AI Shopping Assistants: Adding Revenue or Stealing It?

The numbers behind the headlines

Let's start with the statistic that keeps making the rounds: Amazon generates 35% of its revenue from product recommendations . That's roughly $70 billion of their $200 billion annual revenue stream. But here's the crucial insight most people miss: that 35% isn't new money materializing from thin air. It's the extra shirt a customer buys after landing on Amazon to grab a pair of jeans. It's the phone case recommended after checkout. The recommendation engine isn't replacing paid search or organic traffic. It's the multiplier on top of existing channels. Think of it this way: paid ads and organic search bring customers to your site, while recommendation engines increase what they buy once they're there. Remove the paid ads? Traffic dries up. Remove recommendations? Your conversion rate collapses. They're teammates, not rivals.

Why billions "appear" from nowhere

Here's where measurement gets tricky. Traditional last-click attribution has been effectively blind to how AI recommendations actually contribute. Picture this scenario:

A customer sees your product in a Facebook ad
They click through to your site
Your AI recommendation engine suggests a complementary product
The customer buys both items

In traditional analytics, that second product gets 100% credit to "recommendations." But the paid ad initiated the entire journey. As brands implement more sophisticated measurement , they're discovering that recommendations have been contributing value all along - they just weren't measuring it properly. It's not that billions in new revenue suddenly appeared. It's that billions in existing value finally became visible. And that visibility reveals real multiplier effects:

20-30% conversion rate increases on existing traffic
20-40% boosts in average order value
35% reduction in cart abandonment
300% higher email campaign revenue versus generic promotions

The paid search reality: a different challenge

Now, Google's AI Overviews are reducing paid ad clicks , by roughly 34.5% in some cases. This is real disruption, but it's not because recommendations are stealing traffic. It's because users get answers directly from AI without clicking anywhere. A customer searching "best waterproof phone cases under $50" gets an AI-generated summary right on the results page. They don't click. Your paid ad never gets a chance.

How smart marketers are responding

Shifting from informational keywords (where AI provides direct answers) to high-intent, transactional keywords (where users still need to click to buy)
Maintaining 40-50% of PPC budgets in paid search, but with surgical precision
Allocating 5-10% to test new AI platforms

Interestingly, Google's AI-powered bidding shows that while click volume drops, conversion rates improve by 14% . Fewer clicks, but better quality traffic.

The hidden winner: organic search

Here's what anxious marketers miss: organic search is more valuable than ever (yes, really!).

If your content ranks well organically, it becomes the source AI cites, free visibility to millions of users.

ChatGPT Shopping: the genuinely new channel

Shopify integrated 1 million stores into ChatGPT , reaching 700 million weekly users. Brands like Glossier, SKIMS, and Spanx now sell directly in chat conversations, paying only commission-based fees on actual purchases . This is genuine channel expansion, not cannibalization. Early data suggests competitive acquisition costs versus traditional paid search .

The real shift: quality over quantity

The transformation isn't "AI versus paid search." It's the move from volume-based to quality-based acquisition. Users from AI assistants convert at 4X the rate of traditional search traffic. They're pre-qualified, they've researched, compared options, and know what they want before clicking. You're trading high-volume, low-intent clicks for lower-volume, high-intent visitors. The math works:

Fewer clicks from paid ads
BUT 14% better conversion rates
AND 20-40% higher order values

NET: Lower cost per acquisition, higher revenue per visitor

What to do now

Audit paid search keywords – Reallocate from informational to transactional terms
Check organic rankings – Your top 10 positions are your AI Overview foundation
Review product feeds – Data quality is now a competitive advantage
Set up assisted conversions in GA4 – See the full journey, not just last clicks
Create a ChatGPT test plan – Start small with commission-based risk

Final thoughts

The "tens of billions" narrative is true. But AI isn't replacing your channels, it's amplifying smart decisions across all of them. The brands winning right now aren't abandoning proven strategies. They're tightening paid search targeting, investing in organic content, optimizing product data, testing conversational commerce early, and fixing their attribution models. The game is changing from reach to relevance, from clicks to conversions, from last-click to understanding the full journey. The billions are real. But they're being built on top of your existing channels, not replacing them. And that shift? It's an opportunity for brands willing to evolve.

Adrian Gmelch

Adrian Gmelch is Director of Content at Lengow, where he leads content strategy while staying firmly hands-on: reading the research, and tracking the trends that matter before they go mainstream.