PayPal walked into its May 5 Q1 earnings call with a beat on both lines and walked out with a layoff announcement. Adjusted EPS came in at $1.34 versus a $1.27 consensus; revenue at $8.35 billion versus a $8.05 billion consensus. Net income was $1.11 billion, down 14% year-over-year from $1.29 billion. The stock did not love any of it.
Alongside the print, new CEO Enrique Lores — installed in March after the board ousted Alex Chriss — laid out the plan. Roughly 20% of the workforce, somewhere around 4,760 of the 23,800 employees on the books at year-end 2025, will exit over the next two to three years. Target: at least $1.5 billion in gross run-rate savings by 2028. Framing: PayPal needs to “become a technology company again.”
That last phrase is doing a lot of work, so it is worth pulling apart.
Lores’s pitch — and the HP playbook he is running
Lores ran HP for six years before the PayPal seat opened up. The pattern there was consistent: cut complexity, push toward subscription and software margins, lean into AI as the cost-out story, and use the savings narrative to give investors a number to anchor on while the top line slowly recovers. The HP playbook is what is showing up at PayPal, almost line for line.
The Q1 numbers are not the problem he was hired to solve. The problem is that PayPal’s transaction-margin business — the actual fee-bearing payments — has been compressing for six straight quarters as Apple Pay, Shopify, and platform-native checkout eat the share PayPal used to win by default. The $8.35 billion top line obscures the fact that branded checkout volume growth has been single-digit while unbranded (Braintree) volume growth has been double-digit at much thinner margins. The blended business looks fine. The mix underneath does not.
The cut is what Lores is selling investors instead of a credible top-line reacceleration story. That is also what the HP cuts were — a margin-expansion narrative deployed while the business slowly rotates into a different shape.
Three units, one AI team, and the joke nobody is making out loud
The reorganization simplifies PayPal from a tangled product matrix into three business units: Checkout Solutions & PayPal (branded checkout, the core), Consumer Financial Services & Venmo (peer-to-peer plus consumer credit), and Payment Services & Crypto (Braintree, processing, and the crypto wallet stack). Each unit gets its own GM with margin-and-growth accountability.
Sitting alongside those is a new “AI transformation and simplification” team — Lores’s name for the cross-cutting group that is supposed to pull cost out of every workflow inside the three units.
Then comes the dryly funny part. According to employee discussions surfaced by The Workers Rights and corroborated by Storyboard18’s reporting, the first three groups in line for the cuts are: merchant support, customer service, and the company’s existing internal AI team.
Read that list twice. The internal AI team — the people who have spent the last two years building PayPal’s AI capabilities — is being wound down at the same moment a new “AI transformation” team is being stood up. The names changed; the headcount did not survive the transition. This is the standard reorganization pattern: the strategic frame moves, the org chart is redrawn, the people who built the previous frame’s outputs are made redundant by the change of frame itself.
The merchant support and customer service cuts are the more interesting ones for the displacement question. Both are categories where AI agents are a credible substitute for the bottom 60–70% of inbound volume — password resets, refund status, integration questions with deterministic answers. Neither is a category where current agents can credibly handle the top 30–40% — chargeback disputes that touch regulator-relevant edge cases, merchant-side integration breakage that requires reading partner code, fraud-investigation calls. The number to watch over the next two prints is whether merchant NPS and customer-service CSAT hold or drop after the cuts land.
The Q1 print revealed the bigger problem
The earnings beat was real, but the market reaction was not friendly. PayPal shares traded down on the print despite the EPS and revenue beats, which is the market’s way of saying: the beat was not the news, the cut was the news, and the cut is what investors had already priced in expectation of.
That’s the box Lores has built for himself. The $1.5 billion savings target is the floor. If Lores delivers exactly $1.5 billion of savings over two to three years, the stock probably does not move much, because Wall Street will assume the savings are already in the multiple. To re-rate PayPal up, he has to deliver more than $1.5B AND show evidence that the revenue mix is rotating toward the higher-margin branded checkout end of the business.
The piece nobody is selling on the call is that this is now a two-cut company. PayPal already cut roughly 9% of its workforce in early 2024 under Chriss. Adding 20% on top puts the cumulative reduction at close to 29% from the post-pandemic peak. There is a point at which a payments company stops being able to support the merchant network, the fraud-operations stack, and the regulator-facing compliance load with the headcount it has left, and the question is whether the third quarter of cuts is the one that crosses it.
The AI-driven cut is the easy story; the AI-driven business model is the hard one
Most of the May 5 press coverage stopped at the layoff number. The harder thing to evaluate, which Lores hinted at on the call, is what an “AI-native” PayPal looks like as a product, not just as a cost structure.
PayPal’s payment rail is mostly invisible to the consumer at this point — the brand sits inside checkout flows owned by other companies. The AI-leverage opportunity at the product layer is in the parts of the business consumers actually touch directly: the Venmo P2P interface, the consumer credit underwriting (Pay Later), and the crypto wallet. None of those three was singled out as an AI-investment beneficiary on the call. The cuts are funding margin, not funding product.
That is the strategic gap worth watching. Coinbase’s May 5 announcement at least committed to a specific operating-model change — the “AI-native pod.” Anthropic’s $50B coding-agents bet is a product-level claim about what AI does for software development. PayPal’s announcement is a number — 4,760, 20%, $1.5B — without a product-level claim about what the AI is actually for inside PayPal’s user-facing surfaces.
The number is concrete; the product story is empty. Lores has nine months until the next Q1 print to fill it in.
What to watch through Q3
- Merchant churn through summer. If PayPal’s merchant-services support quality slips after the merchant-support cuts, the place it shows up first is in mid-market merchant churn — the segment most likely to switch to Stripe or Shopify Pay when their account manager goes silent.
- Q2 transaction margin. The cost savings won’t fully ramp until 2027. Q2 should show small margin tick-up if the early cuts are real and on schedule. If margin is flat in Q2 the timing slips, and the $1.5B narrative gets harder to defend.
- AI transformation team’s first deliverable. A team stood up on May 5 should have its first shipped product or measurable KPI by the August earnings call. If the only thing it delivers by Q2 is a deck, the AI framing is theatre.
- Lores’s narrative discipline on the August call. The risk between now and the next print is that the savings target keeps growing — $2 billion, $2.5 billion — as a way to keep the cost story alive. Watch whether Lores anchors to $1.5B or starts inflating it.
The dry coda
PayPal’s May 5 was a fine quarter wrapped in a bad announcement. The earnings beat. Revenue grew. The free cash flow held. None of that mattered, because the company is no longer being valued on the trajectory of its existing business; it is being valued on the credibility of Lores’s turnaround story.
The story he is telling is the same one HP told under his watch and the same one Microsoft, Meta, Oracle, Coinbase, and Freshworks have told over the past nine months: AI lets us run the same business with fewer people. The Coinbase variant pushed it to “one human plus a fleet of agents.” The PayPal variant is more conservative — 20% out, AI integrated into development processes, a margin story.
The conservatism is probably correct for a regulated payments company. It also leaves Lores with a number to hit and very little room for the number to move. The 4,760 figure is precise. The $1.5 billion figure is precise. The “becoming a technology company again” framing is the part that is still a sentence rather than a roadmap.
August is the first checkpoint.