Since the advent of AI search models, discussions have arisen about whether major search giants like Google will face significant economic repercussions. The AI economic report released by prominent venture capital firm a16z, authored by partners Justine Moore and Alex Rampell on August 12, reveals that the traffic truly consumed by AI does not come from Google’s revenue-generating search sources, but rather from informational queries that lack advertising value and purely seek knowledge.
The true economic lifeline for Google lies in searches with “purchase intent,” such as “which laptop is a must-buy” or “which air fryer has the highest cost-performance ratio.” Currently, these high-value search traffic sources are still under Google’s control. However, as AI models become increasingly adept at understanding user needs and even proactively recommending products, the search economy is on the brink of a major reshuffle.
AI will first consume knowledge-based searches, not revenue-generating ad queries
The two partners noted that searches can be divided into two types:
1. Informational queries: Questions like “How many protons are in the element cerium?” These types of queries do not attract advertisers and therefore generate no revenue.
2. Commercial queries: Queries like “Which tennis racket is of the best quality?” lead to purchasing behavior and are the most lucrative sources of revenue for Google.
While Google may lose a significant amount of search traffic, as long as it retains these searches with “purchase intent,” it has little to fear from AI taking its traffic.
Five types of shopping behavior, with AI poised to take the middle three the fastest
The two partners categorized human consumption behavior into five levels and analyzed the role AI plays in each:
1. Impulse shopping: This occurs when consumers are influenced by platforms like TikTok or Facebook to immediately purchase trendy items seen on Instagram. This “buy upon seeing” behavior does not require AI assistance for research, but future advertisements will become increasingly personalized, for example, with dynamically customized names and styles, enhancing their appeal.
2. Everyday essentials: Products like pet food and toiletries that consumers buy regularly can be automatically tracked by AI to assist in ordering. In the future, AI may select options with high cost-performance ratios from different channels and automatically replenish stock.
3. Lifestyle shopping: Choosing frequently used cosmetics, sofas, or home decor requires some research and personal preference as there is no standard answer. AI can analyze past consumer records to determine the user’s preferred styles, skin types, and body shapes, curating the “most suitable options for the user.”
4. Functional shopping: Making substantial purchases like bicycles, computers, smartphones, or furniture requires careful consideration. At this point, AI can act not only as a product recommender but also engage in voice dialogue with users as a “product consultant” to understand their needs and help match suitable items.
5. Major life purchases: Buying a house, planning a wedding, choosing a university, or selecting life insurance involves long-term planning processes. In these scenarios, AI acts like a “coach,” helping to research options, analyze pros and cons, simulate choices, and even highlight key points in contracts. Many entrepreneurs are already using AI (like Claude) to flag investment terms.
Food, travel, and medical consumption are too personal to be included in discussion
Consumption habits related to food and travel vary greatly among individuals; some view them as daily necessities while others perceive them as luxury indulgences. Medical consumption patterns are unclear due to uncontrollable health conditions. Although these three areas have potential, the two partners decided to exclude them from this analysis for now.
Amazon and Shopify may face risks, but are more stable than Google
The two partners pointed out that Google’s strength lies in “search intent,” while Amazon and Shopify directly control “consumer actions.” Amazon offers a complete e-commerce service with search, delivery, repurchase records, and product reviews. Shopify helps small and medium-sized businesses list and checkout, gradually connecting with consumers.
They stated that regardless of how much traffic AI takes away from search, as long as consumers ultimately transact on Amazon or Shopify, these companies will remain stable.
Google, with its vast user base and technological advantages, also possesses a wealth of user payment data (via Google Pay). If it were to launch an “AI price comparison assistant” combined with “automatic ordering,” akin to an evolved version of Amazon’s CamelCamelCamel price tracking service, that is certainly a possibility.
For AI to truly become a shopping assistant, it must address four major challenges
They indicated that for AI to transition from “recommendation” to “ordering,” the following conditions need to be met:
1. Clean, reliable data: Current product reviews are rife with fake accounts and spam comments, making AI’s data sources unreliable. If AI can one day organize real user-tested data, it might provide recommendations resembling those of a human.
2. Unified open APIs for e-commerce platforms: AI agents need to access the latest prices, inventory, and even assist in adding items to the cart and checking out while integrating data from various platforms. Currently, the e-commerce ecosystem remains fragmented, limiting AI’s real-time response and automated actions.
3. Users require AI to remember preferences while allowing flexibility in adjustments: AI should remember that you dislike cotton sweaters, prefer high-priced flight tickets but are frugal with everyday items, and be able to adjust flexibly over time. This necessitates a multi-faceted dynamic memory system.
4. Embedded interactions need to be more natural: The ideal AI model would infer consumer preferences from daily behaviors, such as when a user lingers long on a product page or actively asks for feedback after a purchase. Such interactions can further enhance AI’s recommendation capabilities.
AI is set to reshape the e-commerce landscape
Finally, they stated that the current AI search model is entering the market by addressing “non-revenue-generating queries,” but the search traffic for “lifestyle, functional, and everyday shopping” is rapidly being captured by AI. Google, Amazon, and Shopify must now confront the next phase: the “remodeling of the shopping process” warfare.
Whoever can integrate AI assistants, user preferences, platform data, and automated transaction processes will seize the future e-commerce search supremacy.
Risk Warning
Cryptocurrency investment carries a high level of risk, and its prices can be highly volatile, potentially resulting in the loss of your entire principal. Please assess the risks carefully.