The sense of alarm that "Google Search will be replaced by AI" is spreading even among e-commerce operators. But what needs to be verified is what's actually inside that growth-rate figure. On the earnings call held on August 5, 2026, Shopify President Harley Finkelstein revealed that AI-driven traffic and orders had increased 3x year over year. However, compared to past announcements, this growth rate has rapidly shrunk from 15x down to 3x, and traditional search remains alive and well, up 1.3x. Layering these two numbers together reveals the outline of what e-commerce operators should start addressing right now.
What the earnings call revealed about the "3x via AI" breakdown
Shopify announced its Q2 earnings for the period ended June 30, 2026 on August 5, 2026, and held an earnings call that day starting at 8:30 AM Eastern Time. Revenue rose 34% year over year (33% on a constant-currency basis) to approximately $3.58 billion, exceeding the market consensus of roughly $3.45 billion. Gross Merchandise Volume (GMV) also reached approximately $115.6 billion, up 32%. Following the earnings release, Shopify shares surged sharply during that day's trading. The stock closed up 16.98% from the previous day at $144.24, with intraday gains briefly reaching as high as 24.7%.
What drew the most attention on this earnings call was President Harley Finkelstein's account of AI search performance. He positioned AI search not as a replacement for search but as a complement to it, revealing that both AI-driven traffic and orders had increased 3x year over year. Alongside the figures in the earnings materials, a breakdown showing actual purchasing behavior was also disclosed. Half of AI-driven sessions land directly on product detail pages—a rate 2.5 times higher than that of traditional search. Meanwhile, 75% of AI-driven purchases occur outside the top 100 best-selling categories.
Finkelstein offered an explanation for the mechanism behind this difference during the earnings call. AI agents send multiple queries against the inventory catalog, using structured data as clues to find products matching buyer intent. This is fundamentally different from the starting point of exploration used by search engines, which display results by popularity based on a small number of keywords. This difference in search methodology explains both why many sessions bypass category pages to land directly on individual product pages, and why the reach into long-tail products is so high.
Shopify's official blog for Q1 2026 also presented data backing up the high quality of AI-driven traffic. Limited to AI-driven sessions originating from product detail pages, the conversion rate is about 50% higher than organic search. Looking at AI-driven orders overall, the average order value (AOV) is 14% higher. This metric was not updated again in the Q2 earnings materials. But tracking the trajectory of the growth rate itself reveals a pattern that can't simply be dismissed with the phrase "rapid expansion continues."
The truth behind the numbers 7x, 8x, 15x, and now 3x
This isn't the first time Shopify has discussed AI-driven growth rates in its own announcements. In its November 2025 announcement, the company stated that, using January 2025 as a baseline, AI-driven traffic had grown 7x and orders 11x. In February 2026, a Shopify Japan announcement was cited stating that AI-driven orders had grown 15x over the one-year period from January 2025 to January 2026. In the official blog accompanying that year's Q1 earnings, figures showed sessions up 8x and orders up nearly 13x year over year. And now, in this Q2 earnings report, the year-over-year growth rate has shrunk to 3x.
Looking at this sequence of numbers alone might suggest a rapid slowdown, but the four multipliers each use different comparison periods. The 7x figure from November 2025 represents a 10-month change measured against the low baseline of January 2025, and the 15x figure from January 2026 is merely a single-month year-over-year comparison using the same January 2025 starting point. On the other hand, the Q1 figure of 8x is a year-over-year comparison for the January–March 2026 period, and the Q2 figure of 3x is a year-over-year comparison for the April–June 2026 period—meaning the prior-year comparison periods themselves (January–March 2025 and April–June 2025) had likely already factored in substantial AI-driven traffic growth. The larger the denominator (the prior-year baseline) grows, the smaller the resulting multiplier becomes, even for an increase of the same magnitude.
