Three Enterprise Software Stocks Priced For Perfection: The Risk Desk Scan

Snowflake revenue grew from $3.6 billion in FY2025 to $4.7 billion on a trailing basis, a 31% year-over-year growth rate. That is strong in absolute terms but marks a continued deceleration from the 70%-plus growth rates of FY2022-23. The company remains unprofitable on a GAAP basis with operating losses of $1.4 billion despite the large revenue base.

The specific risks facing Snowflake are material. First, data warehouse functionality has commoditised. AWS Redshift, Google BigQuery, Azure Synapse, and Databricks offer increasingly comparable capabilities at lower prices. Snowflake's premium positioning requires continuous innovation to justify the price point. Second, AI workloads consume compute differently than traditional analytics workloads; much of the incremental AI compute spend is flowing to specialised infrastructure (Nvidia, Databricks) rather than to Snowflake. Third, customer negotiations have shifted; enterprise customers are using competitive threats to extract pricing concessions at renewal.

At $137 per share and 80x forward earnings, Snowflake needs to reaccelerate revenue growth to 35%-plus to justify the multiple. The Risk Desk models FY2027 revenue growth at 26-30%. That trajectory does not support the multiple. Fair value range $100-120 per share, current price implies 15-27% downside to our fair value.

Datadog revenue reached $3.4 billion in FY2025, growing at 25%. The company has achieved modest GAAP profitability with operating margin approaching 1%. Datadog's differentiator is observability across cloud infrastructure; it monitors applications, infrastructure, and security events with an integrated dashboard that enterprise customers value.

The specific risks facing Datadog are narrower than Snowflake's but still meaningful. First, AWS and Azure have been improving their native monitoring tools; for customers willing to accept modestly less capability in exchange for lower cost, native tools are increasingly adequate. Second, security-focused monitoring is being absorbed by dedicated security platforms (CrowdStrike, Splunk); Datadog's security module has not gained share as aggressively as the monitoring side has. Third, cost-conscious enterprise buyers are negotiating harder on multi-year Datadog contracts than in prior cycles.

At approximately $147 per share and 59x forward earnings, Datadog trades at a premium to the fair value band of $110-130 that our model produces. The Risk Desk is cautious on the stock at current levels though less bearish than on Snowflake given the stronger profitability trajectory.

CrowdStrike revenue reached approximately $4.8 billion in FY2026 (ending January 2026), growing at 21%. The July 2024 global outage incident created a multi-quarter customer retention concern that has largely been resolved; retention rates have recovered to pre-incident levels. The core product suite remains differentiated and growing, particularly the Falcon platform's AI-driven threat detection.

The specific risks for CrowdStrike are somewhat different from the other two names. First, security spending has bifurcated; large enterprises continue investing aggressively while SMB customers face budget pressure. CrowdStrike's mix has shifted modestly toward larger enterprises, providing some insulation but also creating concentration. Second, Microsoft has been bundling Defender with E5 licences at attractive economics, creating competitive pressure at the lower end of the market. Third, the July 2024 incident created internal compliance and process costs that continue weighing on operating margins.

At approximately $430 per share and 88x forward earnings, CrowdStrike is the most expensively priced of the three names. The Risk Desk fair value range is $320-380, implying 12-26% downside. The business is higher quality than the multiple deserves to be.

Across the three names, CrowdStrike has the cleanest long-term fundamental story but the most expensive current multiple. Snowflake has the cleanest valuation adjustment needed but the highest fundamental deceleration. Datadog sits in the middle on both dimensions.

The three names share exposure to enterprise cloud spending growth, which has been decelerating across the past eighteen months. The deceleration is not catastrophic; enterprise cloud spending is still growing at 12-16% annually depending on measurement source. But the 30%-plus growth rates that characterised the 2021-2023 period are over. The narrower growth rate compresses the expansion potential for each of these names.

Specifically, enterprise CIO surveys indicate a shift in spending priorities. Budget is being reallocated toward AI infrastructure at the expense of traditional cloud services. That reallocation benefits Nvidia, AMD, Databricks (private), and Azure AI services. It works against Snowflake and Datadog at the margin. CrowdStrike is less affected given security spending exists on its own budget line.

The Risk Desk has historically seen this pattern; narrative-driven premium multiples compress when the underlying spending growth moderates. The cloud software premium has moderated gradually but not fully in these three names. The full adjustment is still ahead.

