JumpCloud Put the Company Directory in the Cloud. Now It Wants the Whole IT Stack.

The directory is the part of corporate technology that nobody puts in a recruitment video. It is a ledger: Micha works here; Micha owns this laptop; Micha may enter Salesforce but not the payroll system; Micha left on Friday, so every door should close. Boring, until one entry is wrong. Then it becomes a security incident, a stranded new hire, or a former employee with a working password.

JumpCloud's bet is that this ledger should live in the cloud and command much more than logins. From one console, an IT team can manage employee identities, Windows, macOS, Linux and Android devices, single sign-on, multi-factor authentication, Cloud LDAP, Cloud RADIUS, patches, remote access, SaaS applications and the rituals of joining or leaving a company. In 2026, the company widened the cast again: AI agents and other non-human identities now get their own access rules.

That sounds like a feature list because the target is a feature pile. A growing company may have a Microsoft directory, a separate Apple device manager, an SSO vendor, an MFA service, a patching product and one heroic spreadsheet of SaaS seats. JumpCloud is not merely selling another item for that pile. It is selling subtraction.

The first product was too small

The origin starts earlier than the company. Around 2007 or 2008, Rajat Bhargava and Larry Middle were running the security company StillSecure when they noticed something odd: the directory market behaved like a monopoly. They played with a next-generation alternative and could not make the idea work. The conditions were not ready. They put it away.

By 2013, the office had escaped the office. AWS was normal, Google Workspace was rising, Macs and Linux machines were sitting beside Windows, and employees expected to work from anywhere. Bhargava, Middle and product leader Greg Keller returned to the directory idea, but JumpCloud initially launched around automated server management. That was the first framing to fail - not because servers were irrelevant, but because access to servers was one symptom of a broader identity problem.

Customer conversations changed their minds. The useful object was not the server. It was the person and the rules connecting that person to every resource. JumpCloud introduced Directory-as-a-Service in late 2014 and began selling it in 2015. The company had moved one layer down the stack, where a directory could become infrastructure.

“We couldn't figure it out at the time, but came back to the idea in 2013.”Larry Middle, co-founder, on shelving the concept years earlier

A control plane, drawn in widening circles

JumpCloud still begins with a directory: one source for users, groups and permissions. But each expansion follows an identity outward. If the system knows who a worker is, it can decide which apps they see through SSO. If it knows their device, it can require encryption, patch the operating system or block access from an unhealthy machine. If it can see SaaS accounts, it can expose shadow IT, reclaim unused licenses and remove access during offboarding.

Acquisitions filled important gaps. MYKI brought password-manager and authenticator technology in 2022. Resmo added discovery and security for SaaS, cloud and IT assets in 2024. Stack Identity arrived in 2025 with deeper access visibility, least-privilege controls and identity-threat expertise. Deal terms were not disclosed. In each case, JumpCloud bought another way to answer the same question: who or what has access to which resource right now?

The 2026 product stream makes the strategy unusually literal. A visual Workflows canvas automates repetitive identity and device tasks. Direct connectors reveal usage and estimated costs for AI services such as Claude, Cursor, Gemini Enterprise and GitHub Copilot. An AI Gateway, launched in public preview, validates and logs interactions traveling through the Model Context Protocol. The old employee directory is learning to issue hall passes to software agents.

What it costs - and what the price is really buying

JumpCloud is a per-user SaaS business sold through self-service, direct sales and managed service providers. In August 2026, public annual pricing was $9 per user each month for device management, $11 for SSO and $13 for device identity management. Paying monthly lifted those figures to $11, $13 and $15. The broader Platform, Platform Essentials and Platform Prime packages required a conversation with sales. A 30-day trial offered the full platform.

The arithmetic only works against the system it replaces. JumpCloud-sponsored economic modeling found 40 to 60 percent lower costs versus traditional identity and access setups in modeled scenarios, plus 10 to 20 percent lower IT administrative cost. Treat those as vendor-backed estimates, not a law of nature. The case studies are more tactile. Tamr reported cutting routine account administration from hours each week to minutes. Augeo FI said its old Windows Server infrastructure had cost about $100,000 annually, while password support fell from roughly 100 tickets every two weeks to barely one a month.

The reader can copy the mechanism without buying the product: inventory every identity tool, count the administrator hours, map each onboarding and offboarding handoff, and price the servers and forgotten licenses. Consolidation should win on the whole workflow, not on a prettier per-seat number.

The catch: one console is not always the deepest console

JumpCloud fits best where the environment is mixed and the IT team is small relative to the company: a Mac designer, a Windows accountant, a Linux engineer, Google Workspace for collaboration, Microsoft 365 somewhere in finance, and workers scattered across time zones. It also suits managed service providers that repeat the same controls across many clients. Google has become a strategic partner, pairing Workspace with JumpCloud's identity and device layer.

It is less obvious in a company already pulled tightly into Microsoft's orbit. Entra ID and Intune may arrive through existing licenses, integrate more naturally with Windows, and eliminate the need for another vendor. A heavily Apple fleet may want the specialist depth of Jamf or Kandji. A large regulated organization demanding on-premises control may reject the cloud-first premise. And any buyer expecting every module to match a best-of-breed point product will eventually find a sharp edge.

Security companies are judged on the bad day

JumpCloud had one in 2023. A North Korean state-backed actor breached its systems and used data injection in the commands framework to target a small set of customers. The company said fewer than five customers and fewer than ten devices were affected, out of more than 200,000 organizations then using the platform. JumpCloud engaged CrowdStrike and law enforcement, published indicators of compromise and force-rotated every administrator API key.

That response matters, but so does the concentration risk. A platform entrusted with identity, devices and commands is an attractive route into a customer. Consolidation removes handoffs and configuration gaps; it also raises the consequence of a platform failure. Buyers should test emergency access, audit logs, key rotation, device isolation and the ability to keep working during an outage. JumpCloud holds SOC 2 Type 2 and ISO 27001 credentials, but procurement should treat certifications as the start of diligence.

The directory becomes the product map

JumpCloud says more than 250,000 organizations across over 160 countries have chosen it. The roster spans startups, schools, global companies and MSP clients; public names include Canva, CoreWeave, MiQ Digital, AkitaBox, HAAS Alert and TP. The company raised a $225 million Series F in 2021 at a $2.625 billion valuation, bringing total capital above $400 million. That money financed a move from useful directory service to broad IT platform.

The risk is familiar: breadth can turn clarity into clutter. The opportunity is equally plain. Modern work creates new identities faster than IT teams can govern them - employees, contractors, service accounts, SaaS logins and now agents. JumpCloud does not need every module to be the category winner if the connected system makes a lean team faster and safer.

The company began by asking how to automate a server. It grew by asking the more durable question underneath: who should be allowed to do what, from which machine, and for how long? The answer keeps getting larger. So does the directory.