Many companies experience that cloud costs grow faster than anticipated. As a result, this slows down cloud adoption and/or creates budget overrun. 

A frequent reason for cost surprises is not taking full advantage of all the options for savings available in Azure and assumed in the original business case. 

So, we will take a practical, financial approach. You will benefit from reading this even if you do not have a technical background. 

If you are an experienced cloud cost manager, you will know the concepts. Still, you might learn a thing or two from the real-world examples and how to get it done. This includes a list of free tools like Azure Pricing Calculator, Cost Management Service, and Advisor.

Savings option #1: Azure Reservations

What are Azure Reservations?

We know, we know, you already know about reservations. Yet, even our best customers would benefit from using even more reservations.

Buying Azure Reservations (Azure Reserved Instances) is committing to purchasing a service for a one- or three-year period. Doing so will give substantial discounts, typically 60%+. 

Moreover, you can exchange reservations within each of the 18 categories. So it is not like buying a specific piece of hardware. In some situations, you can even get an unused reservation refunded.

With this level of flexibility, it is essential to approach reservations in the right way. Ask not “Will this VM, Database, Disk, be running in 2 years”. Instead, ask “Will my use of ALL compute on azure (VMs, Functions, Containers, etc.), be even just 2/3rds of what it is today in 2 years?”. If so, buy the reservation and balance them often.

It is also worth noting that the reservations now do not need to be paid upfront, so it is not a problem from a cashflow perspective.

You can read more about reservations in a recent blog post.

Reservations in the real world

Every company we have worked with uses reservations to some degree, but no one leverages the full potential:

  1. There are a total of 18 categories of reservations, each with multiple different reservation types. 
  2. Every quarter new reservation-types are added or changed. 

Add to this a few “tricks” you must know to master reservations, and you understand why.

“A very cost-focused client exercised great discipline and always bought the right-sized VM, which unfortunately could not be reserved. However, a cheaper solution was to upgrade to a more expensive VM type and then get the 65% reservation discount, delivering a 45% saving compared to the cheaper, less performing VM.”

“A large enterprise client was shutting down servers every weekend to save costs. However, buying reservations and keeping the servers running saved them 20%.”

While these examples are not hard to understand the nevertheless illustrate that your task when managing cloud costs is managing a very large number of combinations. The real challenge is not understanding what needs to be done – rather it is finding the needle in the haystack which delivers the the largest savings.

Savings option #2: Azure Hybrid Benefits

Another way to get significant savings in Azure is Azure Hybrid Benefits (AHB). This is a strategic move from Microsoft to get on-prem clients to move workload to Azure by allowing the same kind of licenses to be used on-prem in the cloud. So, this means you can leverage existing licenses with active SA. You can also benefit from renting or buying new licenses.

These savings are huge, in some instances close to 90%! Hence, understanding and leveraging this option is another must-do.

The implementation is straight forward – it is simply a checkmark in the server configuration that you are bringing your own license. Again, do make sure you own that license or rent or buy new ones before setting the checkmark.

AHB in the real world

Managing AHB is quite complex because you need a detailed understanding of licensing types and options. We often see results of non-compliance or deferred decisions.

A cloud-native startup had no on-prem licenses and thought that AHB was not an option for them. They were not aware that they could rent the server licenses from their cloud service provider and use them for AHB. This saved them 92%.

A large, global enterprise customer was leveraging AHB systematically. Based on external advice, they were about to purchase Windows Server Enterprise licenses for AHB, where Windows Server Standard would suffice. Buying the Standard edition meant a saving of 84%.

Whether you are a startup or a global enterprise customer make sure that you get external advice on licensing – and make sure it is from an independent expert who is not directly or indirectly making money on your Azure spend.

Savings option #3: Azure Elastic Database Pools

This is the third of the must-do savings types because the potential savings are significant, 60%+. 

Nearly all customers use SQL Server in Azure. And for most customers, it’s one of the top 5 expenses.

Azure has the problem that frequently, the minimum size of the database you can buy is too big for the workload you want to run. The answer to this problem (and quite a few other problems) is elastic pools.

We will not go into the technical details here, but it is a way for you to get economies of scale and only pay for what you use.

Elastic Pools in the Real World

Elastic pools is the most important way to get some economies of scale in the cloud – make sure you understand all the options here.

A small Software-as-a-Service customer was spinning up a SQL Database for each new customer who ordered a trial. With the smallest standard database cost of €12.4 per month, that adds up. However, none of these customers use the performance available even in this, the smallest instance, and most are very rarely accessed at all. Creating a small vCore Elastic Pool saved them more than 75% of their Azure SQL Database cost and allowed them to keep adding customers with nearly 0 marginal costs.

Other Ways to Save in Azure

There are many other ways to save. To mention a few:

  • Use the newest version of a service. This is how Microsoft introduces lower prices. New versions are cheaper than the versions they replace.
  • Use Azure discounts for test and development servers. If you have Visual Studio Subscriptions for your developers, set up a Test/Dev subscription and avoid paying for Microsoft products.
  • Find the right model for paying for data egress (that is, data you pull out of the cloud). 

