Weekly AI TrendsMarch 25, 2026 · 7 min read

AI Recommendation Trends: Which Categories Are Winning SOV This Week?

This week's Avanti GEO monitoring data reveals a significant shift in AI-generated product recommendations across four major models. Smart home devices surged 18.3% in share-of-voice (SOV) week-over-week, while portable power stations dropped 11.7%. We tracked 42,000+ AI-generated shopping responses across ChatGPT, Claude, Gemini, and Perplexity from March 18–24, 2026 to surface the category-level and brand-level trends that matter most to cross-border sellers.

Key Findings
  • Smart home hubs saw the largest SOV increase at 18.3% WoW, with Aqara capturing 27.4% of all AI brand mentions in the category — up from 19.1% last week
  • Personal care electronics (electric toothbrushes, IPL devices) held steady at 14.8% category SOV, but Ulike overtook Braun in 3 out of 4 AI models for the first time
  • ChatGPT and Perplexity diverged sharply on supplement recommendations: ChatGPT favored Nature Made (31.2% SOV) while Perplexity leaned toward NOW Foods (28.6% SOV)
  • Portable power station SOV fell 11.7% WoW as seasonal camping demand tapered — EcoFlow's brand mention rate dropped from 38.1% to 29.4% across all models

Category-Level SOV Breakdown: Where AI Models Are Pointing Shoppers

Across 42,187 monitored AI shopping responses this week, five categories dominated cross-border ecommerce recommendations: smart home devices (22.1% of total SOV), personal care electronics (14.8%), dietary supplements (13.4%), pet supplies (11.2%), and portable power stations (9.6%). The remaining 28.9% was fragmented across 30+ smaller categories. The standout mover was smart home devices. All four AI models increased their recommendation frequency for smart home hubs, sensors, and automated lighting. Gemini showed the strongest bias, dedicating 26.3% of its ecommerce responses to the category — likely influenced by Google's ecosystem integration signals. Claude was more conservative at 18.7%, but still up 4.2 percentage points from the prior week. Pet supplies quietly climbed 3.1% WoW to claim 11.2% of total SOV, driven almost entirely by automatic feeders and pet cameras. This is the fourth consecutive week of gains for this category, suggesting a durable trend rather than a one-week anomaly.

Brand-Level Winners and Losers: Who AI Models Name-Drop Most

At the brand level, the data tells a nuanced story. In smart home, Aqara dominated with 27.4% brand SOV — a 43% relative increase WoW. SwitchBot held at 21.8%, while Philips Hue slipped to 16.3% as AI models increasingly favored Zigbee/Matter-compatible brands with broader ecosystem compatibility. Notably, Aqara's SOV gain was strongest on Claude (32.1%) and Perplexity (29.8%), where product spec depth appears to drive recommendation logic. In personal care electronics, Ulike achieved a milestone: 24.7% brand SOV across all models, overtaking Braun (22.3%) in every model except Gemini. Ulike's dominance in the IPL subcategory (41.2% SOV for at-home IPL queries) is pulling its overall numbers up. For sellers in this space, Ulike's rise signals that Chinese DTC brands with strong English-language content strategies are now outperforming legacy European brands in AI visibility. The biggest loser this week was EcoFlow in portable power stations. Its SOV dropped from 38.1% to 29.4% as Jackery (22.7%) and Bluetti (19.3%) picked up the slack. The decline correlates with reduced seasonal camping and outdoor queries — EcoFlow's high-capacity models are less relevant for the indoor/emergency use cases AI models are now surfacing.

Seller Implications: Actionable Takeaways for March 2026

First, if you're selling smart home accessories or Matter-compatible devices, now is the time to optimize your product content for AI retrieval. Aqara's SOV surge is directly tied to structured spec sheets, comparison-friendly bullet points, and broad third-party review coverage — exactly the signals AI models weight heavily. Sellers should audit their listings for technical specification completeness and ensure compatibility claims are explicitly stated. Second, the ChatGPT vs. Perplexity divergence on supplements is a critical warning. If your brand shows strong SOV on one model but weak on another, you have a content gap. Our data shows Perplexity heavily indexes Reddit threads and niche health forums, while ChatGPT leans on established authority sites like Healthline and WebMD citations. Sellers need platform-specific content strategies — not a one-size-fits-all approach. Third, the pet supplies category at 11.2% SOV with four straight weeks of growth represents a clear entry window. Brand concentration in this category is low: the top brand (Petlibro) holds only 18.9% SOV, meaning there is room for new entrants. Sellers with automatic feeders, smart pet cameras, or subscription-model pet consumables should prioritize GEO optimization immediately before category SOV consolidates around 2-3 dominant brands.

AI Recommendation Snapshot

March 25, 2026 · Avanti Platform Data

Brand / CategoryAI MetricSignalInsight
Aqara (Smart Home)27.4%STRONG BUYLargest WoW brand SOV gain (+43% relative); Matter compatibility driving AI preference
Ulike (Personal Care)24.7%STRONG BUYOvertook Braun in 3/4 AI models; IPL subcategory dominance at 41.2% SOV
Petlibro (Pet Supplies)18.9%WATCHCategory leader in a fast-growing segment, but low moat — new entrants closing in
Nature Made (Supplements)31.2%WATCHStrong on ChatGPT but only 14.3% SOV on Perplexity — platform-specific vulnerability
EcoFlow (Power Stations)29.4%AVOIDSOV dropped 8.7pp WoW as seasonal demand fades; Jackery and Bluetti gaining share
Braun (Personal Care)22.3%AVOIDLost top-2 position to Ulike in 3/4 models; legacy brand authority eroding in AI outputs

Track your brand's AI visibility

Run a free GEO Score audit — see your AI mention rate and share of voice across ChatGPT, Claude, Gemini, and Perplexity.