Signals Inbox·August 21, 2026·Cybersecurity

Is Horizon3 really worth $2B today?

Horizon3’s $2B valuation looks aggressive but justified today: near-$100M ARR, 120% growth and 310,000 production security tests support it, while Pentera and public-market multiples show how quickly the case weakens if growth cools.

Q1What exactly happened to Horizon3’s valuation?

Horizon3’s move above $2 billion is a major repricing: investors valued the company at $650 million just over a year earlier.

On August 3, 2026, Horizon3 raised a $250 million Series E at a valuation above $2 billion. NightDragon and NEA co-led the oversubscribed round, with seven new investors and five returning investors participating. The previous Series D, announced in June 2025, raised $100 million and was later described by Horizon3 as carrying a $650 million valuation.

That means the valuation rose a little over 3x in roughly 14 months. Horizon3 was founded in late 2019, so it reached the $2 billion mark in around seven years. That is fast even for cybersecurity, although recent companies such as Wiz, Cyera and XBOW have shown that security startups can reprice very quickly when revenue and demand move with them.

What happened between the two rounds is the real story. Horizon3 went from a company still proving autonomous pentesting could become a large software category to one approaching nine-figure recurring revenue and expanding internationally. The latest round is therefore a bet on a business that has already scaled materially, rather than a valuation attached to an early product demo.

Horizon3 funding and valuation progression

Q2Did Horizon3 reach a $2B valuation unusually fast?

Seven years is fast for reaching a $2 billion valuation, although cybersecurity has produced several companies that moved even faster.

Horizon3 was founded in late 2019. Pentera launched its automated security-validation platform around the same broad period and became a unicorn in 2022. Wiz, founded in 2020, grew so quickly that Google ultimately completed its acquisition for about $29.5 billion roughly six years after the company was created.

The newer generation is moving even faster. XBOW was founded in 2024 and reached a valuation above $1 billion around two years later as investors chased autonomous offensive security. Cyera, founded in 2021, has already reached a $12 billion valuation.

So Horizon3’s speed is unusual by normal enterprise-software standards, but fairly believable inside today’s cybersecurity market. What matters more is whether Horizon3 has enough revenue behind the valuation. Recent cyber history is full of very fast valuation jumps.

Q3Is Horizon3 really close to $100M in ARR?

Yes, the near-$100 million ARR figure is credible enough to use as our main revenue anchor, although we should treat it as an approximate historical figure rather than Horizon3’s exact ARR today.

TechCrunch reported during the latest financing that Horizon3 chief revenue officer Matt Hartley said the company approached $100 million in annual recurring revenue in 2025. Since the number came directly from the executive running revenue, we give it considerably more weight than an anonymous estimate.

There is one discrepancy worth keeping in mind. TechCrunch had reported around $30 million of ARR in May 2025, citing a source familiar with the fundraising. Reaching close to $100 million shortly afterward would imply an enormous jump. The two figures probably refer to different points in time, different definitions, or an earlier estimate that understated the business.

For valuation work, we would therefore use roughly $100 million as the latest credible absolute ARR benchmark. We would not present $100 million as Horizon3’s current ARR. The company has continued growing since the period that figure covers, so the real run rate may already be higher.

Q4Is Horizon3’s roughly 20x ARR valuation actually expensive?

Around 20x the latest disclosed ARR is expensive, but Horizon3’s growth makes that multiple much easier to defend than the headline suggests.

Dividing a $2 billion valuation by roughly $100 million of ARR gives us about 20x. Because the valuation is officially above $2 billion, the true numerator is slightly higher. At the same time, the $100 million denominator comes from an earlier period, so using it probably overstates Horizon3’s current multiple.

The repricing gives us another angle. The valuation rose from $650 million to more than $2 billion, slightly above 3x. Horizon3 currently reports 120% year-over-year ARR growth, which means revenue roughly doubled over a comparable one-year interval. Even allowing for imperfect timing, investors have clearly become willing to pay more for each dollar of Horizon3 revenue.

