Average customer downtime less than 26 minutes, including both planned and unplanned, while upgrading 50 times.
![](https://www.socotra.com/wp-content/uploads/2024/06/Reliability-Blog-V4_03.png)
In a first-of-its kind announcement, Socotra, the cloud-based policy administration system (PAS), has shared its system uptime numbers for 2023. Socotra customers on average experienced less than 26 minutes of downtime last year–including both planned and unplanned downtime.
“With customers, producers, and external API-driven distribution channels demanding real-time carrier responsiveness, poor reliability is no longer an option,” said Dan Woods, CEO of Socotra. “System uptime has evolved from a baseline KPI to a critical component of business continuity, customer satisfaction, and profitability.”
The stats:
![computer-icon computer-icon](https://www.socotra.com/wp-content/uploads/elementor/thumbs/computer-icon-qpkgcdunz5p81a93rd2y81pwfh5ued4xvnkvekkq92.png)
Average 2023 uptime across all customers was >99.995%
![](https://www.socotra.com/wp-content/uploads/2024/06/clock-arrow-icon.png)
Average <26 minutes of downtime in 2023, including both planned and unplanned downtime
![](https://www.socotra.com/wp-content/uploads/2024/06/cycle-gear-icon.png)
No Socotra customer had more than 0.03% downtime in 2023
![](https://www.socotra.com/wp-content/uploads/2024/06/graph-arrow-icon.png)
These statistics include combined downtime resulting from the 50 upgrades each Socotra customer did in 2023
![](https://www.socotra.com/wp-content/uploads/2024/06/gears-image-quote.png)
This high degree of reliability and frequent upgrades is usually found only in consumer platforms. In contrast, a recent report showed that 30% of financial and insurance organizations experienced high-impact outages at least once per week. Respondents in the same report shared that those outages generally take more than 30 minutes to resolve.
Socotra also revealed more about how they achieved these numbers:
Reducing Unplanned Downtime
Unplanned downtime is when unforeseen circumstances result in service interruption. Hardware failures, bugs, and human error can all cause unplanned downtime. Furthermore, it occurs more often during times of high load—which greatly increases the business impact–or when key technical staff are unavailable.
Because Socotra customers are always on the latest version with no customization, Socotra automates full testing of 100% of its features, for 100% of its customers. Furthermore, Socotra’s true cloud-native design provides clients with redundancy across availability zones and supports multi-site failover.
Reducing Planned Downtime
Planned downtime is a scheduled period when a system is intentionally taken offline. This typically happens to accommodate upgrades, configuration changes, data transformations, new integrations, customizations, or preventative maintenance. Socotra strategically works to keep planned downtime to a minimum:
![](https://www.socotra.com/wp-content/uploads/2024/06/checkmark-icon.png)
We champion the philosophy of small weekly updates, rather than infrequent, expensive, and disruptive upgrades.
![](https://www.socotra.com/wp-content/uploads/2024/06/checkmark-icon.png)
Socotra API updates are almost always backwards compatible, and those few breaking changes occur only after a long and clearly-communicated deprecation period.
![](https://www.socotra.com/wp-content/uploads/2024/06/checkmark-icon.png)
Our approach is so refined that downtime is rare and, when necessary, lasts only a few minutes—always scheduled during off-peak hours to minimize impact.
To achieve such impressively low planned downtime, Socotra uses a sophisticated method of building a parallel stack over our existing data stores. This allows us to seamlessly switch to the new stack with a single DNS update. During this transition, only new transactions are directed to the updated stack, while in-process transactions continue uninterrupted on the original version. Actions such as quoting, underwriting, and document generation may literally be (and frequently are) in mid-execution at the moment of upgrade—and those actions finish, unaffected.
This is only possible with modern, backwards-compatible storage designs that accommodate updates without disruption to existing applications and data. This storage and upgrade approach is standard for high-reliability platforms such as Gmail, LinkedIn, and Zoom.
![](https://www.socotra.com/wp-content/uploads/2024/06/forward-icon.png)
Looking Forward
Maximizing system uptime is about more than technology—it’s a strategic imperative to safeguard digital assets, ensure continuous service delivery, and sustain customer trust.
Socotra is the first PAS to achieve consumer-grade reliability, because it uses the same techniques, technologies, and talent responsible for such always-available platforms such as Gmail, LinkedIn, and Zoom. We expect more insurance technology providers to follow Socotra’s lead in providing cloud-native platforms that deliver high reliability, high performance and modular microservice design.
The overall software industry knows how to build great cloud-native software, and Socotra is proud to lead the way in bringing these best practices to insurance software.