication. Production implementations might require additional fields including product images, detailed specifications, availability status, shipping estimates, and seller ratings. The tutorial's architecture provides foundation patterns that developers can extend with additional API calls or data enrichment processes.

Security considerations include API key management and data handling practices. The tutorial recommends environment variable storage for API credentials rather than hardcoded values. CSV files containing product links and pricing represent sensitive business intelligence that competitors could exploit. Production systems should implement appropriate access controls and data retention policies.

The broader context includes substantial investment in commerce-enabled advertising formats throughout 2025. PayPal debuted shoppable storefront ads on June 16, 2025, enabling direct purchases within display advertising units. Digital advertising revenue reached $258.6 billion in 2024, with commerce media including retail media networks growing 23 percent to $53.7 billion. Retail media networks embraced real-time bidding for sponsored products through programmatic solutions enabling activation across multiple networks.

The SerpApi tutorial represents practical implementation of capabilities that industry analysts project will reshape digital commerce. The combination of generative AI for discovery and structured APIs for fulfillment addresses fundamental e-commerce challenges around product selection and price optimization. Whether this architecture becomes standard practice depends on adoption patterns, platform policy evolution, and consumer acceptance of AI-mediated shopping experiences.

The code remains accessible on GitHub at sonika-serpapi/build-an-ai-powered-shoppingcatalog, with complete documentation available through SerpApi's official blog. Developers can contact the author at sonika@serpapi.com for technical questions. The platform offers free accounts with 250 monthly search credits for experimentation and prototyping.

Timeline

• February 24, 2023: Botify participates in Bing's content submission API pilot, enabling real-time indexing notifications
• January 23, 2025: Apple introduces Advanced Commerce API for large-scale app transactions
• February 1, 2025: Microsoft integrates Grounding with Bing Search into Azure AI Agent Service
• June 3, 2025: Amazon officially deprecates Posts program after declining performance
• June 9, 2025: Microsoft ends Bing Search APIs, alternative costs 40-483% more
• July 30, 2025: Microsoft search advertising revenue climbs 21% reaching $13.9 billion
• August 6, 2025: E-commerce app installs drop 14% despite 2% session growth
• August 22, 2025: Google launches Merchant API signaling transition from Content API
• September 29, 2025: OpenAI launches instant checkout capabilities through Agentic Commerce Protocol with Stripe
• October 6, 2025: Skepticism grows over AI shopping agents as ChatGPT checkout launches
• October 3, 2025: Microsoft unveils redesigned Bing Places for Business with new domain and import tools
• October 15, 2025: Bing adds data-nosnippet attribute for search and AI control
• November 11, 2025: Microsoft Advertising requires Clarity for third-party publishers
• November 13, 2025: Google launches agentic checkout and AI shopping tools
• November 13, 2025: Facebook Marketplace introduces AI features and collaborative shopping tools
• November 25, 2025: AI shopping adoption accelerates with 85% consumer trust for autonomous ordering
• December 9, 2025: SerpApi publishes tutorial on building AI-powered shopping catalogs using Bing Copilot and Shopping APIs

Summary

Who: SerpApi, a web scraping platform service, published the tutorial through engineer Sonika Arora. The guide targets e-commerce developers, platform builders, and technical teams implementing holiday shopping features or product discovery systems.

What: The tutorial demonstrates building AI-powered gift buying catalogs by combining SerpApi's Bing Copilot API for generating trending gift ideas with Bing Shopping API for finding optimal prices and purchase links. The Python implementation creates CSV catalogs containing product titles, external links, pricing, and seller information.

When: SerpApi published the comprehensive tutorial on December 9, 2025, during peak holiday shopping preparation periods when e-commerce competition intensifies and advertising costs climb across platforms.

Where: The tutorial appears on SerpApi's official blog with complete code repositories available on GitHub. The implementation uses Bing's search infrastructure through SerpApi's intermediary service, which handles proxy management and CAPTCHA resolution for developers globally.

Why: The tutorial addresses fundamental e-commerce challenges around product discovery and price comparison that intensify during seasonal shopping periods. Consumers either know exactly what they want but spend excessive time hunting deals, or lack direction entirely and scroll endlessly seeking inspiration. The automated system compresses research time by combining AI-powered recommendations with comprehensive price comparison in unified workflows, potentially increasing conversion rates by reducing decision fatigue and cart abandonment.