Spotinst has spent years collecting data and analyzing spot instance availability and interruption patterns for AWS, Azure and Google Cloud Platform. As enterprises began shifting their focus to containerized applications, the company added a platform-agnostic service to manage scaling of container deployments while taking advantage of spot savings. With $35m of new investment from a round led by Highland Europe, Spotinst is plotting aggressive growth and building out new capabilities, including a product for hands-off serverless container scaling and an on-premises implementation.

The 451 Take

Spotinst's single-minded pursuit of savings based on datacenter overcapacity has taken many forms since its inception three years ago. The company came to market with its flagship Elastigroup service, which uses machine learning to 'score' the likelihood of AWS or Google Cloud Platform spot instances being interrupted, then applies that information to create cost-oriented clusters offering deep discounts with 100% availability. When Microsoft Azure launched low-priority VMs for batch workloads in 2017, Spotinst repurposed them as a mechanism for production-level cloud cost savings; soon thereafter, the company launched a Kubernetes- and Amazon ECS-compatible container service to automatically scale underlying infrastructure while taking advantage of spot resources. With a pay-as-you-save pricing model, Spotinst is motivated to reduce its customers' cloud spending in a way that few companies in the hyperscale ecosystem are, and this has caught the attention of investors and customers, including Sony, IBM, Intel, Samsung and Ticketmaster.


In the three years since its inception, Spotinst has honed its ability to use low-cost ephemeral instances – called spot instances on AWS, preemptible VMs on Google Cloud Platform, low-priority VMs on Azure, preemptible instances on Alibaba Cloud and transient servers on IBM – together with machine learning to construct highly available and economical compute clusters capable of running production workloads. Spot instances are often priced at discounts of 80% or more versus on-demand rates, but they come with a catch: when the provider needs to reclaim the spare capacity it can terminate the instance, in some cases without warning.

Spotinst uses predictive analytics to forecast when excess capacity is likely to be available, then dynamically pools spot resources with on-demand and reserved instances to create resilient, cost-effective compute clusters that are suitable for production environments, which the company claims results in savings of up to 80% versus on-demand rates.

Few enterprises have the stomach to manage spot pricing manually or with the help of provider tools: 451 Research's recent Voice of the Enterprise: Cloud, Hosting & Managed Services, Organizational Dynamics 2018 survey found that only 9% of enterprises were using spot resources to pay for public cloud services, versus 52% paying for on-demand usage billed in arrears. As Spotinst describes it, the opportunity for customers is 'like money on the floor that we're picking up for them.'

The company has raised $52.6m of venture funding in four rounds, including a $35m series B led by Highland Europe in August.



Spotinst takes a customer-centric approach to development, building features, integrations and plug-ins to popular third-party tools based on demonstrated demand. Its longest-running service, Elastigroup, works at the VM level and can manage resources on AWS, Azure or Google Cloud Platform; in 2019 it expects to add support for Alibaba preemptible instances and IBM transient servers. A recent survey by the company found that 70% of its customers are using or plan to use multiple public clouds, with different providers hosting different applications, so Spotinst maintains algorithms wherever spot pricing is available.

Earlier in 2018 Spotinst introduced its Managed Container Service, an enhancement to Elastigroup that hooks into either Amazon ECS or third-party container orchestration platforms (Kubernetes, Docker Swarm, Nomad or Rancher) to provision, manage and scale containerized applications without having to deal with the underlying servers. The service continually monitors CPU, memory and port availability, scaling up infrastructure to handle incoming pods or tasks and scaling down when instances are underused. Spotinst says its Kubernetes footprint has grown over 10x since the beginning of the year.

For lift-and-shift, database or dev/test workloads, Spotinst created an option for running stateful workloads on excess capacity, which is currently supported on AWS but slated to work on other clouds soon. With this service, Spotinst takes periodic backup snapshots of the instance; if the machine is terminated, either by the provider or on a schedule, Spotinst will automatically spin up a replacement and can reattach the root volume, private IP address and EBS volumes to replicate the application and its dependencies. To the customer the switchover looks like a reboot; the company says some enterprises are using the service to periodically refreshdev/test environments during a scheduled maintenance window.

