Z
Zscaler, Inc.
ZS · Technology
Q2 FY2026
Earning ended Jan 31, 2026Report generated Mar 10, 2026
The Tollbooth That Microsoft Is Building a Road Around
The bull case on AI tailwinds is obvious and priced — what the market systematically underweights is that Microsoft does not need to beat Zscaler on features, only on procurement inertia, and it already has a structural advantage in every enterprise that runs Azure AD, Teams, and Defender under a single EA negotiation.
Flowly Score
Apr 16 Close
+67%$134.33
Fair Value
$225.00
62
Averageat $134.33
Business Strength7/10
The switching cost moat is structurally real — enterprises who demolished their VPN infrastructure cannot easily rebuild it — but counter-positioning has meaningfully eroded as Palo Alto absorbs the self-cannibalization pain to compete. Gross margin stability across five years of intensifying competition is the single most honest signal that the moat is holding.
Financial Resilience7/10
The GAAP loss story is almost entirely a non-cash SBC artifact — the underlying cash generation is genuine, with FCF roughly quintupling over five years as operating leverage asserts itself. The balance sheet evolution deserves a watchful eye: cash has dropped sharply while debt has grown meaningfully, a shift from the fortress position of prior years.
Growth & Trajectory8/10
The AI security pillar hitting its full-year ARR target three quarters early is not a marketing talking point — it reflects genuine product-market fit with enterprises navigating a new attack surface they don't yet know how to secure. RPO growing faster than revenue is the clearest leading indicator that tomorrow's growth is already locked in.
Valuation6/10at $162.62
The stock sits in a narrow band where the neutral DCF scenario implies meaningful upside but the pessimistic scenario implies brutal downside — that asymmetry is unforgiving for a business where a single competitive variable (Microsoft win rates) could rapidly compress growth assumptions. You are paying a fair price for a genuinely good business, which is very different from paying a bargain price.
Risk4/10
The Microsoft threat is not theoretical: Entra Private Access and Entra Internet Access are functional substitutes bundled into licenses enterprises already own, and 'acceptable quality at zero marginal cost' is the most lethal competitive dynamic in software history. Combined with Palo Alto's aggressive platformization discounting, Zscaler faces a squeeze from above via distribution leverage and from the side via price — both already underway.
Zscaler is a high-quality business trading at a price that demands near-perfect execution. The cash generation is real, the switching costs are structurally embedded in destroyed infrastructure rather than contractual lock-in, and the platform expansion into AI security has outpaced even management's optimistic internal targets. The quality of the business earns respect; the valuation offers only a sliver of margin for error, with the neutral DCF scenario implying modest upside and the pessimistic case implying a genuinely painful outcome. The trajectory is genuinely encouraging in ways the income statement obscures. Operating leverage is finally asserting itself after a decade of deliberate market-capture spending — the compressing gap between gross and operating margins is not accounting noise, it is the natural maturation of a platform that has achieved sufficient scale to grow revenue faster than its cost structure. The RPO growth outpacing revenue growth tells you the installed base is deepening, not churning, which is the most important thing to know about a recurring-revenue business facing competitive pressure. The single biggest risk, named precisely: Microsoft's bundling strategy requires zero product improvement to work. Every enterprise security buyer who already pays for E5 licensing faces a choice between paying separately for Zscaler's best-of-breed platform or accepting an integrated capability at no incremental cost. The question is not whether Zscaler is better — it almost certainly is — but whether enterprise IT buyers will pay a meaningful premium for that quality gap when budgets are tight and the CFO is asking why the security stack has six vendors. If that calculus shifts at mid-market scale, the new logo growth assumptions embedded in any optimistic scenario evaporate faster than any DCF model captures.