Across three prior software multiple compressions (2001-2003, 2008-2010, 2022-2023), the pattern has been that names trading above 50x forward earnings have underperformed the broader software sector by 20-40 percentage points during the compression window. The compression typically takes 18-24 months from the initial deceleration signal to the full multiple reset.

The current cycle's deceleration signal began in late 2024. By that timing, the full compression should complete by mid-2026. The Risk Desk is positioned for continued relative underperformance from premium-multiple software through that window. The opportunity sets improve once the compression completes; quality software businesses at normalised multiples are attractive long-term holdings. These three names would become attractive at the fair value ranges identified above.

The analogy we reference most often is Workday in 2022. Workday traded at 140x forward earnings in November 2021, then experienced a 55% drawdown over eighteen months as revenue growth decelerated from 20% to 17%. The multiple compressed to 40x, at which point the stock stabilised. The eventual re-rating to current 32x was supplemented by continued execution rather than multiple expansion. These three names are earlier in a similar cycle.

Dollar-based net retention rate is the cleanest metric for enterprise software health. It measures how much existing customer spend grows (or shrinks) year over year, before new customer acquisition. For the three names in this scan, the trends are: Snowflake at 126% in the most recent quarter, down from 165% in FY2022. Datadog at 116%, down from 130% in FY2022. CrowdStrike at 113%, relatively stable.

Net retention is the leading indicator for future revenue growth. When net retention approaches 100%, revenue growth decelerates toward new customer addition rates alone. Snowflake's net retention trajectory is the most concerning; the 39-point decline over three years suggests continued compression ahead. Datadog's 14-point decline is more manageable but still negative. CrowdStrike's stability is the most constructive in the set.

The mathematical implication is that Snowflake's reported growth rate will likely decelerate below 25% within four to six quarters. Datadog's will decelerate toward 20-22%. CrowdStrike's should stabilise around 18-22%. These trajectories are what produce the fair value estimates in the scan. The multiples need to compress to match.

Each of the three names has been positioned as an AI beneficiary. The reality is more mixed. Snowflake's Cortex AI product is generating modest incremental revenue ($200-300 million annualised estimated) but is also cannibalising some traditional warehouse queries. Net contribution is small. Datadog's AI monitoring features are adding product depth but have not meaningfully accelerated net new customer acquisition. CrowdStrike's AI-driven threat detection is a genuine differentiator but has been partially absorbed into existing product pricing rather than creating new revenue streams.

The AI monetisation paradox is that enterprise software companies need to show AI revenue contribution to support their multiples, while simultaneously absorbing AI investment costs that pressure near-term margins. The three names in this scan each face this tension. Snowflake faces it most acutely; AI-related capex has grown faster than AI-related revenue.

The Risk Desk's base case is that AI monetisation for these three names is real but smaller in absolute magnitude than the market multiples imply. The incremental $1-2 billion of AI revenue each company might add over three years does not justify the 50-100% multiple premiums they currently carry.

For investors currently holding any of the three names at overweight levels, the Risk Desk recommends trimming toward market-weight. Capital gains tax considerations matter; for tax-inefficient accounts, consider gradual reduction rather than immediate exit. For tax-deferred accounts, the reduction can be completed more quickly.

For investors building new software sector exposure, avoid these three names at current prices. Look instead at quality names trading closer to fair value: Microsoft, ServiceNow (at lower relative multiples), or Salesforce for longer-duration software exposure. The opportunity in the high-multiple cohort is ahead of the compression, not inside it.

The sector scan does not call for exiting enterprise software entirely. The sector will remain a meaningful component of any growth-oriented portfolio. The scan recommends more selectivity within the sector and patience on entry points for these three specific names.

Each of these three names will provide specific quarterly data points that the Risk Desk will monitor to update the scan. For Snowflake: sequential Cortex AI revenue disclosure, customer count growth versus consensus, and net retention rate. For Datadog: security module revenue contribution, multi-module customer penetration, and pricing per workload trends. For CrowdStrike: retention rates post the July 2024 incident anniversary, module attach rates for new modules launched in the past twelve months, and net new customer acquisition.

If any of these data points shows meaningful acceleration, the Risk Desk will update the scan in that name's favour. If the trajectories continue on the current slope, the multiple compression thesis will strengthen. The next update is targeted for post Q1 2026 earnings for all three names.

The consistent theme across the scan is that premium multiples require accelerating or sustained-high growth rates. Deceleration, even from still-healthy absolute levels, is incompatible with 60-90x forward multiples. That is the core of the thesis.