Simplifying the mechanics makes this clearer. Suppose the prior-year index is 10 and this period's figure is 30—the growth rate comes out to 3x. But if the prior-year index has already expanded to 20, then even if this period grows to 60, the calculated multiplier remains the same 3x. No matter how much the numerator (this period's figure) has grown, if the denominator (the prior-year figure) has already ballooned, the surface-level multiplier will appear to shrink. A decline from 7x to 3x doesn't necessarily mean growth has stopped—it may also be the result of a shifting comparison baseline.
The sharp jump from 7x (November 2025) to 15x (January 2026) coincides with the year-end shopping season. A Shopify Japan survey found that 51% of Japanese consumers planned to use AI for product discovery during Black Friday and year-end shopping events, supporting the hypothesis that seasonal factors may have temporarily boosted the AI-driven growth rate. That said, confirming this hypothesis with hard numbers would require the absolute volume of AI-driven traffic itself. A relative figure like a multiplier alone cannot separate out how much of the change is due to denominator shifts versus actual growth.
Traditional search is also up 1.3x—Google isn't dead
While the rapid rise of AI-driven traffic tends to dominate the conversation, another figure Finkelstein mentioned on the same earnings call shouldn't be overlooked. Traditional search sessions have increased 1.3x over the past two years, and their share of total sessions has remained steady at roughly one-third. While the narrative that "Google Search will be swallowed by AI" spreads across the tech industry, Shopify's own data shows no sign of the search channel shrinking. The growth of AI-driven traffic and the growth of traditional search are phenomena occurring simultaneously over the same period.
An analysis published by Adobe's analytics service, Adobe Analytics, on June 17, 2026, corroborates the rapid expansion of AI-driven traffic as independent data. As of May 2026, AI-driven traffic to U.S. retail sites was up 138% year over year, and the cumulative increase since October 2024 reached 1,324%. The average conversion rate for AI-driven traffic is said to be 54% higher than non-AI traffic. While AI-driven traffic is expanding rapidly, there's no data anywhere indicating a decline in traffic from traditional search. For e-commerce operators, this is not a number that justifies cutting investment in search engine optimization (SEO).
Figures from Amazon's AI shopping assistant "Rufus" point in the same direction. In 2025, more than 300 million customers used Rufus, and Amazon states it contributed an estimated $12 billion in incremental annualized sales. This is one more example showing that AI-driven purchasing behavior is expanding in parallel across the e-commerce industry as a whole.
The fact that both channels are growing simultaneously has direct implications for how e-commerce operators allocate their budgets. Treating AI-optimization efforts and Google search optimization as competing for the same budget risks missing growth opportunities in both. Shopify's data can be read as showing a structure where new channels are being added on top of existing ones, not replacing them.
The "absolute share" Shopify never discloses
Even after lining up all these numbers, one gap remains unfilled. While Shopify repeatedly announces the relative growth rate of 3x, the absolute proportion that AI-driven sessions or sales represent of the total is nowhere to be found—at least not in any of the earnings materials, official blog posts, or earnings call statements reviewed here.
Finkelstein explicitly stated on the same earnings call that traditional search sessions continue to account for roughly one-third of the total. The fact that Shopify can calculate search's share of total sessions means the company already has the denominator broken down by traffic source. The same calculation should technically be possible for AI-driven traffic as well, yet that particular figure is conspicuously absent from published materials. It's hard to attribute this to any technical difficulty in aggregation.
Independent data sources like Adobe Analytics also show growth rates and conversion-rate differentials, but do not disclose share figures either. Determining whether the decline from 7x to 3x represents genuine deceleration or merely a shifted comparison baseline requires the absolute figure of what percentage of total traffic is AI-driven. What readers are seeing is only the relative growth rate that Shopify has chosen to disclose. The interpretation of the multipliers discussed in the previous section remains, in this sense, suspended in mid-air.
The 16.98% closing-price boost is itself partly a result of investors relying solely on the relative figure of "3x year over year" as their guide. Whether AI-driven sales account for just a few percent of the total or have reached tens of percent, the published expression "3x" alone offers no way to distinguish between the two. As long as the absolute figure remains undisclosed, this pattern of the market reacting to relative growth rates alone will keep repeating.