And we have not even moved into how you can save by reducing how much you buy. If you are interested in diving a bit deeper into this subject, you can read our article on Finops.

The bulk of savings are made from the top three categories. Therefore, our advice is clear and simple:

“Make sure you master Azure Reservations, Hybrid Benefits, and Database Elastic Pools, and make quarterly systematic reviews of these”

How to get savings done?

It should be clear that there are substantial savings opportunities in Azure, but you probably recognize one or more of these challenges:

  • Time and Priority: Your focus is delivering and operating cloud services, not managing costs.
  • Complexity and Changes: It is hard to get a complete overview of the options and new services and savings opportunities.
  • Knowledge: Even when you set aside enough time to identify savings, you frequently run into a question you would like to get answered before implementing it.

If so, you are not alone. 

Azure tools can help you with some of the savings

Azure offers a comprehensive suite of tools that can help you to get the job done. 

  • Azure Pricing Calculator can help you estimate the costs of a service you consider setting up. Remember to include things like Storage and Back-Up (these are not  included by default)
  • Azure Cost Management and Billing can help you understand your utilization and cost in your environment. We haven’t touched on the benefits of saving on how much you use of each service in this blog, but it can be substantial.
  • A note on Azure Advisor. While this tool is supposed to include cost optimization recommendations, we find it often actually increases costs. Be careful when following the advice for cost recommendations.
  • Cloudyn is now free for your Azure spend, and you can get it to cover your AWS, and Google Cloud spend as a paid service. 

No tool will deliver savings

… just like your bathroom scale in itself doesn’t lose you any weight. 

If you google azure savings, cloud cost management, finops, or other keywords related to saving on Azure and other public clouds, you will find an endless list of tools. You will find Apptio, CloudHealth, Densify, and many more. They are all great tools!

This mere list of free Azure tools indicates that the problem is not getting data points, ideas, and recommendations – even for free.

The problem is that you don’t have the time to properly understand the recommendations or filter through the multitudes of options to find the ones that make the most sense.

So then, how do you do that?

Well, consider this. What if you were getting a service where you quarterly scheduled a couple of hours in your calendar to go through a finite list of material savings recommendations with a guide that can answer any questions or concerns, you might have? You know, like the Kostner Cloud Cost Management Service

Final Reflection: Managing Azure Cost is just one of a New Trend

If you look around your company, you buy more and more technology “as-a-service”. There are good reasons for this, but it introduces a new paradigm for cost management that matters for everything you buy as a service. If you want to read more on this, please download our whitepaper on “Digital Transformation: Everything-as-a-Service”.

For each of these new services you are buying you will be left with the choice: Should you do it yourself or get external assistance.

Surely, you could do it yourself, but with more and more services, is managing costs really what you should be spending your time on?

If you are like most, you probably experience IT costs not being stable as they used to be. You might experience some budget overruns or have a nagging feeling that you are overpaying for some of the new cloud services. 

But don’t worry. This blog post will outline a simple roadmap for your FinOps journey, so you can focus on delivering business value.

What is FinOps?

FinOps is a concatenation of Financial Operations, just like DevOps is Development and Operations. It is the name of a new task that all consumers of the cloud must do.

 

In the old on-premises times, optimizing IT costs was mostly about optimizing how you bought technology. You negotiated large hardware or software deals with teams from multiple departments, including legal, procurement, finance, and IT. 

In the cloud, there is no single large purchase. Instead, you buy services continuously as developers deploy workload, and bills are monthly and therefore seem “small” compared to the hardware purchase made once a quarter. 

Make no mistake, though; the total amount is large.

The Cloud Requires You to Think Differently about Cost Management

Before we move on, we have a few statements to consider regarding how IT, finance, and procurement are working together to manage your cloud costs.

In our company we agree that:

This is not the time or place to detail each of these critical questions, but if you want to dive further into these questions, you can read our blog “Five critical elements in your cloud cost thinking.”

Surely, this requires a vastly different approach than in the on-prem world. Welcome to FinOps!

If you do not manage cloud costs with the right approach, you run the risk of overpaying. Maybe this is the reason why the average budget overrun reported in a recent survey by Flexera is 23%!

The Four Key Components in FinOps

FinOps is the practice of managing cloud costs in a variable spend model. Or simply put, FinOps is about managing this equation:

FinOps

 

In this equation, Price is about keeping your unit costs down, i.e., getting the services you buy today as cheaply as possible. You can do this in many ways, but fundamentally it boils down to: 

  • Optimizing price by working the cloud service catalog, price and discounts models (applies to all organizations), and 
  • Negotiating enterprise-specific discounts and special terms (if you have sufficient scale). 

The great thing about price optimization is that you can do it without impacting the business.

