by Minda Zetlin

IT budgeting mistakes to avoid

Nov 12, 2018
BudgetingIT LeadershipIT Strategy

Rising costs can sneak up on you. Here’s how to prevent runaway technology spending before it hamstrings your IT operations.

piggy bank on top of a calculator
Credit: Thinkstock

Are you in the midst of digital transformation? Using data to drive new insights? Moving more functions to the cloud? Helping support a wide range of SaaS applications selected and purchased by people outside IT? If you’re like most IT leaders, the answer to most of these questions is likely yes.

Now for the tough one: How are these changes and initiatives affecting your company’s technology spending? With rapid innovation and more digital capabilities than ever before, the opportunities to use technology in exciting new ways are seemingly endless — and so are the demands on IT leaders’ time and attention. Which means it’s easier than ever for simple mistakes to send IT costs spiraling upwards.

Here are four common pitfalls when it comes to making IT budgeting decisions.

Mistake No. 1: Moving to the cloud without adapting to the cloud

“Many folks that I see find cloud is costing them more over time than on-premise,” says James Anderson, senior director and analyst at Gartner. “Cloud itself is not a cost-optimization strategy. It can be a component of a cost-optimization strategy.”

Many organizations are moving to the cloud or adopting a “cloud first” strategy, motivated by a host of benefits that go beyond cutting costs. But, experts say, if you’re moving to the cloud to reduce spending — or just don’t want to wind up spending more than before the move — put careful thought into how you do it.

“Sometimes an organization will have a simple mandate that ‘We need to move to the cloud,’” says Mark Sami, vice president of Microsoft and cloud solutions at digital transformation firm SPR. “If I’m a traditional IT operations service manager, I’ll just take all my servers and move them to the cloud piece by piece.” But that approach, commonly called “lift and shift” can be bad for the bottom line. “It would likely cost most organizations more money than staying in the data center, because you’re not taking advantage of what the cloud was created for, which is the ability to scale consumption up and down according to need.” That dynamic has led some companies that have moved to the cloud to reverse course and migrate at least some of their workloads back to their own data centers, he adds.

But if you do it right, an infrastructure-as-a-service (IaaS) deployment can help cut costs and increase availability, especially if you have widely varying capacity needs. That’s how it worked for one of Sami’s clients that manages shopping malls. “Literally one week of the year, from Black Friday on, is when their servers get 90 percent of their traffic, whereas every other week out of the year, they could use a much smaller footprint,” he says. That company moved its infrastructure to the cloud and was able to vastly increase its capacity for that single week, and reduce it the rest of the year, saving costs over its data center. “Users never noticed,” Sami says. “And it was the first time they didn’t have an outage during Black Friday.”

But there’s a catch. For most organizations, taking advantage of the cloud’s scaling capabilities means making changes to applications. “In the data center, you’re running 6 CPUs of capacity because it’s hard to provision more — you’d have to buy it and justify it,” says Michael Cantor, CIO of data center hardware maintenance provider Park Place Technologies. “In the cloud, you spool that up and down. But you probably will not have the skills in-house to do that kind of tuning.”

Mistake No. 2: Deploying SaaS without proper planning or negotiation

For many companies, moving to software-as-a-service (SaaS) applications can make more sense than a lift and shift to the cloud. Moreover, software vendors are increasingly putting more resources into the SaaS versions of their applications than the licensed ones. Because of this, SaaS customers get updates and improvements more frequently — without requiring update work from their IT staff.

But do SaaS applications cost less than on-premises ones? It depends how you look at it, says Yvonne Scott, CIO at accounting and consulting company Crowe. “If I had five network engineers managing servers, I don’t need them if I’m in the cloud,” she says. “But you know that’s built into the cost. Apples to apples, is that cheaper? I’m not sure. But I don’t have to manage those people — they do.” SaaS can provide meaningful cost savings to smaller organizations with a very small IT staff, she adds, because in effect, it allows them to hire a fraction of a person.

Another difference between SaaS and licensed software is that monthly subscription fees count as operating expenses rather than capital expenses. Depending on how your organization is structured and how it reports its financial performance, this could be a very big deal. “Some organizations will never move to the cloud because it’s better for their balance sheets to have capital expenditures versus operating expenditures — even though there could be overall cost savings,” Sami says.

To keep SaaS costs from getting out of control, put some time and thought into your contract negotiations, experts advise. “What we’re finding is it’s really, really important to make sure you can lock as low a price increase as you can,” says Brad Whitehall, CIO of the multinational uniform supplier UniFirst. “If you can lock in two or three years of zero increases, that’s good. If you can lock in 2 or 3 percent, or tie it to the consumer price index for a few years after that, that’s good.” It’s also smart to negotiate up front what will happen if you decide to leave a SaaS provider, Scott says. “What will it cost to get your data, and how much time do you have to get it?”

Cantor says he starts over from zero every year when it comes to negotiating contracts. “I verify: If we paid for 80 users last year but only had 60, let me negotiate that down.” With longer-term contracts, he sometimes renegotiates before the contract is up. “You have a three-year deal, you get one year into it and say, ‘If you knock this price down, I’ll sign another three-year deal.’” Many vendors will agree, he says. “They get a four-year deal.”

