IBM Refines its Tailored Fit Pricing Model for Z Systems

Executive Summary

In May of 2019, Clabby Analytics wrote a review of IBM’s then new “tailored fit pricing” (TFP) model (see our review here) that shifted IBM Z pricing from a model where customers were being charged for the amount of system capacity used, to a model where customers can instead be charged for the amount of computing resources consumed (consumption-based pricing).  IBM has since refined this “TFP” model with a consumption-based software licensing model with cloud-like pricing for on premise computing – and a hardware-based model that allows customers to access additional capacity when needed without having to buy a fixed amount of capacity to handle workload spikes.

The reason why these refinements are important is that they better provide IBM Z customers with the means to structure their IBM Z software and hardware environments for true consumption-based computing.

Modern workloads

In days of yore, IBM Z executives purchased enough capacity to enable their systems to handle workload spikes.  In order to exploit that capacity, Z administrators focused heavily on balancing workloads such that capacity was available to handle workload spikes – while other workloads could be processed when a system was less busy.  To maximize an investment in a Z system, the whole focus was on keeping a system busy (performing on-line tasks in real time – and pushing batch tasks off to slack periods).  This model was great for highly-predictable environments – but today’s workloads have become highly-unpredictable.  IBM understands this market shift, which is why it introduced TFP pricing several years ago.

The refinements

There are two elements involved in TFP pricing: Tailored Fit Pricing for IBM Z Software; and Tailored fit Pricing for IBM Z Hardware.

TFP for IBM Z Software allows IBM Z customers to pay for computing consumption based upon how much capacity they consume.  Using IBM’s Subcapacity Reporting Tool (SCRT), a system baseline is determined (how much capacity a customer needs over the course of a year).  IBM customers who prefer to pay on an monthly license charge (MLC) basis agree to an annual Millions of Service Units (MSU) utilization rate and a monthly license charge (MLC) rate – and are free to consume the MSUs they have purchased at whatever rate they wish (using a lot of MSUs when they have workload spikes and consuming others at slower periods when the system is less busy).  The key here is that the customer no longer has to purchase capacity headroom to deal with spikes. They consume MSUs when needed.  Also noteworthy are that IBM is happy to sell additional capacity at aggressive (favorable) rates, and IBM allows unused MSUs to be rolled over into a following year.

For IBM software users who have purchased their software (these are One Time Charge – OTC – customers), their investments in software can also be used on a consumption metric across multiple Z 15 and 16 environments – instead of being restricted to capacity used within a specific Z.

TFP for IBM Z Hardware enables IBM customers to create a Z hardware environment with capacity needed to handle most of their needs, while additional capacity is made available by creating a fixed-size “capacity corridor” on top of the client-owned capacity in order to deal with workload spikes.  This environment is available on a subscription basis or on a usage basis.  The subscription model charges a monthly fee per system for activated TFP hardware capacity on full system engines.  The usage model employs a cloud-like usage charge on a per hour basis, charging customers only for actual usage, not full engine capacity.

Nicely done!

The way a computing company “wins” in a given industry is to capture the most workloads possible and by providing access to those workloads at reasonable rates (while also meeting client Quality of Service requirements for security, reliability and performance).  What IBM has done with this latest round of Z pricing addresses this “reasonable rates” requirement.  It enables its Z customers to take full advantage of the Z hardware that they own without having to purchase a lot of extra capacity to deal with peaks.  IBM aligned usage and billing with the actual computing power consumed.  As for the hardware consumption model, customers should see improved and more predictable response times with lower latency, faster transaction processing and optimized workload handling within an agreed-to MSU profile.

Overall, by modifying its pricing program, IBM has enabled its Z system architecture to compete more cost-effectively with competing cloud environments while offering the superior QoS of IBM Z System hardware and software.