Quantity is making sure that you are not buying more than you need, i.e., by shutting down unused services  (life cycle management) or purchasing the necessary capacity (right-servicing). These types of decisions typically require the involvement of the business owner and come in two maturity levels:

  • Inform the business owner about what they are spending, and this will immediately have an impact on the quantity consumed
  • Manage quantity in a structured process where you monthly or quarterly sit down with the business and go through material costs

A Roadmap to FinOps

No matter if you are a large or small enterprise, private or public, technology-based or traditional products and services, we experience the following roadmap providing the best results:

STEP

01

Optimize Price

In order to optimize price you need to take a snapshot of your current environment, optimize it using a tool with a complete up-to-date service catalogue and price/discount list, validate the findings and make the adjustments in your portal.

STEP

02

Inform about Spend

When you have optimized price – and taken the largest part of the savings available – you can start creating transparency on who spends how much on what.

STEP

03

Manage Quantity

This is where you need to get the business involved to make decisions about shutting down services (lifecycle management), shutting services on/off, and buying the right capacity services (rightsizing).

No matter if you are a large or small enterprise, private or public, technology-based or traditional products and services, we experience the following roadmap providing the best results:a

Want to know more about each of the steps? Read our blog on “Fasttrack Roadmap to FinOps.”

This roadmap provides excellent results because price optimization delivers considerable savings, making it easier for you to move on to business-critical decisions on quantity.

And if you get external assistance, you can optimize price very fast!

Get external help for price optimization

Another reason to start with price optimization is that it is well suited for external assistance.

You can get more savings, you can get them faster, and you need to spend only a 10th of the time or less.

The reason is simple – you share the cost of the time-consuming parts with other companies. Don’t worry about tracking changes, getting your spend, analyzing it, reading the fine print, etc. 

You want to retain control of your environment by validating the recommendations and ensuring that they do not run contrary to any plans you might have.

You might argue that you already have a person or a team doing this. Indeed, they are doing a great job on this. However, ask yourself if they can create more business value if they spend their time on other tasks than price optimization.

FinOps in the Real World

All of our customers have one thing in common – they care about not overpaying and want to stay in control. Some of them have large teams, other just a single person with limited time. They understand cloud technologies and all the concepts required to manage cloud costs. Yet, there is always time and money to be saved.

Let us share with you a couple of examples.

New Possibilities

 A Media client in Azure was spending large amounts on App Services. There had been no ways to save on these services other than reducing the number you ran. In the fall of 2020, Microsoft suddenly introduced reservations. Now, by upgrading these instances from standard to premium and reserving them, clients could save 60+% on these resources. We notified our client, and within one week, the customer implemented the change and saved hundreds of thousands of euros annually. Continuous monitoring of new announcements is an essential part of FinOps – but also one you can easily outsource.”

A highly technical save

One of our global clients run a large data warehouse in the cloud. It runs on massive machines with very high performance. However, on the cloud vendor’s advice, they had configured the environment with so much storage performance that they exceeded the limits of what the database servers could handle. The result: they were paying tens of thousands of dollars for the performance they were physically unable to utilize. The fix? Adjust the performance setting for the storage drives, instantly lowering costs by 50%.

The Aha! moment

We recently facilitated one of the first Showback review meetings for a client. They showed their development teams the costs they were driving, including the resources with the highest spending. What did they find? “That, that has been inactive for at least 18 months, I thought it was gone!”. You would be surprised how often this happens. No-one has the job to review all running services, so even with a team more than willing to help, this is not their priority. A 20-minute online meeting saved nearly $100k, by merely creating transparency and responsibility.”

FinOps is much more than delivering savings

The fact is that most cloud journeys do not have a cost focus. Rather, it is about digitalization, speed, and agility. Therefore, you might state the success of your FinOps journey in the direction of:

  • Getting more digitalization for the same money,
  • Avoiding digitalization slowdowns due to cost overruns,
  • Getting more time in the cloud teams to building and operating your cloud rather than managing costs.

Is this true for you? If so, working with an external partner like Kostner will help you focus more on success and less on the commodity parts of price optimization.

Want to know more?

Hopefully, it should be clear by now that we are here to make your life simpler. You will save more, faster, and your team will spend less time on FinOps and cloud cost management.

To make your decision as simple as possible, we have two completely risk-free ways of finding out if you want to work with us:

  • Book a free meeting, where you can learn about how we would work with you and ask any questions you may have, or
  • Find out exactly how much you can optimize price by taking our free assessment. We will go through our optimization results, and you can ask any questions you might have.

We are looking forward to making your life a little simpler.

In a recent blog post I described why you need a quarterly cloud cost tune-up.  In short: Your cloud environment is dynamic with new services being added continuously. The cloud providers add new generations of services, new licensing options and changed discount programs. Hence, these dynamics requires a dynamic cost optimization approach.

The question of the week is who should be responsible for the quarterly tune-up? And if you remember my question from last week … how do you get it done in 3 hours?

What are other companies doing?

From our research and customer work we have found that the budget more often than not is owned by the CIO, Head of Infrastructure, or in some instances in a Cloud CoE (Center of Excellence). Sometimes procurement has some responsibilities in price optimization as well and even Finance is occasionally involved.

However, what is almost always true is that the teams actually deploying services to cloud are essentially responsible for the cost optimization effort – irrespectively of who has the budget or process ownership.

Is this really the right way?

