TD Cowen raises Snowflake stock price target on AI tool momentum

TD Cowen raised Snowflake’s price target to $370 from $300 and kept a Buy rating as the stock trades at $325.01 (+48% YTD) ahead of its Sept. 2 Q2 earnings report. The firm pointed to bullish “data cloud” momentum tied to AI catalysts like CoCo (introduced in February) and Cortex AI Gateway (released last month), alongside upward earnings revisions (31 analysts raised estimates) and expectations for profitability this year. Street projections call for ~36–37% YoY Q2 revenue growth, with FY2027 revenue growth projected in the high-30% to 40% range, supporting continued valuation pressure.

Analysis

Snowflake is now trading more like a consensus AI compounder than a cyclical software name, which changes the burden of proof: the stock needs evidence that AI features are not just engagement tools but are pulling through to higher consumption and better retention. In the next 1-3 months, the key mechanism is not revenue growth itself but whether management can show acceleration in large-account spend and attach rates from AI workloads; if that comes through, the multiple can stay elevated even with only modest EPS progress.

The second-order effect is competitive: if Snowflake’s AI layer is gaining traction, it pressures adjacent data/analytics vendors that monetize the same budget pool, especially names with weaker cloud-native consumption profiles. The likely losers are not obvious from the headline, but any slower-growing enterprise data platform with less visible AI monetization will face relative multiple compression as capital rotates toward the clearest AI-through-revenue story. The biggest structural winner may actually be the hyperscaler ecosystem if Snowflake’s usage growth translates into more underlying cloud compute demand.

The risk is expectation saturation. At this valuation, a merely solid quarter can disappoint if the market has already priced in a step-function in AI monetization; that risk is highest into the September print and over the subsequent 30-60 days if guidance remains unchanged. The contrarian view is that the current move may be underestimating the durability of usage-based AI spend: if AI features become embedded in production workflows, the spend curve could remain linear-to-upward for several quarters rather than fading after initial adoption.

What would falsify the bullish case is any sign that the AI narrative is not converting into core consumption metrics: weaker net retention, softer remaining performance obligations, or guidance that implies growth deceleration despite favorable product checks. For the next 6-18 months, the stock likely trades on whether Snowflake can prove that AI is expanding the addressable wallet rather than just improving sentiment.

Key Decisions for Investors

• Stay long SNOW into the September 2 earnings window only if pre-earnings channel checks continue to point to accelerating consumption; otherwise reduce exposure into the print because expectations are now high enough that a merely in-line quarter can de-rate the stock 10-15%.
• For a relative-value expression, pair long SNOW vs. short a basket of slower-growth data/analytics software names (e.g. MDB, DDOG) over the next 1-3 months; thesis is that AI monetization visibility should support SNOW’s multiple premium while peers remain vulnerable to budget scrutiny.
• If you want upside convexity without paying full spot risk, consider a call spread on SNOW through the earnings date; structure should target a post-print re-rating while limiting loss if the AI narrative fails to convert into hard consumption metrics.
• Set a falsification alert on any guide that implies growth deceleration or flat-to-down net retention; if that happens, fade the AI multiple expansion and expect a rapid 8-12% reset as crowded longs de-risk.
• Watch MS and the broader software factor exposure as a sentiment hedge: if the market starts punishing expensive AI stories on weak prints elsewhere, SNOW’s beta to the 'show me' AI trade increases and the entry point likely improves after the event.