Okta is being run like Cisco, for low growth, profitability and buybacks, rather than as a high-growth tech company trying to win, after three years of layoffs. Its AI thesis is generic and soft, and beyond customer inertia there is no strong answer for why it wins in five years. Its model rests on switching costs, yet net retention keeps declining, and outsourcing implementation to GSIs gives up the proprietary deployment knowledge that could deepen its moat. x.com 

Okta's valuation is not justifiable at a P/E of 114 with only 10% revenue growth; it is speculation at its finest. x.com 

Okta should capture durable value from agentic AI because every agent touching enterprise data needs authentication, authorization and auditability, making identity infrastructure. Okta is the only independent identity platform with real enterprise distribution, a launched non-human identity product, and early signs of contract-value expansion. It trades far below its historical multiple despite 27% FCF margins, priced as if Microsoft consolidates identity, unlikely in multi-cloud firms. x.com  … Show more

Okta's risk/reward is no longer attractive after doubling in three months, at 10 times FY27 sales and over 50 times earnings. Even if agentic AI products reach half of its 20,000 enterprise customers within two years with a 40% contract-value uplift, implying 13-15% annual sales growth, it is not cheap; a 50-60% uplift and 17-20% growth still isn't. Agentic AI is exponential, so the growth duration may be underestimated, but two-thirds of the position has been sold. x.com 

Okta, Salesforce and Atlassian look insanely expensive at nine times forward revenue against roughly 10% year-over-year growth. Many SaaS firms also have high sales and marketing costs. A commenter flags inference spending, which SaaS investors worry could compress margins; if free cash flow, growth, retention or gross margins weaken, their price-to-sales multiples could collapse. x.com  x.com  x.com  x.com 

Okta's backlog reaccelerated, with booked demand now outrunning recognized revenue. RPO rose 17% to $4.86B and current RPO 14% to $2.59B, both above 11% revenue growth. Identity is the lowest-hanging AI value layer: whoever wins at building AI, every agent will need a trusted identity and controls over what it can access, as Okta's founder says. x.com  x.com 

Okta's valuation is hard to justify at over 100x trailing GAAP earnings for a company growing about 11%. It has $296M of trailing GAAP net income against a $30B market cap, so its bottom line must outgrow revenue for years just to support today's price, which already prices that in. x.com 

Okta looks expensive against its cash earnings at $150.77. The identity company trades at 90 times cash earnings and 77 times next year's, while software companies usually get 30 times. Thirty times next year's cash earnings implies fair value of $58.47–74.42, leaving the stock 103% above the high end. x.com 

Okta and SailPoint should benefit as AI agent traffic explodes, because agents generate vast numbers of requests for tiny files, which favors identity security rather than bandwidth plays. At about 8-9x sales they are not superficially cheap, but with change spanning orders of magnitude and business models that scale well, they could trade around 15x today and return about 300% over time. Cloudflare, by comparison, trades near 40x. The long-avoided theme is about to light up. x.com 

CrowdStrike, Palo Alto Networks, Fortinet and Okta offer terrible risk/reward because their valuations need a lot to go right. The thesis is correct: more AI agents mean more identities, endpoints and attack surfaces, so security spend should keep growing. But CrowdStrike trades at 147x forward earnings, Palo Alto at 79x, Fortinet at 44x and Okta at 41x. These are good, some exceptional, businesses, yet a great business can still be a terrible investment at the wrong price. x.com