Or put differently … should the same people who are responsible for deploying new services and their technical operations, spend weeks every month on all the tasks associated with cost optimization, ie. the addition of new services, license models, discount model changes, discount percentages changes, migration options, redundancy options, reservation utilization, reservation rebalancing etc.?

The answer is no.

Why? Because the one thing customers tell us more than anything else is: ”My focus is on deploying to cloud. We are so stretched for resources, we have to keep our focus where we grow the business and put new services and features in play”.

Another option is to turn to the company providing you with cloud services or cloud services operations. We have numerous examples that this from a cost efficiency stand point is not an optimal solution. Consider that they often get paid based on how much you use, and this may have the potential to influence how much of a priority they give to reducing what you pay (and consequently what they make). And again, they too are short of ressources, and prioritize their time in accordance with your priorities – getting things running in cloud.

Either way, the result is that money are left on the table. Money that could be put to much better use.

What are the most successful companies doing?

“The companies that best combine speed and agility with cost efficiency have realized that you should not burden the DevOps or infrastructure teams with an ongoing task of cost optimization. Rather this is done as a Tune-Up, quarterly or monthly, where you review what has been provisioned and ensure this is purchased in the optimal way.”

The most crucial thing is to establish clear responsibility for not only budget ownership, but in particular the process of getting the quarterly tune-up done. This responsibility typically sits in the CIO office, with IT managers/architects or in procurement.

Who is responsible in best-in-class-companies?​

Budget ownership is not the critical part. What really distinguishes successful companies are that there is clear authority to get the tune-up done. This authority is typically held and exercised by the CIO or procurement. They typically rely on external assistance, so that the tech teams can focus on building and operating cloud solutions.

And of the two areas of responsibility, what really distinguishes successful companies are that there is clear authority to get the tune-up done. They do not waste valuable time discussing if it is necessary. They just get it done.

And the way they get it done is always with some level of external assistance.

Why?

Because a quarterly tune-up requires massive data collection and analysis, which is quite generic – and it is simply not cost efficient to do this internally.

To illustrate this: Kostner offers a Quarterly Tune-Up service where you and your team only need to spend 3 hours every quarter on cost optimization. Yes, three hours. The time you free up, you can spend on delivering services to the business. This is a great way of balancing speed with cost efficiency.

What is critical when selecting external assistance?

Make sure the provider is focused on cloud cost optimization, so you do not a) have to pay for “generic” efforts (gathering data or building a tool). b) get recommendations that are not technically feasible, and c) have conflicting interest with your intent to save on cloud

In larger enterprises we sometimes meet dedicated teams, frequently known as FinOps teams. However unless you run a (very) large cloud environment, odds are, you can’t dedicate a person (let alone a team) to this alone.

Even when there is a dedicated FinOps team they almost always rely on external assistance. Typically they focus on the larget ticket items and rely on external assistance for finding additional savings, and providing an independent proof, that they are running a cost efficient cloud setup.

 

What should I do from here?

Well, this really depends on why you are here. A few ideas might be to:

  • Learn more about cost optimization and the quarterly tune-up. You might want to participate in a free webinar on this, or continue reading the blog.
  • Assess your potential for savings. In our universe we would recommend that you simply order a free trial. You will need 2*30 minutes and get a complete review of your environment. No credit card required. No hidden subscriptions. 
  • Discussion with your colleagues if a Quarterly Tune-Up is relevant for you. Again, if you want we are happy to host a facilitated discussion of this where you can ask all your questions. Book a meeting

… or you might want to do something completely different.

Interested? Subscribe to our newsletter or follow us on LinkedIn.

Did you like this article? please share with your friends

Yesterday, I was listening to Kevin Kelly (founder of Wired magazine and other accomplishments) talk about the future of Moore’s law, and how a slowing Moore’s law would indeed be an extremely bad thing for how we use technology. You can see the full talk here.

Moore’s Law a Key Driver in Digitalization

As Kelly states, a key driver in digitalization and the rapid adoption of technology has been the relentless progress summarized by Moore’s law – every couple of years, you get twice the computing power for the same price.

What is Moore's Law? ​

The observation that every two years, compute power doubles – and importantly, the cost is cut roughly in half of each compute unit.

Digital technology has followed this curve, and sometimes even exceeded it as better algorithms have led to even higher increases in performance.

Another key component is that pricing nearly unanimously has been cost+, that is, the cost of production plus an acceptable margin. In this way, customers have always benefitted directly from advances in technology, and as a result, we have ”smart” everything, and digital technology saturates nearly all industries. The role of relative affordability in this digitalization cannot be overstated.

Cloud Subscription Services Changes Who Benefits from Moore’s Law

As a result, it is with some concern I am witnessing the shift to cloud subscription services in the cloud, and the pricing strategies that go along with them.

As customers no longer buy the component parts, but rather a sliced-up portion of hardware or even managed service, they no longer automatically benefit from Moore’s law. They are relying on cloud providers to pass on the savings that Moore’s law gives.

