We use cookies on this site to enhance your experience. Visit our Privacy Policy for more info.

AWS – AWSome or AWFul Setup? Let us help.

Steve Rabin | August 29, 2018| 1 min. read

Insight SaaS companies don’t need a lesson on the importance of efficient infrastructure management.  What you may not know is that Insight has a tight relationship with Amazon because AWS is an increasingly large cost item for many companies, and strategically important to service delivery.

More than 85% of portfolio companies host on AWS and in 2018 companies will spend close to $150 million. Managing AWS effectively can companies save money, IT resources and ensure a quality customer experience.

Onsite’s Tech Center of Excellence collaborates closely with Amazon’s team. In fact, they have a group focused exclusively on serving elite venture capital firms, like Insight, and our portfolio companies. This team's mandate is to ensure that Insight companies get the most value from the AWS platform as well as strategic access to Amazon more broadly for GDPR help, access to skilled resources, early release software, and much more. 

In 2018, we ramped up our collaboration for both technology and go-to-market strategy.

Technology Strategy

The tech focus is on cost and service optimization, which includes: 

  1. Architecture: Companies can understand the current architecture and AWS service usage and determine if more effective outcomes can be achieved. Serverless architectures for elastic scaling is one example, while more efficient usage architecture is another one. Taking advantage of new services (like AWS’ machine learning service suite) falls into this category. 
  2. Purchasing: Companies can identify areas of spend throughout the AWS stack where predictability can be a cost advantage with reserved instances (RIs), spot instances, and private pricing, all of which drive significant savings versus on-demand spend. This also includes options multi-year agreements for additional savings. 

It is worth noting that these are not sales programs, so the goal is not to sell Insight companies more. The goal is to optimize efficiencies and spend through ensuring companies are aware of, and avail themselves, of all features and options. 

More than a dozen companies are program participants, and some measurable results are emerging. In total, the program will likely save 8-15% spend with some example results below: 

  • One of Insight’s cyber-security companies will save an estimated 30% on its AWS spend over the next 3 years. This is a result of correctly sizing database environments as well as purchasing compute and database service capacity upfront via reserved instances (RIs). Additional savings come from a multi-year agreement. 
  • A data protection/ backup company, will save an estimated 13% on its AWS over the next 4 years. This is a result of optimizing data storage, utilization of spot instances and savings from a multi-year agreement.

Go-to-Market Strategy

Amazon’s team is focused on activating 'AWS-as-a-channel'. The goal is to drive incremental opportunities and revenue for portfolio companies selling ‘Built-on-AWS’ solutions to existing AWS customers. 

  1. AWS Partner Network (APN): Join APN to benefit from joint collateral development, campaign execution (including AWS co-investment) and promotion on AWS marketing and social channels.
  2. AWS Marketplace:  By selling through AWS Marketplace, AWS customers can offset/add partner solution purchases directly to their AWS bill. Importantly, AWS sales teams attain quota for sales of partner solutions so there’s no channel conflict. 

One of Insight’s large-data analytics investments is availing itself of these capabilities for their U.S. launch. This includes working closely with the APN partner engineering team to structure a proof-of-concept test on a massive-scale data set. This is occurring ahead of potential engagements with large AWS customers.

Marketplace and APN don’t only apply to AWS customers but to selling to Amazon itself. Insight’s Amazon partners are committed to connecting Insight companies with the right decision makers inside Amazon. Several companies have leveraged this opportunity to win new Amazon LOBs as customers.

For the techies who like to go deeper, we’ve added more information below on the different service options and what's involved.  If you’ve read enough and interested to learn how your AWS infrastructure and spend could be optimized, please email Steve Rabin at srabin@insightpartners.com

Insight’s goal is to ensure your experience working with AWS is AWSome. IT Teams, read on below. 

In order to reduce spend it’s necessary to analyze where the spend is concentrated and how to minimize the expense. This involves exploring hosting attributes like services being used, workload, regions, data management, etc….  

Amazon will provide a free assessment service to understand the what, when, where and how of the hosting environment. Once completed, Amazon and the company collaborate to: (i) ensure their cloud environment is optimized from the services utilization perspective, and (ii) explore the programs available to reduce spend. 

In addition to service usage optimizations there are three facilities worth considering.  These are-

  1. Reserved Instances (RI) - basically, companies reserve capacity globally or in specific “Availability Zones”. In addition to the discount (up to 35%) compared to On-Demand pricing, instances can be reserved based on a schedule or tied to peak usage.
  2. Spot Instances - this represents spare EC2 compute capacity that is highly discounted compared to On-Demand prices. Spot instances can be interrupted if/when Amazon needs the capacity back. While this limitation doesn’t work well for some production use cases, it is easy to mix and match spot instances with ‘regular’ uninterruptible instances. Spot instances are ideal for test and development environments, stateless web servers, analytics, machine learning, batch processing and high-performance computing (HPC) workloads. 
  3. Enterprise Discount Program- EDP provides discounts based on reserving future capacity. These discounts are available for both compute and storage workloads along with the use of AWS services (Lamda, for example).  

This is not a one size fits all exercise. Results will vary from company to company. To manage expectations, and based on the hosted environment, some portfolio companies will see marginal optimization opportunities. However, many companies have already seen meaningful spend reductions.