Shopify at 112 Times Earnings: The AI Shopping Boom Is Real, the Price Leaves No Room for Error

Add preferred source
Shopify has become the invisible plumbing behind AI shopping assistants — its catalog of more than a billion products feeds ChatGPT, Copilot, Gemini and Meta’s Muse — and the early data looks extraordinary: orders from AI search up roughly 15x since January 2025, brand websites beating marketplaces by four to eight times on AI referral traffic, and five straight quarters of 30%-plus GMV growth at about $115 billion per quarter. Hosts Shawn O’Malley and Daniel Mahncke of The Intrinsic Value Podcast walk through the bull case — the Universal Commerce Protocol co-launched with Google, the buyback that finally offsets stock compensation, and enterprise wins from GM to Starbucks — then lay out the bear cases that still matter: Amazon’s conflicting incentives, Meta becoming the first platform to put a toll on the transaction, and the risk that AI turns from a referral channel into a distributor sitting between merchant and customer. The verdict, however, is arithmetic: at 112 times operating profit, unchanged since they first passed on the stock, O’Malley’s fair value sits near $100 a share against roughly $160 in the market. The business beat every operating assumption; the price did not compress. That gap is the whole story — and the Muse revenue split is the variable that will decide whether Shopify’s take rate holds.

Key Elements

The business grew faster than anyone modeled. The stock did exactly what it did eighteen months ago. That combination — not a triumphant or a humbled one — is the uncomfortable conclusion Shawn O’Malley and Daniel Mahncke reached when they went back to the Shopify call they got wrong, on The Intrinsic Value Podcast.

Shopify turned over $378 billion of merchandise in 2025, up 29%, on $11.5 billion of revenue and $2 billion of free cash flow. It has now strung together five straight quarters of 30%-plus growth in gross merchandise volume at roughly $115 billion a quarter. And yet the stock trades at 112 times operating profit — precisely the multiple that made the hosts pass on it the first time.

O’Malley’s own model cannot get to the current price. “In this case, I can’t even in good conscience put numbers into my model that spit out the current share price,” he said in the earlier episode, a line replayed here. “It just feels too optimistic.” His fair value sits around $100 a share; the stock was about $160 when they recorded. He would open a tracking position near $100.

The interesting part is not the verdict. It is how much about this company the hosts were right about — and how the arrival of AI shopping agents may have handed Shopify the strongest hand in e-commerce.

The growth didn’t mature. It re-accelerated.

The original thesis was simple: Shopify would climb from small merchants up to large enterprise accounts. That migration has been faster and larger than modeled.

Mahncke offers the single line that best captures the scale: Shopify’s Q4 2025 revenue alone exceeded its entire 2020 revenue. One quarter, bigger than the pandemic-boom year, five years apart.

The enterprise names support the “forever home” pitch — the idea that once a large merchant moves onto Shopify, it never leaves: General Motors, Starbucks, L’Oréal, Estée Lauder, Coach, Michael Kors, Keurig, Dr. Pepper, Sonos, Amer Sports, Guess, Avon, Canada Goose. O’Malley frames switching platforms as a major enterprise hassle, which is exactly why signing these accounts matters more than the near-term revenue they bring.

The retention data is the tell. Merchants who reach $1 million in annual sales stay with Shopify 92% of the time; at $10 million, 97%. The cohort that joined in the first quarter of 2015 now generates 3.3 times the quarterly revenue it did then.

The economics underneath keep widening. Shopify Payments processes 68% of platform GMV, up four points in a year, live in 40 countries. Shop Pay crossed $400 billion in lifetime volume and handles more than half of U.S. payment volume. The merchant-solutions take rate rose from about 2.2% to 2.4%. On more than $400 billion of annual sales, every hundredth of a percentage point is worth roughly $40 million of revenue.

O’Malley is candid about which half he got wrong. On business quality, the hosts were directionally right: Shopify captured the up-market move, raised its take rate, and built switching costs plus an app-store network effect. On numbers, they were wrong. He had assumed deceleration; he now reckons he underestimated 2026 revenue by about $2 billion. Part of that, he says, came from covering less e-commerce at the time — Coupang, MercadoLibre, Amazon have since been added to their research — and defaulting to conservatism against his own ignorance.

