AWS Cost Management – How to Shorten the AWS Maturity Model and Save Costs Right From the Start


AllCloud Blog:
Cloud Insights and Innovation

Starting with and having a fully automated and cost efficient AWS Cloud service in place takes a lot of experience, and not only with AWS, but also around AWS’ Technology Partner Ecosystem. The biggest challenge is to keep pace, while AWS provides new services nearly every day, and that’s not even considering all the partners and their product updates and all of this has an impact on your current AWS spending too.

Expectations are high when you “move to the cloud.” They get even higher after the first deployments are up and running successfully and no harsh costs have hit you yet. So the expectations are:

  1. Business needs will be covered by IT in shorter time frames:
    Weeks instead of months and years, new features and functions are expected to be available much faster,  40% expected improvements.
  2. Innovations: 
    Speed of innovation and its availability is expected to grow by 35%.
  3. IT SLAs: 
    IT service quality, stability, availability should increase by around 25%.
  4. Cost: 
    Cost reduction is usually expected to be around a minimum of 30%.

On one hand to meet these expectations you are challenged by focusing on your core applications, their improvements and new features, and on the other hand to get your cloud operations running smoothly and cost effectively. Do you really want to do this: monitoring and logging all your cloud infrastructure services, operating worldwide 24/7? Especially when it comes to security, you want to have a specialist at your side. Taking care of threats is very time consuming, even though you are operating within one of the most secure and economical cloud services.

Bottom line, are you the one who wants to analyze cloud cost consumption and associated costs by the minute or hour or on a daily bases? I don’t think so.  Besides, you do not want to have your finance or purchasing manager chasing you, questioning costs increases and you finding yourself  having to justify costs from 1 or 2 months ago.

So let´s have a look at how and when cost optimization usually shows up during the typical AWS maturity model.

1. The Starter:

  • Characteristics:
    Starting to use AWS-Cloud Services is usually done in a kind of chaotic, ad hoc trial process. The mentality is best described as “In the past, I have done things like that!“ Traffic is still unpredictable. You enjoy the speed of on-demanded IT services that are available to you and you test deployments are running smoothly.
  • Other Challenges:
    You start worrying about whether your setup “has enough security around it.” and if  your cloud services are acting according to corporate governance?
  • Cost Challenges:
    You start thinking about how to gain a better cost overview.

2. First Core IT-Service-Deployment:

  • Characteristics:
    The first core IT service is running within the AWS Cloud. You and your team are congratulated for working with cutting edge technology. Your current focus is to keep the system running smoothly and add new AWS services when needed and increase performance.
  • Other Challenges:
    You are starting to invest in your security setup and the DevOps team will define processes individually, somehow caring about company governance and other regulations.
  • Cost Challenges:
    To a great deal costs are still sporadic and unpredictable. You think, it just needs some time to estimate future costs properly.

3. First Automation on IaaS:

  • Characteristics:
    Availability of your provided services to your customers is looking reliable and stable.
  • Other Challenges:
    At  traffic grows, vulnerability and attacks from the outside may increase and you need focus on security and start putting into place processes according to corporate governance.
  • Cost Challenges:
    As far as cost is concerned, you are still maneuvering in a kind of grey zone, sometimes being surprised at month end. You hope costs will grow linearly as traffic grows on a steadily. Now’s the first time you realize that AWS costs are not increasing linearly as you thought.

4. First Automation on PaaS:

  • Characteristics:
    More and more IT services are being deployed within the AWS cloud in turn reaping the benefits of lower IT costs. Periodic schedules are defined to analyze performance based on KPIs and benchmarks. A cost control system is in place to regulate your cost management operations.
  • Other Challenges:
    You have processes for Iaas and PaaS not only institutionalized but you are starting to look into automating as many tasks as possible. You are also focusing a great deal on your existing security setup.
  • Cost Challenges:
    You are beginning to look into cloud usage and cost trends before you start deploying new services.