That extra premium is where the valuation debate really sits. A 20x multiple would be hard to justify if Horizon3 were already slowing toward ordinary cybersecurity growth. With triple-digit growth still being reported, the multiple looks aggressive rather than absurd.

Q5How expensive is Horizon3 compared with public cybersecurity companies today?

Horizon3 currently sits between mature security vendors and the market’s most highly valued cyber platforms, which is roughly where we would expect a private company growing this quickly to trade.

Using the latest available market capitalizations and trailing revenue, Tenable trades around 4x sales, SentinelOne around 8x and Qualys around 9x. Their recent revenue growth is approximately 10%, 21% and 10%, respectively. Horizon3’s roughly 20x benchmark is obviously much richer.

The other end of the market looks very different. Palo Alto Networks is around 30x trailing revenue and CrowdStrike around 43x. Their latest trailing growth rates are around 20% and 23%. Public investors are currently paying enormous premiums for cyber companies they believe can dominate large platform categories.

Horizon3 is still a much smaller and riskier company, so we would never apply CrowdStrike’s multiple mechanically. Still, a company reporting growth several times faster than those public vendors can reasonably trade above Tenable or Qualys. The real danger would come if Horizon3’s growth fell toward 20% while its valuation continued assuming premium-platform economics.

Horizon3 valuation versus public cybersecurity companies

Q6Is Horizon3 more expensive than Pentera?

Yes. Pentera is the comparison that keeps the Horizon3 bull case honest because the two companies sell unusually similar forms of automated offensive security.

Pentera raised $60 million in 2025 at a valuation above $1 billion. It then announced that it had crossed $100 million in ARR, with more than 1,200 enterprises using the platform. Its last disclosed valuation therefore sits around half Horizon3’s latest mark despite revenue of the same broad order of magnitude.

That comparison is imperfect because Pentera’s valuation is older. A new financing today could easily price the company higher. Still, we cannot simply ignore the gap.

Horizon3 has a credible reason to command a premium. Its recent ARR growth is much faster, while Pentera said its revenue increased more than 300% over four years. Horizon3 is also pushing quickly from pentesting into broader exposure validation, remediation and web applications. Yet even after giving Horizon3 credit for those advantages, Pentera suggests that $2 billion already prices in a meaningful amount of future execution.

Q7Do Cyera and Wiz make Horizon3’s valuation look less extreme?

Yes, recent cybersecurity deals show that investors are willing to pay far more than 20x revenue when they believe a company controls a strategically important new security layer.

Cyera recently completed a $600 million financing at a $12 billion valuation. TechCrunch reported that the company had surpassed $150 million in ARR, although Cyera disputed the accuracy of the financial numbers reported around the round. If the estimate were close, the implied multiple would be around 80x.

Google provides a cleaner example of strategic scarcity. It completed the acquisition of Wiz for approximately $29.5 billion. Wiz had already built one of the fastest-growing cloud-security businesses ever, so Google was paying for a company with a much broader platform, customer base and strategic position than Horizon3 currently has.

These comparisons do not prove Horizon3 is cheap. They show that 20x sits comfortably inside the range investors are currently willing to entertain for high-growth cybersecurity. Horizon3 still has to earn its place in that premium group.

Q8Is Horizon3 still growing unusually fast right now?

Yes, Horizon3 is still growing at an exceptional rate, and the latest numbers suggest the business has not yet hit the usual slowdown that comes with getting larger.

In March, Horizon3 reported 102% year-over-year ARR growth and more than 5,200 organizations using NodeZero. By the latest financing, the company reported 120% ARR growth and more than 7,000 organizations.

That customer increase is worth isolating. Going from 5,200 to more than 7,000 represents at least 35% growth in less than five months. Revenue growth also accelerated from the 102% rate reported earlier to 120% in the latest announcement.

Looking further back gives us another check. Horizon3 reported 2,962% three-year revenue growth for the 2025 Inc. 5000 and later cited 19,939% three-year growth for Deloitte’s Technology Fast 500 measurement. The exact periods differ, but both independently show that the recent triple-digit growth is part of a multi-year expansion rather than one unusually good quarter.

