Confluent (CFLT) Q4 2025 earnings review
IBM Acquisition Caps a Year of Decelerating Cloud Growth
Confluent agreed to be acquired by IBM for $11 billion ($31.00/share), ending its run as a standalone public company. The Q4 results highlight why a sale may have been timely: while profitability surged (8.8% operating margin vs 5.2% a year ago), the core growth engine—Confluent Cloud—continued to decelerate, dropping to 23% YoY growth from 38% in 24Q4. The company beat revenue estimates, but the narrative has shifted entirely from 'growth at all costs' to 'strategic integration' within IBM's portfolio.
🐂 Bull Case
The $31.00/share all-cash offer from IBM represents a definitive outcome for shareholders, removing execution risk in a decelerating growth environment.
Confluent proved its business model works, delivering $27.6M in Non-GAAP operating income (8.8% margin) and $35.5M in adjusted free cash flow in Q4, up significantly from prior years.
🐻 Bear Case
Confluent Cloud revenue growth has decelerated for five consecutive quarters (38% → 23%). The 'hypergrowth' phase appears over as large customer optimization persists.
Growth in large customers (>$100k ARR) slowed to 10% YoY (1,521 customers), suggesting the high-end market saturation or increased competitive friction.
⚖️ Verdict: ⚪
Neutral. The operational results show a mature company trading growth for profitability, but the stock is now tethered to the $31.00 acquisition price. Fundamentals matter less than deal consummation.
Key Themes
Acquisition by IBM
IBM will acquire Confluent for $11 billion enterprise value ($31.00/share), expecting to close by mid-2026. This move validates the data streaming category but signals that Confluent management likely saw limited standalone upside given the slowing cloud growth trajectory.
Cloud Growth Deceleration
Decelerating. Confluent Cloud revenue growth slowed to 23% YoY in Q4, down from 38% a year ago and 24% in Q3. While total subscription revenue grew 20%, the cloud engine—central to the valuation multiple—is losing steam faster than expected.
Operational Efficiency
Accelerating. Management has successfully pivoted to profitability. Non-GAAP operating margin expanded to 8.8% in Q4 (up 360 bps YoY). For the full year 2025, operating margin reached 7.4% compared to just 2.9% in 2024, demonstrating significant operating leverage.
Agentic AI Demand
CEO Jay Kreps cited 'emerging agentic AI applications' as a key driver for mission-critical workloads. The company is positioning its data streaming platform as the 'context layer' for AI, though financial quantification of this specific segment remains vague in the press release.
Large Customer Saturation
Decelerating. The count of customers with >$100k ARR grew 10% YoY to 1,521. This is a slowdown from the 12% growth seen in 25Q1 and 24Q4, suggesting it is becoming harder to land or expand large enterprise accounts at the same pace.
Other KPIs
Decelerating. Up 23% YoY, but sequentially growth is flattening. This segment now represents 56% of total subscription revenue, up from 55% a year ago.
Stable. Margin of 11.3%, roughly flat vs 11.1% in 24Q4. The company has established a consistent cash generation profile, a key requirement for M&A attractiveness.
Stable. Full year growth of 21% YoY. The company successfully crossed the $1B revenue threshold in 2025.
Guidance
Due to the pending acquisition by IBM, Confluent is not providing financial guidance for Q1 2026 or the full fiscal year 2026. No conference call will be held.
Key Questions
Cloud Growth Floor
Cloud revenue growth decelerated to 23% this quarter. Was there a specific churn event or further optimization from a large customer that drove this, or is this the new normal for the category?
Deal Rationale vs. Standalone Plan
Did the Board view the decelerating growth profile as structural, prompting the sale to IBM, or was the $31/share offer simply too high to refuse relative to internal standalone projections?
Flink Monetization
How material was Flink revenue in Q4? Management previously cited it as a major growth vector—did it fail to offset the core Kafka deceleration?