The balance sheet crossed a line most investors missed

In 2025 Shopify repaid its convertible notes mostly in cash and ended the year with about $6 billion in cash and investments and no debt. In February 2026 the board authorized the company’s first-ever buyback at $2 billion; another $3 billion followed in June, taking the total to $5 billion, with roughly $1.9 billion already executed in the first half.

Mahncke frames this as a symbolic crossing: for the first time in Shopify’s public life, buybacks roughly offset stock-based compensation, moving the company from cash-burning disruptor to self-funding compounder — a transition he likens to Uber’s. O’Malley expects the share count to be flat or declining, but questions whether buying back stock at around 100 times earnings is the best use of the cash versus reinvesting it.

Shopify is building the plumbing for AI agents — not competing with them

This is where the story gets structurally interesting. Agentic commerce, as Mahncke defines it, means an AI assistant doing the search, the comparison and the checkout inside a conversation — Muse, ChatGPT, Gemini, Copilot — rather than today’s intermediate stage where a chatbot merely helps you start your research. Shopify’s CFO draws the same distinction. The fully autonomous version, where you hand an agent a budget, does not exist yet.

Shopify has built three things for that future.

The first is Catalog: a continuously updated, machine-readable index of more than a billion products from millions of merchants. O’Malley says AI searches powered by Catalog convert to purchases twice as often as those relying on scraped web pages — and that shopping inside ChatGPT today effectively means browsing Shopify’s catalog. “Everyone else has to scrape the internet for product data, basically sending bots to read web pages, while Shopify is the

Catalog also carries detail that scraping cannot reliably supply: fabric, fit, whether a shoe suits a long run. That is the qualitative layer a shopping agent needs to make a recommendation rather than return a link.

The second is the Universal Commerce Protocol, co-launched with Google in January 2026. It is an open standard covering everything from discovery through cart, checkout and returns, deliberately payment-agnostic so merchants keep their own checkout, subscriptions and discounts. Amazon, Meta, Microsoft, Salesforce and Stripe joined the body overseeing it; Target, Etsy and Wayfair opted in. Any developer can build on UCP and plug into Shopify’s catalog without approval.

O’Malley is careful here: joining costs nothing and implies no toll owed to Shopify or Google. That makes broad sign-on a low-commitment gesture rather than an endorsement of Shopify’s position.

The third piece is checkout retention. Shopify says it is currently the only platform letting merchants sell directly inside ChatGPT, Copilot and Google, managed from one place. The chatbot handles the conversation; Shopify processes the order, takes payment through Shopify Payments and handles fulfillment. O’Malley’s test is blunt: does a dollar sold through ChatGPT pay Shopify the same as a dollar sold on the merchant’s own site?

The data says AI favors the little guy — and that means Shopify

The most surprising evidence in the episode is that AI shopping may reverse the internet’s long consolidation toward Amazon.

MoffatNathanson data shows independent brand websites outpacing legacy marketplaces by four to eight times in AI-platform referral traffic, against only about two times faster in regular Google Search. “The fact that brand.com is outperforming Amazon on AI referrals is really a testament to the importance of what Shopify has built in Shopify catalog,” O’Malley said.

The mechanics matter. Traditional search ranks by popularity and rewards big SEO budgets. Agents match on specific intent. Mahncke’s example: a parent looking for a screenless phone for a 10-year-old gets routed to Tin Can Phone, a small Shopify merchant that would never have won that Google search. “AI agents, they match on specific intent,” he said. “So in some ways, AI discovery favors this long tail of small specialized merchants. And guess what? That’s exactly what Shopify’s customer base is.”

That is why 75% of AI-attributed purchases come from outside the top 100 product categories.

“AI chatbots are driving more traffic to small online stores and not less,” Mahncke said. “AI chatbots may actually be helping independent brands.”