Other tools in the Spotinst portfolio complement the company's focus on optimization and automation. They include the Multai Load Balancer with support for hybrid environments including on-premises, Functions for serverless development, and the Spectrum monitoring service for cross-cloud visibility, analytics and alerts. For DevOps teams, Spotinst supports a variety of provisioning and CI/CD tools including AWS CloudFormation, CloudFoundry BOSH, Terraform and Ansible templates, Chef integration and a Jenkins plug-in.

Spotinst notes that the pure functions-as-a-service market has not grown as fast as expected. Instead, driven in part by the broad success of Kubernetes, the company is working to enable serverless containers – an area where the hyperscalers themselves are rolling out platform-specific offerings. On the roadmap for Q4 2018 is a container scaling service that will alleviate the need to select servers for running containerized applications – it monitors container workloads and automatically and dynamically selects the best instances to host them, similar to AWS Fargate and Azure Container Instances. Another roadmap item, also scheduled for Q4 2018, is an on-premises implementation of Spotinst for VMware environments. The tool is being developed for quick deployment on any Kubernetes cluster to enable real-time dashboards and analytics detailing usage, performance and potential cost savings of on-premises containerized workloads.


Business Model

Prior to the recent $35m funding round with Highland Europe, Spotinst hadn't yet touched the $15m invested last summer by Intel Capital. It plans to use the cash for R&D, to execute on its roadmap and to grow aggressively, organically with the addition of marketing and sales personnel and possibly also via acquisition. The company cites Highland as a strategic partner that can help it expand further into Europe and accelerate growth while maintaining its customer focus.

Spotinst uses a pay-as-you-save billing model, taking a 20% share of the savings achieved with Elastigroup versus on-demand rates.

The company has a full complement of technology, MSP and reseller partners. It is an AWS Advanced Technology Partner and is available via the Azure Marketplace. The company says the biggest challenge to working with MSPs is that few are incentivized to save customers money, but it cites Logicworks and Israel's AllCloud as exceptions.

Spotinst has more than doubled its head count since 2017 and now employs over 110 people across four offices in San Francisco, New York, London and Tel Aviv. The company says its customer base has grown rapidly and now numbers over 1,000, primarily in the US and EMEA. We estimate Spotinst's trailing 12-month revenue at around $20m.




Spotinst's singular focus on using large-scale providers' excess capacity to save customers money is unique in the industry. The company had two AWS-focused rivals that were acquired: AWS purchased ClusterK in 2015, and Cisco bought (formerly known as Batchly) in 2017.

Spotinst does compete with parts of some larger cloud management platforms, including Platform9, which just released a capability for running Kubernetes clusters on AWS spot instances. Cloud vendors themselves have also improved their budgeting and workload-optimization capabilities, most notably AWS, which has added Spot fleet (for automatic bidding on spot instances) and AWS Batch (for automated scheduling and processing of batch workloads) to its portfolio.

Jean Atelsek
Analyst, Cloud Transformation and Digital Economics Unit

Jean is an analyst working across 451 Research's Cloud Transformation team and Digital Economics Unit. In addition to producing the quarterly Cloud Price Index deliverables, Jean covers vendors and cloud providers that offer technology or services to manage or improve public and private cloud TCO, performance or consumption. She has developed a niche in new private-cloud pricing models, including pay-as-you-go and build-transfer-operate.

Jeremy Korn
Research Associate

Jeremy Korn is a Research Associate at 451 Research. He graduated from Brown University with a BA in Biology and East Asian Studies and received a MA in East Asian Studies from Harvard University, where he employed quantitative and qualitative methodologies to study the Chinese film industry.

Aaron Sherrill
Senior Analyst

Aaron Sherrill is a Senior Analyst for 451 Research covering emerging trends, innovation and disruption in the Managed Services and Managed Security Services sectors. Aaron has 20+ years of experience across several industries including serving in IT management for the Federal Bureau of Investigation.

Want to read more? Request a trial now.