Okta: Like Palantir Two Years Ago.
Deep Dive
Edited by Brian Birnbaum.
The proliferation of AI agents will likely unlock exponential growth for Okta. Okta is positioned to become a foundational platform of the AI economy.
I’ve found that Okta wields capabilities at least somewhat akin to the Nvidia Algorithm. As such, a sudden technological shift presents non-zero odds of exponentiating the company’s earning power. Okta has been focused on identity management for 17 years. Now that AI agents are set to empower the workforce by orders of magnitude, Okta’s revenue per customer may increase non-linearly while costs remain relatively flat. Trading at a P/S multiple of 5.5X, the market is not discounting this possibility.
Okta creates software that helps companies manage the identities of their employees and customers in the cloud: this is an increasingly vital endeavour as everything moves to the cloud, because every single system depends on identity management. Identity management is a forklift upgrade–it usually requires modifying a company’s entire IT system–which has held Okta back from exponential growth. Companies are slow to change and thus acquiring new customers has been difficult.
However, as AI continues to proliferate, AI agents will replace a large percentage of a company’s workforce and clients, with two downstream effects:
Multiplying the number of identities that existing Okta customers have to manage.
Making Okta’s product even more critical, since managing the identity of AI agents requires an AI-powered tool.
After nearly two decades in the identity management business, the company only launched Okta AI in Q3 FY2024. Okta AI works much like Crowdstrike’s XDR engine. Okta AI constantly generating data to train AI models that detect threats. Upon detection of potentially malicious activity, it simply logs users out automatically.
Typical to the moats built around today’s best tech companies, every collected byte of data accelerates the improvement of Okta’s AI models. This in turn drives higher customer retention and more upsells, which yields more data, better models, and so forth.
Crowdstrike is a perfect example of how an ever-improving AI-engine allows novel features to be created and sold at a marginal cost, further increasing operating leverage. This has been the main driver behind Crowdstrike’s rapidly rising cash from operations:
Over time, the variety of problems associated with AI agents will also likely grow exponentially. This will not only further increase Okta’s revenue per customer, but enhance its proprietary data advantage. An exponentially growing problem set on an exponentially growing base of AI agents allows incremental advantages to translate into nonlinear advantages over the competition’s AI models.
The Nvidia Algorithm is an operational pattern that explains some of the most astounding corporate successes of this decade, including not only Nvidia but also companies like Palantir. When a company invests decades into developing a technology that significantly enhances humanity's ability to process information or harness energy, a sudden technological breakthrough can trigger exponential growth that seems to emerge unexpectedly. In reality, these companies set the agenda years prior to their respective inflection points.
The core force behind this growth is the relentless increase in demand for computational power. Regardless of the specific label we use—whether AI, data science, or something else—the need for computation continues to rise at an exponential rate. The flowchart below illustrates a highly simplified version of the Nvidia Algorithm.
Whether Okta’s earning power increases exponentially from here is a function of the strength of the two dependent variables depicted below. My impression is that Okta has the best cloud identity management platform and that the rise of AI agents is highly likely. In Q3 FY2025, half of Okta’s top ten deals came from US federal customers, which somewhat reminds me of Palantir a few years ago. When I learned that Palantir’s software was running the US nuclear stock pile, I understood their digital twin software was second to none.
I haven’t seen quite a strong signal for Okta at present, but their traction with the US government is likely indicative of a strong competitive advantage. In the Q1 FY2023 earnings call an analyst asked CEO and co-founder Todd McKinnon how they were planning to compete with Microsoft’s recently launched Entra. The response is typical of founder-led companies that are highly focused on a single task and end up out-competing much larger players, with abundant resources. McKinnon essentially pointed out that Microsoft didn’t have the adequate level of focus on cloud identity management:
The competitive environment we're seeing really hasn't changed in a number of years. I know we seem to get this question every time we chat and we kind of give the same answer.