But cloud services are increasingly priced via Value Based Pricing – not cost+. The result? You no longer get twice the technology for the same price every two years. In some instances, price reductions have slowed compared to classic models, in other instances they have stopped, and in some instances, prices increase every year.

As services replace hardware and customizable software, we need to reestablish a competitive landscape that means prices actually drop in accordance with Moore’s law, which has so far driven technology adoption. Otherwise, we run a very real risk of cost becoming a prohibiting factor in further digitizing our companies. This is an outcome that no one other than shareholders of large cloud providers would find attractive.

What Can You Do to Benefit (as Much as Possible) from Moore’s Law?

The first thing you can do is to manage the purchase of your cloud services with higher frequency to ensure that you actually reap the portion of price reductions the cloud providers are passing on to you. You can do this in a quarterly (or monthly) cloud cost tune-up.

What is a Quarterly Cloud Cost Tune-Up?

Every quarter you a) Review the overall growth of the environment, b) Check if there are newer versions of services that offer better performance and lower costs, c) Buy missing reservations and rebalance existing ones, and d) See if there are things that can be combined – decisions that cannot be made when deploying services individually

The other options to reestablish a competitive landscape are a lot trickier. Creating your IT infrastructure to accommodate a true Multi-Cloud setup is where you ensure that you can provision services where they are cheapest. It frequently involves the use of co-connection centres, managed infrastructure services, and building with components like containers, which can run on multiple platforms

Where should I start?

Where to start? The Quarterly Tune-Up is where you can rely on external assistance. Start with that and take the first level of savings here. This can be done in just a few hours per quarter if you choose the right service provider. This will free up resources, budget and time, for you to focus on the other tasks.

Interested? Subscribe to our newsletter or follow us on LinkedIn.

Did you like this article ? Please share with your friends.

Contrary to popular belief, prices for cloud services generally do not drop over time. Instead, new services, or new generations of existing services, are introduced at lower unit costs.

In fact, this is the primary price reduction mechanism for most major cloud providers.

Making sure you are benefitting from each new service or service generation is an important step in maintaining a cost-efficient Cloud Architecture. With thousands of services available on most cloud platforms, keeping track of new services is a time-consuming task.

Why is "checking for updates services" important?

In the cloud prices generally do not drop. Rather new generations are introduced at better price/performance. Make sure you leverage this. Time consuming? Yes, tools or external assistance can help.

We recommend you review your cloud estate once every 3 months, and see if there are new services, or new service editions available that offer some of the following benefits.

Lower Prices on New Generations

Look for pricing benefits that match your consumption model: New generations of existing services often come with lower Pay-as-you-go pricing. In terms of reservation discounts, newer generations often offer much higher discount rates.

An example of VM pricing across 3 generations of D-series VMs with 2vCPU and 8GB of Memory,
all in Western Europe:

As you can see, keeping up with each generation is a good way to save money.

The Quarterly Cloud Cost Tune-Up​

In the cloud prices generally do not drop. Rather new generations are introduced at lower prices. This is why you need a quarterly tune-up to ensure you get the most out of your budget.

New Price Models Become Available

Not only do new generations of services often mean lower prices, they also tend to open up new price models altogether. This is particularly prevalent on AWS. Here new editions may now be eligible for reservations, which the previous generations were not. In other cases, new editions may now allow for more flexible ways of paying for reservations, i.e. no up-front payment.

An example would be ElastiCache on AWS, a service most of our AWS customers use. Here older service family generations like r3 or r4 only offer the legacy reservation type “HeavyUsage” with 50% up-front fee, whereas the latest generation r5 offers “No Upfront” pricing, where – as you might have guessed from the name – there is no up-front fee.

Better Specifications on New Editions

Higher performing, newer processors, more included storage, access to higher tier external storage, better networking performance, these are all things that come with new editions of services. But, there are sometimes tradeoffs, and understanding what the technical changes to a service are from generation to generation is an important part of deciding whether to upgrade to the latest edition.

An example would be the move from 1 virtual CPU per core to virtual CPU per thread – not something you would suspect, if you didn’t read the technical specifications. Most of the time, this has zero real world impact, and the new service can be adopted – it is, however, one of the things you should check for, when you are conducting your quarterly review of which services to upgrade to the newest generations.

Better Fit Features

Services are often introduced with the highest tiers of gold-plating in terms of redundancy etc. Even though you may not use this high tier of feature on other services, you’ll have no choice on specific services. New editions open up the opportunity for lower-tier features.

A good example of this is the recent opportunity to reduce retention and availability for back-up of SQL Server Managed Instances on Azure. Until recently, back-up was only available in the highest tier of redundancy – far exceeding the redundancy or availability choice taken by the majority of our clients. Now, however, you have a choice, and can reduce the availability of back-up storage to match your other services. Savings? Up to 50% or more.

In Summary

By reviewing your existing service and resource portfolio regularly – we suggest quarterly – you can get lower prices, lower overall cost, better performance, and better specifications.

Don’t forget, however, as you change your environment, you will need to be sure to re-balance your reservations and license choices. 

Interested? Subscribe to our newsletter or follow us on LinkedIn.

