Dear Software: The War Called. It Wants Its Excuse Back.
Dear Software: The War Called. It Wants Its Excuse Back.
Markets & Macro — April 23, 2026
TO: Enterprise Software Sector
FROM: The Market
RE: Q1 2026 Earnings Season — A Structural Concern
CLASSIFICATION: Urgent
Let's be precise about what happened to ServiceNow last night, because the headline — beat on revenue, beat on cRPO, raised full-year guidance, stock down 14% — reads like a logic error until you understand what investors were actually voting on.
NOW reported $3.77 billion in Q1 revenue, up 22% year-over-year. Subscription revenue came in at $3.67 billion, ahead of consensus. Remaining performance obligations hit $27.7 billion, up 25%. The company raised its full-year subscription revenue outlook. By every conventional measure of a healthy SaaS business in 2026, this was a good quarter. ServiceNow got punished like it had confessed something.
What it confessed was this: non-GAAP gross margins came in at 79.5%, down from 80.5% the prior quarter. And tucked inside the earnings release, a line that did real damage — Q1 subscription revenue growth suffered roughly a 75 basis point headwind from delayed closings of several large on-premise deals in the Middle East, due to the ongoing conflict. The company said it expects those geopolitical headwinds on deal timing to persist for the rest of 2026.
Seventy-five basis points. A war disrupted enterprise software contracts in the Gulf. And for a stock already down 34% year-to-date, already on trial for its future relevance, that was sufficient.
The SaaS-pocalypse thesis — the market's working hypothesis that AI agents will hollow out traditional enterprise software the way streaming hollowed out DVD rental — has been grinding through valuations since February. ServiceNow has been one of its central exhibits. The irony being that ServiceNow is also one of the more credible AI monetisation stories in the sector: its Now Assist AI annual contract value more than doubled year-over-year last quarter; it has a stated $1 billion ACV target for AI products in 2026; the deal count for million-dollar-plus contracts went from 72 in Q1 2025 to 244 in Q4. That is not a company being disrupted. That is a company — theoretically — doing the disrupting.
But the market has stopped rewarding what happened and started taxing what might happen. The fear is not that ServiceNow is losing today. The fear is that the floor is moving. That the 82% subscription gross margin that was once the envy of the sector could drift toward 79%, then 76%, as AI infrastructure costs rise faster than AI-related revenue. When margins compress — even marginally, even with revenue accelerating — in an environment of existential suspicion, the stock doesn't get the benefit of the doubt. It gets sold.
Down 14% after hours. Down roughly 33% for the year. Trading at $88 and change.
Thirty of thirty-five Wall Street analysts still have a Buy rating on this stock. The consensus price target implies more than 65% upside from current levels. That gap between analyst conviction and market price is either the trade of 2026 or the most expensive consensus error since someone rated Peloton a Strong Buy in January 2022.
Meanwhile, across the sector, United Airlines filed a different kind of damage report. UAL beat Q1 estimates — $1.19 adjusted EPS against $1.09 expected, revenue of $14.6 billion against $14.45 billion forecast. Premium revenue up 14%. Loyalty revenue up 13%. Genuinely good numbers for a global carrier operating through a war that doubled its fuel bill in a quarter.
The full-year EPS guidance got cut from $12–$14 to $7–$11. That $4 midpoint reduction reflects a $340 million fuel cost increase in Q1 alone. Jet fuel was $2.39 a gallon on February 27th, the day before the first strikes on Iran. It hit $4.78 on April 2nd. It sits at $3.51 today. The graph of that price spike is also, roughly, the graph of every airline's earnings revision this season.
CEO Scott Kirby's pitch is that United can pass through 40–50% of elevated fuel costs in Q2, 70–80% in Q3, and 85–100% by Q4. The pricing power argument has some merit — premium demand is holding, and unit revenue is up across every segment including domestic. But the model requires yields to increase 15–20% to fully offset sustained higher fuel. That is a very large ask from an economy where the IMF has already cut global growth to 3.1% and UK inflation just hit 3.3%. At some point, the consumer paying the yield increase is the same consumer getting squeezed by the energy shock.
The airline is cutting 5% of planned capacity in the back half of 2026. Pulling seats off the market to protect what margin it can. It is a logical response to an impossible situation. It is also exactly what you do when you have stopped growing and started managing.
Here is the underlying architecture of this earnings season, now clearly visible:
Companies exposed to physical energy — airlines, industrials, shipping — are reporting strong demand and horrifying cost structures simultaneously. The revenue is there. The margin is not. Their guidance ranges have become comedically wide ($7 to $11 for UAL is a 57% spread) because nobody can price oil in a ceasefire that may or may not hold.
Companies building AI software are reporting strong top-line growth and margin pressure simultaneously. The revenue is there. The story is there. The market doesn't care, because it is busy stress-testing what those businesses look like in three years if a better-capitalised AI-native competitor undercuts the entire pricing stack.
In both cases, the war is doing something that will outlast the ceasefire: it has introduced genuine uncertainty into forecasting, and markets hate uncertainty more than they hate bad news. Bad news can be priced. Uncertainty cannot.
Alphabet, Amazon, Meta, and Microsoft report next Wednesday. If the hyperscalers show AI monetisation accelerating — revenue growing faster than infrastructure spend — the software selloff may have finally found its floor. If margins disappoint there too, the SaaS-pocalypse becomes a self-fulfilling rout.
ServiceNow beat its quarter. The market didn't care.
That is the sentence that explains 2026 so far.
Thursday, April 23, 2026