And I think when we see competition, we see point players that don't have the breadth and scale or we see the big platforms. And Microsoft, as you mentioned, it probably has the most focus of the big platforms. They [have been] traditionally in the legacy world [and they] had the best identity franchise with Active Directory on-premise. And so they've tried to use that to force their way into cloud identity.
As AI agents proliferate and the fractality of the digital space continues to increase, cloud identity management is set to become one of the foundational layers of the economy. Identity management is the top of the funnel for any digital system on Earth, which makes it particularly attractive for large corporations as a customer lock-in mechanism. Microsoft’s approach is to lock customers in as much as possible, while Okta’s approach is to stay neutral.
Long term, I believe that Okta’s exclusive focus on cloud identity management coupled with the neutral stance will be a substantial competitive advantage. I do not yet have a conviction on Okta, but I plan to study the company in depth quarterly going forward. It looks a bit like Palantir a few years ago, and the odds of a powerful investment thesis arising are meaningful.
Okta’s enhanced profitability despite the macroeconomic headwinds is indicative of process power, which makes exponential growth more likely than otherwise.
Okta’s financials have softened post-pandemic as a result of Okta customers not hiring as many new employees. Reading the quarterly earnings call transcripts of FY2022 (natural year 2021) reveals that much of Okta’s growth at the time came from seat expansions per customer. See the remarks of CFO Brett Tighe during the Q1 FY2023 earnings call:
Our dollar-based net retention rate for the trailing 12-month period remains strong at 123%.
This was driven by the strong upsell motion we are seeing with our existing customers across both Okta and Auth0 as they expand on both products and users.
As a result, dollar-based net retention rate for the twelve trailing months has declined since Q1 FY2023, coming in at 108% in Q3 FY2025. Although not the best news, it makes sense in an environment where companies have been firing employees to get leaner. Further, in my research I have not come across signs of structural degradation: the decline in the retention rate seems to be driven by less seat expansion.
Additionally, in Q3 FY2025, 15% of bookings came from new products that Okta has been releasing during this downturn. Management shared this metric in the earnings call with a celebratory tone, later clarifying that the number was “higher” than in previous quarters. A further increase in this metric over the coming quarters would serve as tangible evidence that there’s nothing structurally wrong with Okta and that the lower retention rates are cyclical.
This will be the most important metric to watch in the short to mid term. Here’s what CFO Brett Tighe said about how they expect the percentage of bookings from new products to evolve:
I would actually just add the percentage is up year-over-year because basically, last year was just governance. And Todd just listed out like 6 products that are doing really well right out of the gate.
So it's pretty easy to see that the percentage is going up and to the right, which is really exciting for us for all the reasons that Todd has said just in his last answer and what he said earlier on the prepared remarks.
Despite this downturn, Okta has also managed to meaningfully enhance its profitability over the past two years via efficiency initiatives. As you can see in the graph below, gross margin (orange line) is now higher than it was before Q1 FY2023. In turn, cash from operations (blue line) has skyrocketed from nearly zero in natural year 2022 to over $600M in the twelve trailing months. That’s approximately 23% of total revenue during the same period. Fairly impressive.
Should the % of bookings coming from new products evolve as management indicated in the last earnings call, much of that revenue should be highly accretive to the bottom line. Indeed, the last two years have put Okta in a position to reap even more financial rewards from the rise of AI agents than otherwise
For reference, Spotify has converted approximately 12% of its revenue to cash from operations in the twelve trailing months. Spotify wields extraordinary process power, as evidenced by a recent increase in profitability in their own right. Okta’s numbers currently point to a company with considerable process power as well, greatly increasing the probability that Okta will capitalize on the AI agent opportunity ahead.
As evidenced by my deep dives on Netflix, Costco, Amazon, Spotify and Palantir, companies with extraordinary process power tend to successfully increase shareholder value over time, far beyond expectations and oftentimes against all odds. This is one of the main things I teach students of my Tech Stock Goldmine course and is one of the most powerful components of the mental framework behind my successful multibagger investments.