Did you like this article? Please share with your friends.

The Quarterly Cloud Cost Tune-Up

In this blog, we will focus on how you can make your life a little easier by getting help on the six crucial elements for the quarterly cloud cost tune-up by implementing a) tools, getting outside help from b) service providers or c) consultants.

The Quarterly Cloud Cost Tune-Up​

The dynamic nature of cloud requires a dynamic approach to cost management to ensure you are not wasting money; A quarterly cloud cost tune-up.To do this you need to have six crucial elements in place, ie. data, tools, and people.

Tools​

Using technology is a must for tuning a cloud spend. Most of our customers generate millions and millions of rows of usage data. Far more than can be managed in a spreadsheet. Some organizations are implementing tools like Apptio Cloudability or Densify. Others rely on Azure Cost Optimizer, or AWS Trusted Advisor, both included in your cloud subscription. Tools are crucial – we’ve made one of our own that we use to calculate savings for our customers.

What tools give us is an overview of things you COULD tune, and you will then need to sort through this to find the ones that make the most sense, have the biggest impact in the shortest amount of time, and that is consistent with your future strategy and limitations.

The real challenge in getting savings done is not the tool; it is mainly time and priorities. There is massive pressure on the teams to deliver with the highest possible speed, which in turn means that the cost optimization moves down the priority list – at least until the cost is becoming a significant issue. Then a lot of money has already been wasted.

Hence, rather than asking yourself “what tool should I use”, the most important question for you to answer is, “should I put a team in place that has the time and priority to drive down our cloud costs?”

When should I implement a tool?

A tool is not a solution to getting savings done. It requires a team with sufficient time, and with savings as their top priority. Hence, implement a tool if you have a dedicated team in place.

Cloud partners / Cloud Service Providers

Some service providers in the cloud space include some form of cost optimization in the price. This is a convenient solution. They know your environment; they know the services and prices with your cloud provider, and they have the necessary access to implement any changes directly.

The feedback we get, however, is is that operationally this approach is not working as well as it could in terms of cost management. Unfortunately, the data we have confirm it – there are nearly always double-digit percentages in savings to be made.

Exactly why it does not deliver the best savings is unclear, but there might be a couple of reasons:

  • The partner is focused on the migration/operations/development of solutions more than cost-optimization if nothing else then because this is where you as a customer focus.
  • Like it or not, there is often a conflict-of-interest. Service providers get a commission based on the money you pay for cloud services. If they also operate your environment, their operational mark-up may also be dependent on your spending. Hence, saving you money will take time for your partner and reduce their revenue significantly.

How do I best handle the conflict-of-interest?

The best way to handle conflict-of-interest is by either a) having an independent third party do a quarterly review, or b) unbundling the cost optimization from the service contract

Consultants

You can use external consultants to assist you in your quarterly reviews. It isn’t easy to categorize different types of consultants, but you should at least consider the following when selecting external assistance:

  • Do they have a tested methodology to get a picture of your environment with existing data without installing sniffers, agents, etc.?
  • Have they already a database with up-to-date services and pricing from your cloud provider?
  • Do they have a tool in place to optimize your infrastructure?
  • Do they have the necessary technical expertise to recommend technically feasible optimizations?
  • Is their methodology well enough tested to save you time, so you only need to spend a few hours every quarter?

These questions are critical because you want the consultants to spend their time on your specific issues – not paying for generic data gathering or the consultants learning about this space.

What is critical when selecting a consultant?

Make sure they are focused on cloud cost optimization, so you do not a) have to pay for “generic” efforts (gathering data, or building a tool), and b) get recommendations that are not technically feasible.

Final thoughts

Years ago, when I was living in the Middle East, I discussed with a personal fitness trainer and friend, “What is the most effective form of training?”. He laughed and said:

You know, I love this question, because many think of this the wrong way. They are looking for me to say, “set a massive goal, complete an Ironman”. That’s not the way at all; the most effective form of training is the one you get done!

With the pressure to deliver with speed and agility, one of the main challenges is not to understand that there are significant savings to be made on your cloud bill – the challenge is getting it done.

In next week’s blog, we will discuss in more detail what it takes to get cloud savings done and who should be responsible for what. Until then, I will leave you with the question, “How do I get cloud cost savings done?”. And could I get it done in 3 hours every three months?

Interested? Subscribe to our newsletter or follow us on LinkedIn.

Did you like this blog post? please share with your friends.

Six Crucial Elements in your Quarterly Cloud Cost Tune-Up

Last week we talked about why a quarterly cloud cost Tune-Up is needed (Find out why). This week we will focus a bit more on what it takes to do a quarterly Tune-Up.

The great thing about cloud is that you can provision new services with focus on speed and agility and leave the cost optimization for this quarterly Tune-Up. And after your quarterly Tune-Up you will see savings materialize immediately.

The Quarterly Cloud Cost TuneUp

Just to reiterate, the reasons why we need the quarterly Cloud Cost Tune-Up are:

  1. Cloud pricing is dynamic. Prices and discount models are changing, and new (and cheaper) services are added
  2.  Your environment is dynamic. You migrate infrastructure, change architecture and add new applications and services.