For a company already around the $100 million ARR scale, that pace is the main reason the valuation works.

Q9Are Horizon3’s 7,000 customers really 7,000 enterprise buyers?

No, we should not read Horizon3’s 7,000-plus organizations as 7,000 direct enterprise contracts, because a large share of the business reaches customers through MSSPs and other partners.

Horizon3 said earlier this year that approximately 70% of customers were serviced through managed security service providers. It also reported that 32% of Q4 bookings originated through channel partners. Soon after the Series E, the company announced another $20 million investment in its global partner ecosystem.

That model helps explain how Horizon3 can reach thousands of organizations without building a huge direct-sales force. One MSSP can use NodeZero across multiple client environments, which means an “organization using NodeZero” does not necessarily represent a separate large software contract signed directly with Horizon3.

A crude $100 million divided by 7,000 organizations produces only about $14,000 of ARR per organization. Pentera’s $100 million-plus across more than 1,200 enterprises produces a much higher figure. Those numbers are not directly comparable because the customer definitions and distribution models differ, but they show why raw logo counts can mislead us.

The partner model still looks attractive today. It gives Horizon3 distribution, local relationships and access to smaller customers that would be expensive to acquire directly. We simply put more weight on ARR and retention than on the 7,000 headline.

Q10Are Horizon3 customers actually spending more after they buy?

Yes, Horizon3’s retention data says existing customers are expanding their spending rather than quietly churning after the first pentest.

The company reported 125% net dollar retention and 94% gross dollar retention earlier this year. A 94% gross retention rate means Horizon3 keeps roughly $94 of every $100 of recurring revenue before counting customer expansion. Once upsells and additional usage are included, that same customer base ends up at roughly $125.

That is a healthy combination. Horizon3 can lose some accounts or contract value while still expanding the remaining customer cohort by enough to produce 25% net growth before adding new customers.

That retention profile answers a basic product question. Autonomous pentesting could easily have turned into something companies try once, produce a report from, and then cancel. The retention data points in the opposite direction. Customers appear to increase usage after seeing what NodeZero does.

Q11Is NodeZero being used enough to prove this is a real product?

Yes. NodeZero has now accumulated enough real production usage that Horizon3 has moved well beyond the demo stage.

Horizon3 passed 100,000 autonomous pentests in February 2025. By June 2025, that had risen above 150,000. The latest company figure is 310,000 production security tests.

The acceleration is more interesting than the total. Horizon3 added roughly 160,000 tests after June 2025, meaning it performed more tests during the following 14 months than it had accumulated during its entire earlier history.

The company also says those 310,000 tests were completed without causing production disruptions. That safety claim comes from Horizon3 itself, so we would like to see independent verification eventually. Still, executing hundreds of thousands of authorized attacks inside real environments creates experience that a new competitor cannot reproduce overnight.

The product is also widening lately. NodeZero WebApp Pentesting was introduced in July, extending autonomous testing into web applications and linking application weaknesses with identity, cloud and infrastructure attack paths. That expansion makes Horizon3 look increasingly like a platform rather than a single-purpose pentesting tool.

Q12Is Horizon3 spending efficiently enough for a $2B company?

So far, Horizon3 looks relatively capital-efficient for a company growing this quickly, although private-company disclosure prevents us from calling the economics proven.

Before the latest $250 million raise, Horizon3 had raised about $178.5 million in total funding. It had already built toward the $100 million ARR range. The company also said around its Series D that it was Rule-of-40 positive, meaning its growth and profitability profile together cleared a common SaaS quality threshold.

TechCrunch reported that Horizon3 had spent roughly $100 million on R&D over six years. That is substantial, but modest compared with AI companies that consume similar amounts every few months on infrastructure and model training.

But a few numbers are still missing. Horizon3 does not publicly disclose gross margin, operating margin, free cash flow, customer acquisition cost or payback periods. A $250 million Series E is also a very large injection relative to current revenue.

So we see decent evidence of efficient scaling, but we cannot yet verify that Horizon3 has CrowdStrike-quality economics underneath the growth.

