“Reducing costs” is the corporate version of dieting. Every expert has a secret formula to share with you. Don’t eat carbs. Fast for 16 hours a day. Eat animals. Don’t eat animals. Consume nothing but maple syrup, lemon juice, and cayenne pepper for 10 days (seriously, people do that). And the list goes on and on. The diet gurus tend to bill their formula as “the one” right way, not one of many ways.
Let’s take the opposite approach with reducing IT costs. There are many ways, and I encourage you to be skeptical of each one. I, nor anyone else offering you tips in an article, know your situation. I’m CEO of SysAid, a provider of IT service management (ITSM) technology, so that colors my lens. Principally, I see how 10,000 customers approach costs in the context of ITSM. I’m also conscious of how SysAid’s spending affects team performance, the quality of our product, and the experience we offer.
Two principles underlie the advice below. First, just as you must eat food to lose weight, you must spend some money to spend less on IT. We will not starve your IT department. Second, the modern IT department is an innovator, not a ‘cost center.’ What good is cutting $10,000 in IT expenses if doing so kills $100,000 in opportunities? I will not offer suggestions that undermine your effectiveness and value.
With that in mind, let’s talk about five ways to cut costs:
1. The cloud, of course
The only companies that genuinely want to own data centers are in the data center business. If you still have on-premises infrastructure, consider these questions:
- What software licenses do you pay for annually to have on-premises infrastructure?
- What do you spend on installation, maintenance, staffing, monitoring, security, etc.?
- What’s the payoff?
If you can explain why you create more value by having your own data center versus using software-as-a-service (SaaS), fine. Keep doing it. If your industry is subject to laws that forbid using cloud infrastructure, there’s no choice. But, if you’re like the 99.9% of companies I’ve worked with, inertia probably keeps you on-premises.
Moving data and applications to the cloud will save money. To quantify how much, the folks at TSO Logic, an analytics service for data centers, analyzed 50,000 operating system instances. They found that 45 percent of the instances would run 36 percent more cost-effectively on the cloud. The companies they studied had an average $33 million of on-premises computing costs per year, so moving those instances would save $5.6 million annually. In cases that involve re-platforming legacy applications, migrating to the cloud would lead to savings within five years.
2. Automation and machine learning
One of the best ways to reduce IT service costs is to prevent incidents and service requests. Earlier this year, my team sourced IT service data from more than 10,000 IT departments in 140 countries using SysAid’s IT Benchmark tool. We crunched 86.5 million service requests recorded between 2010 and 2016 to answer a question: Could machine learning aid IT service?
A surprising number of requests take human time and don’t need to. Consider that:
- About 7.8 million of those 86.5 million requests were about granting permissions for printers, shared folders, IPs, and other elements. 4.3 million were from new employees.
- 10.4 million requests were to refresh IT equipment with degraded performance.
- 15.6 million were reset-domain-password requests. In about two-thirds of those cases, an expired password caused the request.
Let’s pretend each password reset consumed five minutes, and the IT admin makes $50 an hour. That would mean password resets cost $64.8 million in labor hours.
Machine learning could reduce or eliminate those costs. An algorithm could, for instance, analyze data on failed attempts, permissions by job roles, and cubicle locations to personalize permissions for new hires and existing employees. Another algorithm could crunch data on usage patterns, error logs, monitoring events, and refresh tasks to measure the average lifespan of printers and other IT equipment. That way, IT could refresh devices before they prevent employees from doing work.
As for password resets, machine learning would be overkill. Try writing a script that forces users to update passwords if they come within 12 hours of expiration.
DevOps has a reputation for increasing the speed and quality of software releases. I would argue that DevOps can reduce IT costs as well.
Consider Infrastructure-as-a-Service (IaaS), a pillar of DevOps. It automates the process of creating new development and test environments, thereby reducing labor costs. IaaS dynamically adjusts the computing resources used, meaning you can run more servers during usage spikes but not pay extra during slow hours. If you swap containers for traditional virtual machines (VMs), you’ll squeeze even more computing resources out of fewer servers. [Note: IaaS applies to companies that offer cloud services; general cloud computing and SaaS, mentioned above, apply to any company].
Introducing test automation in a continuous integration and continuous delivery (CI/CD) pipeline is another saver. Functional testing takes roughly two to three hours when performed manually, and the automation script for that task doesn’t require frequent updates. The investment in automation would cut labor hours.
The DevOps savings do not come instantly. A full DevOps ‘transformation’ can take one to three years for small businesses and three to seven years for large companies.
4. DIY service
One IT admin can address one incident at a time. One knowledge base article might solve a recurring incident for hundreds or thousands of users. Self-service clearly is more cost-efficient – if people use it.
Millennials tend to troubleshoot their IT problems. They Google error messages and browse support forums to avoid human contact. Gen Xers and Baby Boomers are more likely to submit service requests.
Therefore, DIY service only pays off if you invest in high-quality knowledge base articles and direct users to them. Notice that the “Support” page on most websites leads visitors to search bars and FAQs rather than contact information. Note, too, that many companies lower the costs of knowledge base content by publishing service conversations as articles.
Send your IT pros to tradeshows, trainings, and seminars. Conversations with real people, in person, are crucial because they invite context. If you tell a group of IT people that you’re trying to reduce costs, someone will ask questions. What are your top three costs? What KPIs are you trying to optimize for? What does your service workflow look like? And so on. These conversations can uncover unnecessary expenses that are so obvious you couldn’t see them.
I can’t predict with any accuracy how much you’ll ‘save’ from investing in your employees. There’s no guarantee that they’ll meet the magic vendor you’ve always needed.
Rather than obsess over ROI, remember this story. One day, a CEO goes to the HR manager and says, “Why do we invest in training when everyone leaves so quickly?”
The HR manager replies, “Imagine if we didn’t invest in these people, and they stayed.”
From cost to value
The demand for an article about “ways to reduce IT costs” speaks to IT’s reputation. I think non-IT people distinguish between the innovation team that invents products and the support team that makes them work. Perhaps CEOs and CFOs think we can slash operational and support costs without any effect on innovation, the moneymaker.
We must reframe IT costs as deliberate spending, not dead weight. To borrow the dieting analogy again, someone who believes that the only purpose of an eating regimen is to lose weight has a myopic view of food. What about using a diet to increase energy, prevent chronic sickness, or fuel athletic pursuits? We can apply similar questions to IT costs. Rather than view every IT expense as a cost to be slashed, let’s emphasize the purpose and value of our spending.
Sarah Lahav, CEO at SysAid Technologies
Image Credit: Bbernard / Shutterstock