Executive summary – what changed and why you should care

OpenAI and Perplexity have each added AI shopping capabilities directly into their chatbots: visual and conversational product search, personalized recommendations, and in‑chat checkout enabled by Shopify (OpenAI) and PayPal (Perplexity). For retailers and platforms, this converts conversational engagement into a direct purchase path; for startups, it raises the stakes in the run‑up to the holidays and beyond.

  • These feature launches turn general chatbots into commerce touchpoints with conversion mechanics built in.
  • Adobe predicts AI‑assisted online shopping could grow ~520% this holiday season, amplifying near‑term impact.
  • Vertical startups claim data quality and merchandising expertise still win for nuanced categories like fashion and home.

Key takeaways

  • Immediate impact: Users can research and buy inside ChatGPT or Perplexity, shortening discovery to checkout and improving funnel conversion potential.
  • Competitive pressure: Big platforms’ existing user bases and retailer deals give them scale advantages that can rapidly capture market share.
  • Specialist advantage: Startups with curated domain datasets (fashion, furniture) retain quality advantages in nuanced categories.
  • Risks: monetization via ads or paid placement could degrade search quality; privacy, PCI compliance and hallucinations create operational and legal exposure.

Breaking down the announcement

Functionally the features align: users can ask conversational queries (e.g., “new gaming laptop under $1,000 with a 15+ inch screen”), provide photos of garments for similar lower‑priced items, get recommendations tailored by memory, and complete transactions inside the chat UI. OpenAI’s Shopify integration enables native cart and checkout flows; Perplexity’s PayPal tie provides a comparable in‑chat payment experience.

Why now

Timing is deliberate: holiday shopping increases transaction volume and advertiser interest, and LLMs plus visual search are mature enough to provide usable recommendations. Platforms want to monetize high‑engagement chat sessions-e‑commerce partnerships offer a direct revenue route and a playbook already used by Google and Amazon.

Where specialist startups still matter

Founders of vertical AI shopping companies argue that domain datasets and merchandising logic produce materially better outcomes for fashion and home goods. “Any model or knowledge graph is only as good as its data sources,” Onton CEO Zach Hudson said, noting Onton cataloged hundreds of thousands of interior items into a cleaner pipeline. Daydream founder Julie Bornstein emphasizes that fashion is “uniquely nuanced and emotional,” requiring silhouette, fabric and outfit‑level reasoning that general search indexes don’t capture.

Risks and governance considerations

Operational and regulatory risks multiply when discovery and checkout converge. Payment integrations raise PCI and fraud‑prevention requirements. If platforms monetize placement, discovery quality can deteriorate and regulators may scrutinize transparency and deceptive practices. LLM hallucinations or misattributed product claims create liability around false advertising and returns. Data privacy issues also grow when memory features personalize recommendations across contexts.

Competitive context

Compare three archetypes: 1) Big platforms (OpenAI, Perplexity) offering scale, attention and retailer integrations; 2) Verticals (Phia, Cherry, Deft, Onton, Daydream) offering domain‑tuned catalogs and UX; 3) Retailer‑owned experiences (Shopify merchants) that can plug into any channel. Early advantage favors platforms because of reach and partnership velocity, but verticals can sustain higher conversion quality in niche categories.

What this means for operators and buyers

Retailers should expect new referral channels and test in‑chat checkout pilots; merchandising and attribution teams must adapt. Startups should double down on proprietary datasets, product ontologies and shopper intent signals where they can sustain higher precision. Legal, fraud, privacy and customer‑care teams need updated playbooks for chat‑native commerce.

Recommendations – who should act and how

  • Retailers: Run A/B tests integrating with platform checkouts (Shopify/PayPal) now to measure conversion lift and CAC before the holiday peak.
  • Vertical startups: Invest in data quality, merchandising logic and attribution to prove higher lifetime value per customer; consider partnerships where your catalog adds unique value inside larger platforms.
  • Product leaders: Update compliance and fraud controls for conversational checkout and validate claims handling to minimize returns from AI recommendations.
  • Execs and investors: Monitor ad/placement monetization—if platforms pivot to paid promotion, expect degraded organic discovery and increased ad spend to maintain visibility.

Bottom line: OpenAI and Perplexity converting chat into checkout is a structural shift that accelerates AI commerce. It expands channels—and rapidly concentrates control over discovery—so startups with narrow, high‑quality data must move from “better recommendations” to demonstrably higher conversion and retention to remain competitive.