Q13Is Horizon3’s market big enough for the company to become much larger?

Yes, Horizon3 has plenty of room to grow from its current size, although it will need to take market share rather than simply ride category growth.

Fortune Business Insights currently estimates the global penetration-testing market at about $3.1 billion and expects it to reach $7.4 billion by 2034. Grand View Research puts the adjacent continuous threat exposure management market at roughly $3 billion currently, growing to $7 billion by 2033.

Those markets overlap, so adding them together would inflate the opportunity. Horizon3 is also moving into the broader security-testing market through web applications, exposure management, Rapid Response and remediation workflows. Fortune Business Insights estimates that wider security-testing market at more than $23 billion.

Growth is the more revealing comparison. The core penetration-testing market is expected to grow around 12% annually, while Horizon3 has lately been growing many times faster. Market expansion alone therefore cannot explain the company’s trajectory.

Horizon3 needs to convert spending that currently goes toward manual pentesting, vulnerability management, breach-and-attack simulation and exposure management. If customers start testing weekly or monthly instead of hiring humans once a year, the effective market expands even without cybersecurity budgets growing dramatically.

Q14Is AI actually increasing demand for autonomous pentesting now?

Yes, AI is creating a real reason for companies to test security more frequently, and the evidence goes beyond cybersecurity vendors telling us that AI is scary.

The World Economic Forum’s latest Global Cybersecurity Outlook found that 87% of respondents saw AI-related vulnerabilities as the fastest-growing cyber risk during the previous year. The share of organizations with processes for assessing AI security rose from 37% to 64%. Another 94% expect AI to be the biggest force changing cybersecurity this year.

Horizon3 is benefiting from a broader operational problem. Companies are changing applications, infrastructure and AI systems faster, while attackers can discover and weaponize weaknesses faster as well. An annual penetration test becomes less useful when the environment has materially changed several times before the next test arrives.

Horizon3 is already adapting its product around that shift. Recent releases added web-application testing, larger-scale external-asset testing, expanded Rapid Response coverage for newly disclosed vulnerabilities and stronger workflows for verifying remediation.

That combination makes the demand story credible today. The opportunity comes from increasing the frequency and scope of security validation, which is a much stronger thesis than simply putting “AI” in front of pentesting.

Q15What can Horizon3 do that competitors cannot easily copy?

Horizon3’s best moat today comes from years of production attack data and the trust required to let software autonomously attack live systems.

The underlying AI models are unlikely to remain proprietary enough to protect the company forever. Competitors can access strong models, hire offensive-security researchers and build autonomous agents of their own.

The harder part is getting a bank, government agency, hospital or large industrial company to let those agents run real exploits in production. Horizon3 has spent years accumulating a record across hundreds of thousands of tests. It has also obtained FedRAMP High authorization and says NodeZero is used by four Fortune 10 companies and sensitive government environments.

Every production test potentially improves Horizon3’s understanding of how credentials, configurations, identities and infrastructure combine into actual attack paths. The recently released WebApp capability also connects application attacks with cloud, identity and host compromise, widening the data the system can learn from.

So Horizon3’s defensibility comes from operating history, trust and attack-path data. Those advantages can compound if usage keeps accelerating.

Q16Can Tenable or CrowdStrike make NodeZero less valuable?

Yes, larger cybersecurity platforms can squeeze Horizon3, especially if autonomous validation becomes a feature customers expect inside broader security suites.

Tenable already combines exposure management with attack-path analysis, vulnerability prioritization and security-control validation. CrowdStrike has expanded its own exposure-management products and increasingly connects endpoint telemetry, identities, vulnerabilities and cloud risks inside Falcon.

Horizon3 currently goes deeper into autonomous offensive execution. NodeZero actually attempts attack chains in production, rather than mainly calculating which weaknesses look exploitable from telemetry and vulnerability data.

The gap can still shrink. CrowdStrike, Palo Alto, Tenable, Microsoft and other large vendors already sit inside thousands of enterprise environments. They can bundle adjacent capabilities, discount them and make customers question why another security platform is necessary.