Though it’s not fun to think about, a move to SaaS can create future cost problems in case of a downturn in your business or in the economy at large. “The move to cloud has been hitting since the last big economic downturn,” Scott says. In the past, she adds, IT was an easy place to do cost-cutting. “In 2008, what a lot of people did was they decided to stop paying for the maintenance service on their licensed software. It was painful, and eventually I’m creating technical debt, but I can wait to pay down that debt when the company is in a healthier financial situation.”

With SaaS, of course, you can’t choose to stop paying unless you’re also willing to stop using the software. “That’s the conversation I’m having with my CEO as we’re migrating more and more things to SaaS,” she says. “There’s not much we can do about it, but just be aware as we get to the next downturn, your flexibility will be limited.”

Mistake No. 3: Not working with business leaders to ensure tech spend aligns with business needs

These days, when employees can buy SaaS applications on a credit card, and when vendors often bypass IT and go straight to business users, it can be tough for IT leaders to know about, let alone control, all the technology spending going on in their organizations. The CIOs interviewed for this story all acknowledged that there was likely at least some shadow IT in their companies. But they all do their best to stay on top of all tech spending so they can vet, support, or perhaps offer an alternative to anything business users buy on their own.

“Regardless of whether it’s in IT’s budget or not, it’s the CIO’s responsibility to make sure all the money is going to the right place at the right time,” Cantor says. “The traditional thing is, ‘I don’t like what my IT department offers so I’m going to get something else.’ Managing that spend, consolidating it quickly and getting the best pricing is important.”

As Cantor notes, there are plenty of tracking tools that will tell you where network traffic is going and alert you to, say, an unsanctioned AWS deployment. But the most effective strategies for managing non-IT tech spend all involve working with the leaders of other departments and lines of business (LOB) to control shadow IT, while making sure IT can provide or at least support anything users really need.

“We work with the financial organization so when people try to expense something that’s out of policy it gets rejected,” Scott says. “If my accounting department says, ‘That’s nice you bought that, we appreciate it but we’re not giving you your money back,’ that gets around pretty quickly.”

But, she adds, “We also have a mechanism for people telling us what they need. We say, ‘Here’s the product we use. If it doesn’t work for you, come back and tell us about it.’ Ninety-nine percent of the time they never come back.”

Mistake No. 4: Paying for technology that doesn’t add value

Getting rid of anything that doesn’t bring value to your organization sounds like a simple thing to do but most IT leaders know it’s one of the biggest challenges of their jobs. Prior decisions lead to technical debt. Mergers leave you with redundant and/or incompatible systems. And user preferences sometimes trump IT best practices.

Still, the ongoing effort to streamline, consolidate, and eliminate anything non-value-adding is a battle well worth fighting, CIOs say. Michael Reagin is CIO at Sentara Healthcare, a not-for-profit regional healthcare organization that operates 12 hospitals, 300 care centers, and a health plan, among other services. Given the goal to reduce tech spending by 20 percent, Sentara’s IT team launched a multi-pronged program that includes breaking down IT silos and training up functional specialists who can perform the same tasks across different technology functions. Sentara is also part way through a move to become 100 percent cloud-based in two years.

But a big part of the effort is orchestration of IT and relationship management with the lines of business, Reagin says. “We have IT people who live in the field with our [internal] customers. They help create the value plan with the customer point of view and help us understand value. Everything we do, we attach a value statement to it, whether it’s a full-on project or minor work request, and that makes its way into our system.” Combined, he says, these efforts have led to a 10 percent productivity gain.

Another big initiative is simplifying the organization’s application portfolio. Sentara had 1,600 applications when the IT team started this process. It now has 1,000 and Reagin’s goal is to get to 800. “We asked, Is there something we have on a different system that could do 80 percent of this job? A lot of applications didn’t make sense and we were able to sunset them or migrate them.”

The secret, he says, is to keep asking the question over time, since software providers are continually adding functions once only available as third-party plug-ins. For example, Sentara’s electronic medical records (EMR) system turned out to have 100 bolted-on applications when the IT team reviewed it. “Five, six, seven years ago, those functions weren’t available within the EMR, but the core system vendors over time catch up,” Reagin says. The review found 70 of the add-on functions were now duplicated within the EMR system Sentara already had. “Most organizations aren’t good at going out and asking, ‘Is this still innovative, or is it baked into the core system now?’” he says. “We did a lot of work with our operations team and our executive team.”

He also uses incentives to get users to give up their cherished older applications. “We say that when we buy a new system, we want to sunset a minimum of two old ones,” he says. “For the most part, that’s been successful.”

Beyond redundant systems, Scott also evaluates whether the business process the system supports is still relevant. Time tracking, for example. “Because we’re a professional services firm, people still track time for certain purposes,” she says. “You need to keep track of hours for auditing. But we used to send a weekly report to people’s bosses that told them who hadn’t filled  out their time because we needed that information to bill people. We could keep doing that, but most of our billing isn’t by the hour anymore.”

There’s no value now in sending out that automated report every week because most managers won’t do anything in response, she says. “If no decision is being made or no action follows it, why are we still doing it?”

For reasons like this, she says it’s important to regularly review roles and activities and ask if they provide value. “People take on activities and tasks because they’re meaningful at some point in time. Rarely do they say, ‘Let’s stop doing this.’”