As companies accelerate transformation initiatives that will increase resiliency and efficiency, IaaS spending on Microsoft Azure is growing. Meanwhile, IT and IT sourcing leaders are under pressure to streamline IT costs and reprioritize business-critical initiatives. How can Microsoft’s enterprise customers better manage Azure costs and get more value out of their investment?
Microsoft’s Azure business continues to grow at a steady clip. The vendor continues to report quarter-over-quarter revenue growth in the 50 percent ballpark as the market share gap between Azure and market leader AWS narrows, with Azure now capturing more than 20 percent of the market.
The volatility of 2020 and 2021 continues to benefit Microsoft’s Azure business and its customers – a storyline echoed by Microsoft CEO Satya Nadella: "We have seen two years' worth of digital transformation in two months. From remote teamwork and to sales and customer service to critical cloud infrastructure and security, we are working alongside customers every day to help them stay open for business in a world of remote everything.”
This draws attention to a growing concern among IT and IT sourcing leaders. As enterprises accelerate digital transformation initiatives that enable agility and resiliency, cloud spending – particularly, IaaS spend – is growing in lockstep at a time when many businesses are under pressure to reduce IT costs. How can customers better manage Azure costs and get more value out of their investment?
THE CHALLENGES OF AZURE SPEND MANAGEMENT
It’s important to acknowledge the challenges associated with managing and forecasting Azure costs. At the top of the list is the significant technical expertise required. While governance protocols can certainly curb Azure overspending to a degree, other culprits for bloated spend include imprecise understanding of infrastructure requirements, unattached disc storage, stranded assets, older-generation VMs, etc. Furthermore, Microsoft doesn’t provide deep visibility into line-item and component-based pricing, which makes it difficult to fully understand what you’re paying for – much less manage if you’re seriously consuming Azure’s cloud services.
Second, it’s tough to establish discount leverage, even for enterprises with sizeable Azure spend. A good rule of thumb is you need to spend at least $5M on Azure in order to get Microsoft’s attention. But even if your consumption reaches that level, the efficacy of discounts dwindles over the long term.
While it is possible to negotiate an above-standard discount with Microsoft, it tends to have a less meaningful impact on costs (especially in comparison to an EA discount) – it’s a bit of Moore’s-law-meets-constant-SKU-invention. Azure capabilities and SKUs are constantly changing and so are customer requirements. An offering may be faster/newer/better six months from now, and what you choose to buy today may not be enough (or may be overkill) by that time. In Microsoft’s current evolution of solution versus customer requirements, there is high potential for misalignment – and that often starts out of the gate. Enterprises regularly under-subscribe or over-subscribe to services which can negate and/or minimize the impact of discounts.
Finally, for the reasons explained above, it’s difficult for customers to price-protect themselves when buying Azure. The old mechanisms of locking in price/discounts simply don’t apply because the solution and requirement landscape is evolving too quickly. It’s also worth pointing out that as the cloud compute market becomes more commoditized more vendors are taking a COGS-driven approach to pricing versus the land-and-expand approaches of years past. It’s one more reason Microsoft, like AWS and others, is rarely motivated to cut margins.
There are numerous challenges associated with managing and forecasting Azure costs. It requires significant technical expertise, the efficacy and availability of discounts are compromised, and it’s difficult for customers to price-protect themselves.
HOW ENTERPRISES CAN REDUCE MICROSFT AZURE COSTS AND GET MORE OUT OF THERE OVERALL MICROSOFT INVESTMENT
Accurately sizing workloads and eliminating idle spend are the two most effective ways to rein in IaaS costs with any vendor, and the foundation upon which strong IaaS cost management is built. Beyond that are several other tactics to help enterprises derive maximum value from their Azure spend and lower their IaaS costs:
There are hundreds of Azure SKUs to choose from and, similar to other areas of Microsoft’s business, it’s difficult to determine the right choice for your unique technical and business requirements. Choosing the best-fit option can eliminate costly over/underutilization and toxic spend.
Microsoft typically won’t discount Azure costs in an EA unless the Azure spend is well above a certain (and very high) threshold. With that in mind, enterprises should focus on getting the best deal possible on their EA. Performing IT price benchmark analysis, license/subscription optimization and optimization of cost-related business terms can yield significant savings on your EA, which can offset a lack of discount on Azure.
Microsoft now allows some customers to burn 100 percent of their monetary commitment through purchasing via its Azure Marketplace. For certain customers, it will accelerate commitment consumption thereby helping Microsoft lock in more customer spend over the long term. Also, while solution vendors typically offer competitive pricing on marketplaces, there is no guarantee this pricing is within fair market value range. As such, NPI strongly suggests clients perform price benchmark analysis on large marketplace purchases to ensure they’re receiving the best possible deal.
Enterprises that consider themselves a Microsoft shop may default to Azure, but Google and AWS may meet requirements for less cost. It’s important to conduct a thorough competitive analysis to understand how competitors differ (and are similar), how they handle privacy/sensitive data, and their ability to meet certain software/networking requirements. Some vendors have historically demonstrated more willingness to customize customer agreements and even build data centers to customer specifications while others do not. In some cases, Microsoft may exhibit more flexibility if a valid, demonstrable competitive threat is in play.
Advancements in AI have produced a host of tools/software that provide data-driven recommendations and predictions to help you determine idle assets that may be draining your IaaS budget as well as when it makes sense to modernize and move applications to the cloud. In addition to providing cost savings and avoidance, these tools help companies accelerate their digital transformation initiatives and lower risks associated with cloud migration.
One way Microsoft is capturing more Azure revenue per customer is through its Cloud Economics Assessment. While presented to customers as a way to help them plan migration of workloads to the cloud, the outcomes of these engagements are designed to benefit Microsoft’s bottom line in several ways – including overselling compute and storage resources based on suboptimal scenario modeling.
Another way is by using the assessment to sniff out licensing noncompliance, which can lead to vendor-side leverage to overcommit to Azure spend. These engagements are Microsoft-funded, facilitated by Microsoft’s own tools, include an automated discovery process to gain a clear picture of your Microsoft environment and collect license/usage data under the auspices of cloud migration planning. NPI advises clients to approach Microsoft’s Cloud Economics Assessments with caution. Consult a non-partisan compliance analysis expert to proactively identify and remediate any license compliance issues before providing Microsoft access to your deployment environment, and then to validate Microsoft’s findings if any purported issues are identified.
One other area to find savings that can contribute to reducing the cost of your Microsoft estate is to periodically conduct M365 license optimization assessments. Like any large SaaS footprint, M365 administration is challenging and it takes a formal, user-level inspection to identify license currency that can be reclaimed and overpowered license assignments that should be tuned. NPI recently assisted a F500 CPG company with this and identified $4M in annual savings – it can make a meaningful contribution to overall Microsoft cost reduction.