3 Incredible Growth Stocks to Buy Now

The article highlights improving fundamentals in three growth names: Plug Power expects a swing to profitability by late 2028, with 2025 revenue growing 13% and net loss improving 20% vs. 2024. ServiceNow reports Q1 non-GAAP revenue of nearly $3.7B (+19% YoY) with subscription renewal rates at/above 97%, and has a consensus target of $140.38 (over 30% upside). Marvell is positioned to benefit from AI infrastructure demand, with top-line up more than 40% last year and analysts expecting continued double-digit growth.

Analysis

The most interesting setup is not the obvious AI beta trade; it is the dispersion trade inside AI infrastructure. MRVL benefits if hyperscalers keep shifting from one-size-fits-all accelerators toward custom silicon and adjacent networking, but that also makes revenue lumpier and more customer-concentrated than the market usually prices in. The upside is a rerating if design-win conversion shows up in backlog over the next 1-2 quarters; the falsifier is any sign that custom chip adoption is still pilot-heavy while NVDA continues to own incremental spend.

NOW looks less vulnerable than the market implies because enterprise automation is increasingly a workflow budget item, not a discretionary software spend. The real risk is not AI “replacing” the product, but slower net-new adds if CIOs defer projects and if modules become easier to unbundle in procurement. Over 1-3 months, the key catalyst is whether renewal and billings remain resilient; over 6-18 months, sustained mid-90s retention should support multiple recovery even if growth moderates.

PLUG is the weak link: this is a capital-structure story masquerading as a technology story. Incremental revenue quality can improve, but the path to profitability is long enough that dilution, covenant pressure, or project delays can overwhelm operating progress. The market may be underpricing how often alternative power solutions win first in data centers and industrial use cases because they are easier to finance and deploy than hydrogen; if cash burn does not tighten materially in the next 2-3 quarters, the equity remains a financing overhang.

Key Decisions for Investors

• Go long MRVL on pullbacks over the next 2-4 weeks; target 6-12 month hold if hyperscaler/custom-chip commentary confirms design-win conversion. Risk/reward improves if relative performance to NVDA starts widening on earnings revisions.
• Pair long MRVL / short NVDA as a relative-value trade for 1-3 months. Thesis: the market over-owns turnkey GPU exposure while underweighting custom silicon content; invalidate if NVDA reaccelerates or MRVL guide cuts imply slower conversion.
• Buy NOW on weakness or use Jan-2026 call spreads for a 6-12 month recovery trade. Falsifier: renewal/remaining performance obligations slip meaningfully below trend or management signals slower seat/module expansion.
• Avoid outright long PLUG; if anything, use tactical put spreads into rallies over the next 1-3 months. The trade only works if cash burn narrows enough to remove dilution risk; otherwise any operating improvement is likely to be diluted away.