Horizon3 therefore has to keep expanding faster than the incumbents can copy the useful parts. The recent move into web applications, Rapid Response, remediation verification and broader proactive security suggests management understands that problem.

Q17How much revenue does Horizon3 need to grow into a $2B valuation?

Horizon3 only needs roughly $200 million of ARR to support a $2 billion valuation at a much more ordinary 10x revenue multiple.

Starting from the latest credible benchmark of approximately $100 million, that means doubling once. If Horizon3’s current triple-digit growth slowed dramatically to 60%, the same base would still reach about $160 million after one year. At 80% growth, it would reach $180 million.

Timing is the whole game here. Horizon3 can grow into the valuation fairly quickly as long as the business stays fast enough to outrun multiple compression.

At $200 million ARR, even a 10x multiple gets us back to $2 billion. At $250 million, only 8x is required. The valuation starts becoming dangerous if revenue growth slows before Horizon3 reaches those levels.

ARR needed to support a $2B valuation

Q18What would make Horizon3’s $2B valuation look cheap?

The $2 billion valuation starts to look cheap if Horizon3 can keep growing above roughly 60% while turning NodeZero into a broader proactive-security platform.

Imagine the company starts from approximately $100 million ARR and grows 80%. That produces about $180 million. At a 15x multiple, the company would be worth around $2.7 billion. At 20x, it would be worth $3.6 billion.

The product roadmap is important here. Horizon3 is currently moving beyond internal pentesting into web applications, external assets, newly disclosed vulnerability validation and remediation. It has also opened an EMEA headquarters, is entering Singapore and Australia, and has committed another $20 million to partner expansion.

If those moves turn NodeZero into infrastructure that customers run continuously across several security workflows, Horizon3 could keep both high growth and strong expansion revenue for longer than a single-product pentesting company normally would.

That is the bull case we find credible. It requires continued execution, but it does not require fantasy-level revenue assumptions.

Q19What would make Horizon3’s $2B valuation fall apart?

The valuation gets difficult very quickly if Horizon3’s growth drops toward ordinary cybersecurity levels before ARR reaches roughly $200 million.

Consider the public comparisons. Tenable currently trades around 4x revenue while growing around 10%. Qualys trades around 9x with similar growth. SentinelOne is around 8x while growing just above 20%.

If Horizon3 eventually grew 20% to 30% and the market valued it at 8x to 10x sales, $100 million of ARR would support only around $800 million to $1 billion. Even at $150 million, a 10x multiple gives us $1.5 billion.

The next growth figures matter more than another customer-count announcement. Net retention also needs to stay comfortably above 110%, and Horizon3 must keep widening the product before competitors make autonomous validation easier to bundle elsewhere.

The bear case is quite specific: Horizon3 becomes a good cybersecurity company, but the market stops seeing it as a future platform. At that point, $2 billion becomes hard to defend.

Q20Is Horizon3 really worth $2B today?

Our answer is yes: Horizon3’s $2 billion valuation looks aggressive but justified today.

The evidence is stronger than the headline initially suggests. The business approached $100 million of ARR, retention is healthy, usage has accelerated sharply, and product expansion has continued recently into web applications and broader security validation. Public cyber valuations also show that premium multiples remain available for companies the market believes can become strategic platforms.

There are two reasons we would stop short of calling Horizon3 cheap. Pentera has built a directly comparable business above $100 million ARR at a much lower last disclosed valuation, and Horizon3 still lacks the financial disclosure that would let us verify margins and cash efficiency properly.

As we saw above, the recent 120% ARR growth is doing most of the heavy lifting in the valuation. If Horizon3 can stay above roughly 60% to 80% growth long enough to reach $150 million to $200 million of ARR, $2 billion should become increasingly easy to defend. If growth suddenly falls toward 30%, the valuation can compress brutally even if the company remains healthy.

So yes, we think Horizon3 is worth around $2 billion today. Investors are paying ahead for several more years of category leadership, but the current revenue, customer expansion, retention, production usage and product velocity give them enough evidence to make that bet.