Nativo Case study

How Nativo achieved over 22% cost reduction in a week without any code changes or R&D efforts

Download Case Study


Nativo empowers brands and publishers with the world’s most advanced technology for content. For brands, Nativo enables storytelling at scale with the largest native reach and reveals insights that unlock return-on-content. For publishers, Nativo enriches monetization with the most comprehensive platform for next-generation ad formats and breakthrough technology for accelerated web pages. Nativo’s mission is to equip advertising for the age of content, improving the web experience and creating meaningful connections for today’s digital consumer.

We were surprised by the scale of the performance improvements and cost reduction we were able to achieve using Granulate.
Eric Hsu, Senior DevOps Engineer, Nativo

The Challenge

As a leading advertising platform, Nativo serves native ads at scale with strict performance requirements. Nativo’s ad stack is designed and deployed as a fully distributed system that spans multiple geographical locations. This architecture ensures shorter response times (due to lower network latency) and fault tolerance, allowing the system to successfully serve and deliver ads even in case of a catastrophic failure of an entire region or data center.

The Nativo platform is hosted on AWS EC2, heavily utilizing on-demand instances across several regions. Operating with the growing business success and customer base growth leading to constantly increasing scale, Nativo also experienced rising cloud costs and a need to improve performance.

The Results

Intel Granulate was installed on several of Nativo’s AWS EC2 instances using an Ansible playbook. Following several days of autonomous learning and analysis of the resource usage patterns and dataflow, the solution was ready to be activated and begin optimizing performance.

Following the activation, Nativo experienced significant performance gains. The performance improvements directly translated to reduced compute costs due to an automatic cluster size reduction. In addition to the 22% reduced costs, Nativo saw 31% reduction in the response time and 18% throughput increase on instances with activated agents.

After this initial activation, Nativo deployed Intel Granulate on all their EC2 autoscaling groups in three different regions.

Reduced Average Latency
Increased Throughput
Reduced CPU utilization

Nativo turned to Granulate’s real-time continuous optimization solution to reduce cloud costs, but also guarantee performance and stability - Nativo’s workloads have to adhere to strict SLA requirements. Nativo achieved dramatically improved performance while reducing cloud costs by 22%.