I believe Okta’s relatively timid valuation is the result of the market not being sure about the aforementioned cyclicality, together with the (necessarily temporary) focus on net income versus cash from operations. As you can see in the graph below, net income is still in the negative territory despite inflecting upwards since late 2021.
As I have explained previously, a company that has a negative net income but produces positive cash from operations does not actually lose money. The loss often comes from stock-based compensation, which isn’t a cash expense. While it’s dilutive to shareholders, SBC is typically optimal during the growth stage for a couple of reasons:
The stock is typically priced for growth–i.e. above current intrinsic value–meaning the issuance of shares for growth purposes likely increases shareholder value.
Cash is like oxygen for young growth companies. Share-based compensation allows the company to accrue and allocate more cash toward growth, which also increases shareholder value.
As you can see in the graph above, revenue (orange line) has grown tremendously fast since 2017. As previously explained, Okta has two main operations: employee and customer identity. Okta only launched the latter in January 2024 (Q1 FY2025) and just two quarters later, the customer business is at a $1B annual revenue run rate. This suggests Okta’s ability to deploy new verticals and upsell its customers is notable, which in turn is also suggestive of extraordinary process power.
Improved financial performance is not only the result of cutting costs, but simultaneously delivering more value to customers. The ability to keep an eye toward growth of additional verticals during sector-wide challenges points to competent management.
The recent security incident Okta experienced in October 2023 enabled me to see how co-founder and CEO Todd McKinnon thinks about managing the company. Okta's October security incident involved unauthorized access to its customer support system through a compromised service account, affecting 134 customers and leading to potential session token hijacking. Okta started out facilitating customers’ moves to the cloud. Over time, it’s had to evolve into a cybersecurity company as they and their customers have been increasingly attacked by cybercriminals.
After the recent incident, which according to management has had no tangible impact on the financials, McKinnon came up with a plan called Bedrock to ensure Okta becomes one of the most secure companies in the world, a bottom-up approach that enables the best ideas to make their way into the products and a culture that makes security the number one priority for everyone in the company.
The latter component proves McKinnon understands the importance of culture in the pursuit of maximizing corporate value over time. As you may know by now, the culture of a company broadly determines its fate:
When you think about executing well, it can be pretty straightforward. It's like you have to make sure you have a clear vision. We want to be one of the most secure companies in the world. We want–you've got to set clear priority and get the right amount of resources on it. And that kind of sets up this cultural component to be successful.
Okta is a highly asymmetric thesis in the making.
Identity management is a crucial component of the AI-driven economy. Okta has been cutting its teeth for nearly two decades now and the recent efficiency gains have poised the company to meaningfully capitalize on the rise of AI agents. It looks like Okta’s earning power will increase non-linearly in the coming years, with no obvious limit in sight.
Their focus on identity management sustained over two decades is already an appealing barrier to entry. However, as they move down the path of managing AI agents, their platform promises to yield a proprietary data moat/flywheel similar to that of Crowdstrike’s. The key enablers of Crowdstrike’s data advantage have been a lighter agent and a quicker time to value every quarter.
Per his remarks in the Q3 FY2025 earnings call, McKinnon perfectly understand that this is the path to success:
And I think the governance market is like that. There's a lot of software, a lot of the stuff is not implemented. In our product, which is very strongly integrated to access management or access management, it's very quick to implement.
People get tremendous value out of it fast.
It's well integrated to many SaaS applications. And more and more, it's integrated to even on-prem applications as we innovate there. And it's fast time to value. And I think that, in this case, is the winning formula, and we're seeing it play out in the market.
I will continue tracking Okta over time to more deeply understand the competitive positioning of its offerings and its ability to confer the aforementioned data advantage, via its recently launched Okta AI module. However, in my view the company’s execution in the past two years points to notable process power, which makes exponential earning power growth more likely than not as AI continues to progress.
Until next time!
⚡ If you enjoyed the post, please feel free to share with friends, drop a like and leave me a comment.
You can also reach me at:
Twitter: @alc2022
LinkedIn: antoniolinaresc