Revenue reaccelerated to 34% growth and AI consumption is scaling, but Databricks competition and elevated stock-based compensation have lowered our conviction.
Our previous analysis in "Snowflake's Cortex AI Is Creating Value the Market Hasn't Priced" argued that the Cortex AI platform would drive a reacceleration in consumption growth as enterprises deployed AI workloads on Snowflake's data cloud. That thesis was directionally correct but the timing was aggressive. Consumption growth did reaccelerate, but from a lower trough than we expected, and the path to profitability has extended.
The updated numbers: revenue grew to $4.68 billion in fiscal 2025, up from roughly $3.5 billion in the prior year, representing approximately 34% growth. Net margins remain deeply negative at minus 28.4%, though this is an improvement from the prior year's minus 35%. Free cash flow has not yet turned positive on an annual basis. At $49.9 billion market capitalisation and 80.6x forward earnings, the stock is priced for a recovery that has not yet fully materialised.
Three things have changed since our last analysis. One is constructive. Two raise new questions.
Our Cortex AI thesis rested on three pillars: that AI workloads would drive incremental data consumption on Snowflake's platform, that Cortex AI features (vector search, LLM functions, ML model serving) would differentiate Snowflake from competing data platforms, and that the consumption-based pricing model would translate AI adoption directly into revenue acceleration.
Pillar one has largely played out. AI-related consumption on Snowflake has grown faster than traditional analytics workloads, and the company reports that customers using Cortex features spend 20-30% more than those who do not. The flywheel is working: enterprises bring data into Snowflake for analytics, discover they can run AI workloads on the same platform, and increase consumption accordingly.
Pillar two is facing competitive pressure. Databricks has aggressively expanded its AI capabilities, and the open-source Apache Iceberg table format has reduced Snowflake's data lock-in advantage. Snowflake's decision to support Iceberg natively was strategically necessary but removes one of the switching cost barriers that made the platform sticky.
Pillar three is confirmed but with a lag. Revenue growth reaccelerated from approximately 30% at the trough to 34%, and remaining performance obligations (RPO) growth exceeded revenue growth, a forward indicator that consumption commitments are increasing.
TickerXray Report
Get the complete Snowflake report with all 12 quantitative models, AI-generated investment thesis, and real-time data.
The constructive change: dollar-based net revenue retention rate has stabilised and appears to be inflecting upward. After declining from 170%+ to approximately 125% during the cloud optimisation cycle, retention has ticked up to an estimated 127-130%. That inflection, if sustained, has significant implications for the revenue growth algorithm. At 130% net retention with 15-20% new customer growth, the math supports 30-35% total revenue growth sustainably.
The first concerning change: the competitive landscape with Databricks has intensified materially. Databricks raised additional funding at a $60+ billion valuation and is investing aggressively in both AI/ML capabilities and traditional analytics (historically Snowflake's core strength). The overlap between the two platforms is increasing, and price competition is emerging in segments where both platforms are technically adequate. Price competition in cloud data platforms compresses margins for everyone.
The second concerning change: stock-based compensation remains elevated. SBC expense consumes approximately 40-50% of revenue, which means GAAP profitability is years away even as non-GAAP margins appear to improve. For a company with a $49.9 billion market cap, the annual SBC expense represents significant ongoing dilution that reduces the per-share value of any future earnings recovery. The adjusted free cash flow margin (excluding SBC effects) overstates the true economic profitability of the business.
At 80.6x forward earnings and approximately 10.7x revenue, Snowflake is priced between the optimistic and pessimistic scenarios we outlined in our previous analysis. The stock has underperformed both Datadog and CrowdStrike over the past twelve months, reflecting the market's concern about the competitive dynamics and profitability timeline.
For the valuation to work from current levels, Snowflake needs to sustain 30%+ revenue growth and demonstrate a clear path to 20%+ operating margins (excluding SBC) within the next 2-3 years. The revenue growth trajectory supports this; the margin trajectory is less clear.
Compared to Databricks' private valuation of $60+ billion on similar revenue, Snowflake's public market valuation of $49.9 billion appears reasonable. But private market valuations are often inflated by round dynamics and are not directly comparable. Relative to public cloud software peers, Snowflake trades at a premium to its growth rate that requires continued execution.
The 52-week range of $100 to $197 reflects the stock's volatility and the market's uncertainty about the fundamental trajectory. The 50-day moving average at roughly $140 and 200-day at $165 show the stock in a downtrend from its recent highs.
Our Cortex AI thesis for Snowflake was directionally correct: AI workloads are driving consumption reacceleration, and the platform's data gravity advantage is real. But the competitive dynamics with Databricks, the elevated SBC expense, and the extended timeline to GAAP profitability have reduced our conviction from high to moderate.
At 80.6x forward earnings, the stock requires continued 30%+ growth and improving margins to justify the multiple. If Databricks' competitive pressure causes growth to decelerate below 25%, the stock could decline 30-40% to the $90-100 range. If Cortex AI drives consumption above expectations and margins improve, the stock could reach $180-200.
We are moving from buyers to holders at current levels. The risk-reward has shifted toward neutral since our last analysis. We would become buyers again on a pullback to the $115-125 range, where the valuation provides more margin of safety for the execution risks we have identified. For new positions, Datadog offers a better risk-adjusted entry point in the cloud data infrastructure space.
Full forensic analysis of Snowflake
+ 6 more models included
150,000+ stocks covered
Global coverage across 60+ exchanges. Every report includes all 12 quantitative models and AI analysis.
View plansEvery report runs 12 quantitative models and generates an AI investment thesis. From Piotroski scores to manipulation detection -- get the full picture in seconds.
12 forensic models
Piotroski, Altman, Beneish, DuPont & more
AI investment thesis
Synthesized outlook on every stock
Manipulation detection
Spot red flags before they hit the news
150,000+ tickers
Global coverage across 60+ exchanges
Expected return
Forward return projections for every stock
Real-time data
Live prices, insider trades, news sentiment
Free accounts get 1 report per month. Pro gets unlimited.
Snowflake at 80x, Datadog at 59x, CrowdStrike at 88x. All three face decelerating revenue growth. The fair value gap to the current price is wider than the premium.
Snowflake, Datadog, and Cloudflare sit at the centre of the broadening software rally identified on 17 April. Each has a distinct setup, but the common thread is a free cash flow inflection that consensus has only partially recognised.
Net retention has stabilised, AI workloads are inflecting at 300% growth, FCF margins hit 25%, and the CEO transition is behind it. The four biggest bear arguments are weakening.