In a dynamic environment with dynamic pricing you need … dynamic cost management.

The Quarterly Cloud Cost Tune-Up is about taking a snapshot of your current environment and finding the optimal price configuration against the service and price catalog of your cloud provider. Following that implementing the changes is easy and fast – as long as you only implement the material savings . Most changes involve merely adding a few checkmarks in your buying configuration – not changing your infrastructure or services.

The ability to complete a Tune-Up is about data, tools and people. You need to have the following in place to cost-optimize:

✔ Data: A complete overview of your current environment

✔ Data: Knowledge of all relevant services and prices from your cloud provider

✔ Tool: A way to combine these two data sets and use it to optimize your cost

✔ People: A person with detailed knowledge of terms and conditions from your cloud provider and technical expertise

✔ People: A person with insight in your technical environment to verify optimization recommendations before implementing them (typically an IT Architect)

✔ People: A person to implement the recommendations

The key question from here is: How do we get it done?

The challenge is getting it done

Most practitioners are well aware of concepts for optimizing their infrastructure. They know of at least some of the opportunities in reservations, pooled resources, looking for new and cheaper service instances and all the other savings options available. Also, they review the recommendations on cost savings available in their cloud portals.

Referring to the checklist above the do-it-yourself (DIY) approach is challenged on two primary factors:

  • With the pressure for speed in delivery is it then realistic to carve out enough time to not only “do” cost optimization, but to keep track of every new pricing structure, licensing model, service introduction etc. for all cloud providers in use?
  • Would you have the necessary tools in place to optimize? This cannot be done by hand unless your cloud consumption is very small.

To get a direction you might want to know the answer to this question:

How do mature organizations get savings done?

They focus on getting the largest ticket items optimized on an ongoing basis and use external validation of overall cost efficiency and getting the last 10% savings home

What we see in mature organizations where dedicated teams are in place, they focus on getting the largest ticket items optimized on an ongoing basis and use external validation of overall cost efficiency and getting the last 10% savings home. Not only is this a sound business case in its own right, but it also gives confidence to have an independent, external third party and an experienced cloud cost architect to validate the trickier parts with.

Three ways to make your life easier

In next weeks blog we will dive into some of the things you can do to make your quarterly Tune-Up a little easier. In fact we will explore three different ways to reduce the time you need to spend.

If you don’t want to miss out on this subscribe to our blog or follow us on LinkedIn

The fundamental equation for what your monthly cloud bill amounts to is:

P*Q

P is the unit price, your pay for the services you buy, and Q is the quantity you buy. In this blog, we will focus on P, i.e., how you minimize the unit prices you are paying for the services you consume in the cloud.

But really, this is not new! No, but … 

It used to be much simpler

In the On-Prem world we have had a quite well-functioning division of responsibilities. IT knew how to architect, build, and operate the environment and every 1, 2 or 3 years procurement ran a process to drive down unit costs.

This all worked well, and with the help of Moore’s law (driving down prices / getting more performance for the same amount) this meant that IT could see costs remain fairly stable or even go down, despite increased consumption from the business..

Buy and Build is changing … a lot

With public cloud a new level of cost optimization is required for a couple of reasons:

  • “Buy” has split into two things: 1. the annual or tri-annual negotiation with major vendors about volume discounts, and 2. The act of buying individual services as you build your infrastructure.
    The volume discounts apply to the public price list and is typically single digit. This should be compared to the discounts you can get leveraging discount and smarter purchase options in the price list that offers far bigger savings potential, up to 70% or more – and this is not only by buying reservations!
  • “Build” is now happening with bigger speed and agility. The pressure to deliver features and performance faster is huge. The individual decisions in the build phase are well made, but the combined infrastructure offers additional price reductions by pooling resources and allocating existing resources efficiently.

A Quarterly Tune-Up of Cloud Cost is Needed

These two factors in themselves drives the need for a quarterly Tune-Up process (or even more frequent) in order to leverage the “buy smarter” options from the price list.

The need for a quarterly Tune-Up is further amplified by the dynamic character of cloud environments and cloud pricing. Let us give you a few examples:

  • Did your cloud provider introduce new reservations you could apply?
  • Did your cloud provider introduce new, cheaper services with better performance? (Yes, that happens almost every quarter)
  • Can you leverage the pooling of services because your environment expanded?
  • Did you provision a new server, where you can save more by applying the BYOL on this new server than where it is used now?

Surely you know of these types of savings, and many of these were considered when you provisioned the service initially, but the changes of pricing and discount models are so plentiful and highly frequent that there is much money to be saved by getting this right. You could say that when cloud providers say that cloud offers speed and agility – this is also very true for their price lists.

What Will You Use Your Savings on?

These periodic Tune-Ups offer significant potential for cost savings.

  • In mature companies spending millions annually, with limited growth and few changes in their cloud environment, and with a dedicated team to manage cost, the periodic reviews generate 10-15% annual savings.
  • In companies at the beginning of their cloud journey, where the focus is on migrating infrastructure and building new services, the savings potential is typically 15-30%.
  • For cloud-native scale-ups focusing on company growth and deploying new features, the annual savings range is even higher at 20-45%.