MoffatNathanson’s Michael Morton upgraded the stock this year on this data, arguing Shopify’s GMV “is going to grow higher for longer than most people think.” O’Malley adds the mechanism that makes it durable: scraping is compute-expensive, so agents have a standing incentive to prefer well-indexed platforms.

Amazon’s incentives are the bear case that survives

The original episode’s biggest worry was Amazon’s Buy with Prime reaching third-party sites — Amazon stepping onto Shopify’s turf. That fear now looks dated, because AI has redrawn the battle lines. When shopping starts in Claude or Muse rather than amazon.com, even Amazon benefits from open rails existing. The rails, awkwardly for Amazon, were built by Shopify with Google.

O’Malley constructs the skeptical case against himself. Standards bodies are cheap to join. Amazon still owns the largest closed commerce ecosystem and will follow its incentives. It blocked Meta’s Muse from its own site, and earlier in 2026 completed a $50 billion investment in OpenAI — after which, O’Malley notes while flagging that he is speculating, OpenAI shelved its Instant Checkout experiment within weeks.

“Amazon is trying to have its cake and eat it too,” he said. “I think it wants its own agents to shop everywhere else while keeping other companies’ agents out of Amazon.”

Mahncke’s counterpoint is that $50 billion is a steep price for a checkout feature. The hosts land on a less binary view: the pie likely grows, both can win, and both could equally lose to a competitor that does not exist yet.

The real open question, as Mahncke frames it, is whether AI becomes a referral channel — a better Google, where the merchant keeps the customer, the checkout, the data and the take-rate leverage — or a distributor wedged between merchant and customer, owning the relationship. OpenAI shelved Instant Checkout and now sends shoppers to the merchant’s own site; Google and Meta have shown how lucrative the referral model is.

Muse is the test case, and Meta just blinked

The sequence is instructive. Shopify shares fell from about $155 to $130 when Muse was announced. Then Amazon blocked Muse. Then Shopify said it would partner deeply with Muse, enabling agentic checkout through Shop Pay across all Shopify stores — and the stock rose 14% in two days.

Two days after that, Mark Zuckerberg said Meta plans to take a small fee on transactions completed through Muse, without specifying the size or who pays. Shopify says its own cut will not change.

Mahncke calls Meta the first major AI platform to say out loud it wants a toll on the transaction — a step toward the distributor model. O’Malley, who has Muse on his phone and demoed it for Mahncke over a screen share, says Muse purchases on Shopify stores run through Shop Pay with the same economics as any other transaction.

That is the number to watch. If the toll comes out of the merchant’s share, Shopify’s take rate survives and the referral-channel thesis holds. If it comes out of Shopify’s, the 112-times multiple has a problem.

The smaller bear cases don’t hold up

The idea that merchants will “vibe-code” their own Shopify with AI-written software — a fear that helped drive the SaaS sell-off hitting Adobe, CoStar and Intuit — does not survive contact with real merchants, in O’Malley’s telling.

The savings are negligible against the risk. Going alone forfeits Shopify’s negotiated Stripe fee structure, which no individual merchant could match, and a one-click connection to Meta’s ad-tracking tools that O’Malley says came from a deal worked out personally between Tobias Lütke and Zuckerberg. Morton’s version, as relayed: even with unlimited resources, an army of engineers and a blank check from Anthropic, you could not rebuild Shopify more cheaply than paying to use it.

O’Malley’s practical argument is that things break. A merchant selling on TikTok, Etsy, Amazon and its own site who has replaced Shopify with a hand-rolled stack and hits a glitch on Black Friday is on their own.

Etsy is the other casualty in waiting. Mahncke notes Etsy earns more than 100% of its operating profit from onsite ads while charging sellers roughly 25% — a model he says will not last when agents, unlike human shoppers, are not swayed by advertising. Successful Etsy sellers have a cheaper path: a Shopify storefront surfaced through AI search.

Sidekick is showing up in margins — with a compute bill attached

Shopify’s built-in AI business partner, Sidekick, knows a merchant’s products, customers and transaction history, and can edit product photos, analyze data and build custom apps. Daily active merchants are up 3.5 times year over year; daily sessions nearly 5 times; nearly 34 million conversations in Q2 2026 alone; and 36,000 custom apps built by merchants through it in a single year.

