“Mainframe” doesn’t make many headlines these days. This could be because mainframes are a relatively “traditional” way of hosting large IT infrastructure and data processing. Another reason might be that innovations in the space are starting to slow as enterprises look to the cloud to replace these once-critical components of a business.

Mainframe users face less flexibility and a harder time finding experienced staff

We’ve seen clear signs that mainframes are starting to be phased out. One example is government organizations. In a 2018 survey from the National Association of State Technology Directors, nearly 80% of state central IT authorities from 38 states did not see future demand for mainframe computing power. The primary reasons for maintaining these systems weren’t costs, capacity, reliability, or even security: 95% of respondents reported that their main reason for continued mainframe use was to maintain legacy applications.

But it’s not just state governments reckoning with their mainframe use. Enterprises are facing a specific talent shortage in skilled mainframe workers, with a 23% loss over the past five years and with roughly two thirds of positions going unfilled. The long release cycles can also limit an enterprise’s ability to make swift changes in response to market changes.

Although a mainframe’s isolated nature might be a big bonus for security, this same isolation offers less opportunity for cross-platform collaboration and innovation. If you’re running on legacy tools to manage your mainframe, you could be falling behind other businesses in terms of automation, visibility, and integration.

It may be true that a majority of business operations still take place on an in-house mainframe. But with an increasingly fast development mindset, a lack of talent, and high costs, it’s not just CIOs evaluating the continued use of mainframes. IBM, the world’s leading mainframe provider, launched a “skinny mainframe” in an effort to offer power and security alongside broader cloud and machine learning capabilities.

What’s the tipping point away from a positive cost-benefit of a mainframe?

If you have a well-planned and well-maintained legacy mainframe system, the benefits of an in-house mainframe can sometimes outweigh the maintenance costs. In addition to the high fixed costs of actually obtaining and installing a mainframe, you must also consider capacity issues as your company grows — on top of paying for repeated software and hardware updates. Of course, the more million instructions per second (MIPS) you’re negotiating with a vendor, the more likely you’ll get a better unit cost.

The fixed costs and continuous maintenance can also make mainframes an easy target for budget cuts, even if they’re performing well. Variable pricing isn’t an option when it comes to physical units. So, with the explosion of flexible ”as a service” infrastructure and storage opportunities, you may also have to spend more time explaining an in-house mainframe’s value.

The balance of a mainframe’s costs and benefits depends on what your enterprise will bear. A good place to start is ensuring you’re tracking the total cost of ownership (TCO) of your mainframe and regularly reviewing the time and money going into upkeep. If you’re at the point where you’re seriously looking at shifting away from an entirely in-house mainframe, bringing in an outside perspective can be beneficial.

Remember the cloud comes with costs too

When we’re comparing new and old technologies, it can be easy to forget or discount the initial migration cost and the total lifetime cost of the more advanced service. That’s why we recommend reviewing all your options, including keeping an in-house mainframe system.

It’s crucial to understand the cost structure of cloud services as they relate to mainframe as a replacement. Instead of measuring cost by MIPS, major cloud providers like Amazon and Google bill by the second or minute of resource use. This is also helpful during short-term projects and variable usage times, such as Black Friday.

This type of billing isn’t without its own challenges, but it still has the potential to undercut your long-term costs of mainframe maintenance since cloud services don’t require the purchase of hardware or software updates. The flexibility and accessibility of these services in comparison to mainframe computing centers — not to mention the relief of internal maintenance workloads — should be enough to warrant seriously considering such a change.

Whether you want to keep your mainframe in-house or move to an outsourced or cloud provider, the Windsor Group can help you assess your current position and find the right vendor that suits your needs. Get our guide to in-house and outsourced mainframes to get started.