Even if you do not have a savings agenda for your cloud initiatives, you could surely put the money to better use.

… But How do I Get the Time to do this?

We understand you are very busy, but join us next week where we will explore some of the options of getting this periodic review done faster, whether you are looking to do-it-yourself or with external assistance. Also, we will touch upon the question of who should be doing what internally.

Interested? Subscribe to our blog or follow us on LinkedIn

Did you like this blog post? Please share with your friends

An informal poll during  a #IACCM (The Global Contract Management Association) event gave the following distribution of who is currently in charge:

I can’t rightfully say this is surprising – it aligns well with our own experience – but it is a big step away from how things were done in the on-premises world. In the old days, if you bought outsourcing services you absolutely would have involved sourcing. And Vendor Management would then likely run things on a day-to-day basis.

What is really interesting here, is that IT Infrastructure teams and Cloud Center Of Excellence’s account for 67% – that’s TWO THIRDS – of those responsible.

Now, IT Ops and COE’s tend to have a core focus and priority that is all about stable operations and fast-deployment architecture. And that is exactly where their focus should be. BUT, when you task people who are more than busy with that, with also being the ones who have to keep up with all the ways you can buy cloud, the discount models, the reservation offerings, and the geographical price differences – not to mention the licensing options, payment plans, long-term contract options etc., you are asking a lot of a very busy team.

As a result, Cloud Cost Optimization will often end up being way down on the list of priorities, and the effort and impact will be limited.

So what to do? Well, in the words of a head of IT Operations at a large SAAS company – Cloud Cost Management is a job in and of itself. It simply takes more time to keep up with ways to save, and analyzing those ways and how it applies to your environment, than someone can set aside on top of other tasks.

If you are a large enough company, it might well make sense to invest in a FinOps (Financial Operations for Cloud) team.

If you are a smaller company, or a large company with a smaller cloud budget, you are going to want to get some help. Because who should be in charge of managing Cloud Cost Efficiency? Someone who has that as their number one priority, and number one task. 

Interested to learn more? Subscribe to our blog or follow us on LinkedIn.

Did you like this article? Please share with your friends


Save on cloud

The question I get most often from companies considering buying our savings-as-a-service is “How are you different from the other companies claiming they can save me money?”.

And sure, I could explain to you how our algorithms are greater/better/deeper, how we cover literally hundreds of ways to save, how we have such and such reference case, how our guides take you through every step, and on and on and on.

But, at the end – I would be answering the wrong question.

The question that should be asked, is “How do we get savings done?”

Whether solution A could save you 18% and solution B could save you 21%, or tool X has so and so many features, while tool Y has slightly fewer – or more, that really has very little impact on the end result.

Your company most likely already has tools today, if nothing else, most major cloud platforms come with some savings initiative recommendation platform. And yet, we’re still saving every single customer money. A lot of money.

How? Well, I don’t lack a ribbed six-pack abdomen because I don’t know how to do a sit-up. I know. It’s not that I’ve never joined a gym or haven’t bought an ab-roller on a home shopping network.

It’s that I don’t get it done. And spending weeks or months evaluating solutions isn’t going to get cost savings done either.

And that is how we’re different: everything we do is about ensuring you Get. Savings. Done. In one week. In just a few hours in fact.

 

Savings-as-a-Service

  • You buy online, with no risk and a full money-back guarantee.

  • We take you through 3 simple steps:
    1. An onboarding call to plan the next two steps and brief all stakeholders
    2. A data collection call where you get data collection done
    3. A results presentation and implementation workshop where we talk through the activities that need to happen, and we start executing. Right there on the meeting.

  • You celebrate actually getting savings done.

No tool to implement, integrate, secure, and learn how to use.

No team to hire, we’ve got the cost optimization experts and you’ve got the technical staff to implement the savings.

No “behind the scenes” optimization by the same vendor who makes money selling you cloud.

Want more differentiators?

We’re independent: We don’t sell cloud, consultancy, licenses, or indeed anything else, other than savings-as-a-service. We don’t have a conflict of interest between the department that makes money of off selling you the very same cloud another department is selling you a savings analysis of. We aren’t scared of jeopardizing our relationship with Microsoft or Amazon if we put the pedal to the metal in helping you save. Independence allows us to focus on just that one thing – making sure you get savings done. It’s the only way we can make money, and the only way you’re likely to recommend us.

A final one?

We cover every angle. All other solutions out there target either a) DevOps staff focusing on things like rightsizing, b) procurement focusing on reservations, or c) asset management focusing on licensing. We cover all these angles and many more, but we’ll only ever show you the things that actually have a material financial impact and can be done in little time.

That’s our side of the story, but the proof is in the eating. So why not try us? You can even get a free trial – a full featured service, limited to €5.000 of savings. Done in one week, of course.’

Interested to learn more? Subscribe to our blog or follow us on LinkedIn.

Did you like this article? Please share with your friends