O’Malley adds that nearly 10% of new Shopify shops now make their first sale within a day of signing up, up 43% in three years. His example: “Da Pope” t-shirts on sale within an hour of the election.

Mahncke is more measured. Shopify’s president has called it possibly the most AI-pilled company in the world, and such claims can be oversold. The supporting evidence: headcount has been roughly flat for two years while revenue grew more than 60%, and operating expenses fell from 42% of revenue in Q2 2024 to 34% last quarter.

O’Malley flags the offset. Lütke has acknowledged that rising AI compute costs from Sidekick usage pressure gross margins — analogous to Adobe giving away compute-intensive creative tools as a loss leader.

The arithmetic that explains the whole story

Here is the fact that ties the episode together. Shopify traded at 112 times operating profits when the hosts first looked. It trades at 112 times operating profits now. The stock doubled because earnings roughly doubled while the market kept paying the same premium.

O’Malley’s model lifts operating margin to nearly 20% within five years from 14% in 2026, with operating profit growing faster than revenue and a flat share count. For the multiple he invokes Ben Graham’s formula — the AAA corporate bond yield plus estimated earnings growth — implying roughly 35 times operating earnings is fair. He stretches to 45 times for Shopify’s quality. He still cannot justify the price.

Mahncke agrees 112 times is too rich and expects a better entry, noting the market swung Meta and Shopify from “big AI losers” to “big AI winners” within a month.

Grading the miss

Mahncke opens with the epistemics: judging a past decision requires separating a genuine analytical error from reality landing on an unlikely branch of a well-considered range. Only one version of reality exists, so the second is unprovable.

O’Malley’s sports analogy carries it — a team with a 10% comeback chance that comes back does not make the bet on it wrong — with the caveat that investing has no final whistle, so investors can cling to losing theses indefinitely. Mahncke’s discipline is writing down the thesis and the specific quarter that will prove or disprove it, as he did with PayPal, and cutting when reality drifts, per Munger’s warning. O’Malley’s Lululemon exit, triggered by faster North American share loss, is his example of a thesis that changed and was sold.

Mahncke pushes back on the framing: the hosts discuss their losers so routinely that listeners may think they have nothing else, and Shopify was a mistake of omission rather than commission. He also does not fully accept that they were wrong, since refusing to chase stocks priced for best-case outcomes is sound even when the best case arrives. The counterexample is Trade Desk, whose chart was up-and-to-the-right until it fell sharply right after their episode — down 91% over two years.

O’Malley’s summary of the tension: giving a company credit for its track record is right for Shopify and wrong for Trade Desk, and the line between them is fine.

“In this case, I can’t even in good conscience put numbers into my model that spit out the current share price,” O’Malley repeated.

“Being too far ahead of your time is indistinguishable from being wrong,” he added, quoting Howard Marks.

The unresolved tension is the one worth tracking. The hosts’ entire case for passing rests on a multiple that has not compressed in eighteen months while the business beat every operating expectation they set. If Shopify keeps compounding earnings at 30%-plus and the market keeps paying 112 times, the wait-for-a-fair-price discipline will look wrong for longer still. And the Muse economics — specifically whether Meta’s transaction toll comes out of the merchant’s share or Shopify’s — is the concrete variable that will determine whether AI is a referral channel or a distributor for Shopify’s take rate. That is what an investor should watch next, not the multiple.

Full content available at:TIVP105: Shopify (SHOP): The Big Winner of Agentic Commerce? w/ Shawn O’Malley & Daniel Mahncke

• US Stocks: All Three Major Indexes Rise as SpaceX and Super Micro Earnings Lift AI Shares
• SpaceX Eyes Access to 80,000 Telecom Towers; All Three Major U.S. Telecom Stocks Fall About 4%
• Japan’s Nikkei Stock Average briefly surges more than 900 points to 64,172
• Anthropic to List on Nasdaq in Mid-November, Raising May Reach ¥16 Trillion