For now, investors and e-commerce operators have limited means of filling this gap. Aside from tracking quarterly earnings materials and indirectly inferring the scale from the trajectory of the multipliers, no highly reliable basis for judgment has been made public. Whether Shopify will eventually disclose its absolute share depends on whether the company judges that the figure would work in its favor relative to competitors.
The difference from OpenAI's retreat from "AI checkout"
When discussing AI-driven sales, treating referral traffic and checkout completion as the same phenomenon leads to a misreading of what's actually happening. On September 29, 2025, OpenAI launched "Instant Checkout," which lets users complete purchases entirely within ChatGPT. Etsy sellers were on board from launch, and major retailers joined too—Walmart enabled roughly 200,000 items for the feature in November 2025. Yet according to Finkelstein, out of the millions of Shopify merchants, only a dozen or so actually went live with it.
In March 2026, Daniel Danker, EVP of Product & Design at Walmart, revealed that the conversion rate for checkout completed within ChatGPT was only about one-third that of directing users to walmart.com (a roughly 66% drop). The promise of completing checkout entirely within a chat interface drew enormous attention, but by March 4, 2026, OpenAI announced it would scale back Instant Checkout in stages and shift its focus toward a checkout model managed by merchants themselves. There's no official statement directly linking the two events, but the timing coincides with when Walmart's underwhelming results were reported.
The data Shopify presented in its earnings supports this referral-based model. The figure showing that half of AI-driven sessions land directly on product detail pages demonstrates that AI is actually directing users to the product pages and checkout flows operated by Shopify's own merchant stores. Perplexity has also rolled out its own purchase feature, "Buy with Pro," but what Shopify's data proves this time is strictly the effectiveness of referral traffic—not the effectiveness of completing checkout entirely within an AI platform. OpenAI's retreat is one example showing that this latter model has yet to be proven out.
This distinction serves as practical guidance for e-commerce operators. To accommodate AI-driven referral traffic, it's enough to smooth out the pathway that guides users to your own existing checkout—a major system overhaul isn't required. On the other hand, supporting in-platform checkout completion within an AI service requires handing over inventory integration and payment-processing authority to the platform, which changes both the scale of investment required and the risks that need to be assessed. What Shopify's data supports is the former approach; and even looking at Walmart's results with the latter (AI-embedded checkout), the conversion rate came in at only one-third of what's achieved by directing users to its own site.
What Japanese e-commerce operators should start doing right now
Shopify's Q2 revenue of approximately $3.58 billion translates to roughly ¥562 billion at the exchange rate as of August 1, 2026 (¥157 to the dollar). The fact that a company of this scale has, for two consecutive quarters, characterized AI search as "a complement, not a replacement" is reason enough to accelerate investment decisions around AI search—but it is not a reason to abandon existing Google search optimization efforts. That's because traditional search is still growing at 1.3x and continues to account for roughly one-third of all sessions.
Half of AI-driven sessions land directly on product detail pages. What this figure teaches us is a priority order: rather than category pages or feature pages, it's more effective to first ensure the structured data (product name, price, inventory, reviews, etc.) on individual product pages is well organized. Because AI agents read this structured data to narrow down product candidates, pages lacking well-organized information are unlikely to even enter the pool of candidates being explored. Given that 75% of AI-driven purchases occur outside the top 100 categories, this is a figure that small and medium-sized businesses dealing in niche products cannot afford to overlook. Compared to the arena where major players fight over top search-result rankings, the pathway where AI agents directly locate individual products is a much better fit for small and medium-sized Japanese e-commerce operators handling long-tail products.
When considering where to invest a limited budget, it makes sense to start with the low-cost measure of structured data first. Adobe Analytics' research also shows that the average conversion rate for AI-driven traffic is 54% higher than for non-AI traffic. There's no need to tackle large-scale advertising spend or system overhauls first.
Maintain your Google search optimization efforts while beginning to organize structured data on product detail pages, and revisit your investment allocation once the absolute share of AI-driven traffic is disclosed—or corroborated by independent data. Rather than being swayed by relative figures like multipliers, moving forward step by step based on verifiable facts is the most reliable course of action that Shopify's own numbers point to.