5. Continuous Improvement Cycle:

  • Characteristics:
    Introducing new intelligence and product improvements to your various IT services can be done faster than ever as long as you have your AWS Cloud Services automated. Not only overall but for each service you have well defined KPIs in place so that you can measure and have regular optimization sessions on IaaS and PaaS with a competent partner.
  • Other Challenges:
    You are investing to assure an institutionalized improvement cycle stays established and you put a lot of energy to keep your stakeholders to stay motivated to play an important part within this continuous improvement cycle.
  • Cost Challenges:
    You have a complete cost overview and in depth costs and trend analysis always at hand. Other non-AWS IT-services want to have access to your cost control system.

Let’s put this into perspective – Do you want this to happen in these stages? Or would you rather it happen right when you start using AWS Cloud Services?

Do the right thing, right away!

A successful AWS deployment means focusing on the right sized AWS setup, automating as much as possible, 24/7 monitoring and having a cost overview always on hand. Of course, it takes a lot to do it right the first time; there are so many areas to cover which is why it’s very helpful and even more efficient to have a trusted advisor by your side. Once you have a good grasp and the manpower available – do it on your own until then let an expert help you.

Let’s look now at how cost optimization on the AWS cloud starts off. Well, it actually starts with an expert who has done this more than a few times, try to find one who does it every day for nearly 500 customers worldwide. This type of cloud expert will understand the following aspects of your IT operation:

Finance:

  • Investment & ROI
  • Capex to Opex – improving efficiency on capital and cost
  • Cash Flow & Ease of Controlling
  • Cost Optimization

Organization:

  • Expanding your governance, while operating in the cloud: new roles, new access management, agile methodologies
  • Process Optimization – improving collaboration
  • Change Management – continuous optimizations and improvements and ongoing self actualization.

Supply-Chain:

  • In- & Outsourcing – new resources to remain in the fast lane
  • Expertise-as-a-Service -on-demand when you need it
  • KPIs & SLA Driven – alerting when it gets real critical
  • Ongoing Improvement Cycle – shorter “Time-to-Value”
  • Lean IT Portfolio Management – improving transparency

Technology:

  • Proper Application-Portfolio Management
  • Reduction of Project Management
  • Standardization & Automation

In order to focus purely on the cost management of your AWS deployments, the following best practices on continuous cost optimization need to be covered:

Step 1

  • Analysis of IT services currently running
  • Identification of potential savings, adjusting used technology and setup
  • Identification of potential savings on used cost models, based on consumption
  • Benchmarking in comparison to other deployments (within eCommerce, SaaS, websites, different industries, etc.)
  • Feasibility and recommendations for future activities

Step 2

  • Deploy recommended activities based on best practices to achieve a fine-tuned and right-sized AWS service setup
  • Deploy recommended activities on optimized cost model to reduce costs
  • Deploy a Cost Control Center, including KPIs on usage and costs, break down in booking accounts if needed, trend analysis on usage and costs, especially underutilization, wasted resources etc.
  • Establish a regular, weekly/bi-weekly reporting and discussion on additional improvements
  • Provide finance, controlling, and purchase access to your cost control cockpit

Step 3

  • Setup an automated deployment process for future deployments on IaaS and PaaS
  • Define a service roadmap and onboarding process with a partner asap to discuss implications on cost and technology and follow their recommendations based on benchmarks and best practices

Choosing an unfit partner will lead to a cost benchmarking which just does not work for you and as a result lead you to working with the wrong KPIs. Ultimately, you’ll find yourself again in a trap, spending more money on services than you would have, had  you done it right from the start.

In conclusion, the proper AWS cost optimization process can be very challenging to find but it’s fairly easy once you have a knowledgeable partner by your side, especially one who does it for hundreds of AWS customers every day.

Reducing the typical AWS maturity cycle together leads to:

  • 1st -ability to remain focused on core products and differentiators
  • 2nd – always being on track with latest technology, especially when it comes to security
  • 3rd – enjoying peace of mind when it comes to continuous optimized cost management

In the end, you’ll find yourself adding a great deal to your company, not only cost wise, but also playing a vital part in increasing their time-to-market for services and products, right from the start.

Robert Klimke

Read more posts by Robert Klimke