I was at a conference for internet retailers two weeks ago and was overwhelmed by the software and hardware solutions promising to solve all their operational problems and turn their ecommerce businesses into a highly profitable, eight figure monsters.

They’re lying.

Technology is not, by itself, the answer. If you have a broken process and you add technology, all you get is a faster (and more expensive) broken process.

Let’s say, for example, that you invest in hand-held scanners in a warehouse. Yes, the system is pricey, but think of the productivity improvement! Think of the increased inventory accuracy! Except, maybe not. 

Consider a client of mine that bought scanners to improve picking and inventory accuracy. That was a great idea, until the pickers decided that it was a hassle to scan 12 boxes individually. Instead, some of them were scanning a box once and typing in 12. . . except that occasionally their fingers were a bit heavy on the button, and they typed in 122 instead. That did wonders for their inventory accuracy. 

Another client of mine invested in a fancy ERP system to help them manage the complexities of their supply chain, and better match supply with demand. One day, they realized that they had 18 months of inventory on-hand for a few products, and it was tying up badly needed working capital. What happened? The software calculated purchases based on customer demand, which was artificially inflated by a bonus program that the VP of sales ran during the normally slow summer months. They were stuck with a pile of closeouts that destroyed their margins for the year.

Even Amazon has run into this problem. They deployed scanners throughout their warehouses, but found that workers weren’t hitting their productivity targets when stowing products on shelves. The problem? Scanner batteries were running down in the middle of the process, forcing them to stop in the middle of putting away products, and walking to the office to find fully charged batteries.

The moral of these stories: your new hardware or software will only deliver the promised results when you combine it with solid processes. If your processes aren’t well-defined, or if they’re not consistently followed, or if they’re not particularly effective, then the new technology won’t deliver the goods.

Here are three steps you should take before investing in new tech:

1. Observe the work.
You think you know how the process works, but are you sure? Most likely, you know how the process is supposed to work. But in the time since you’ve gotten a new phone system, hired new staff, or added a specialized oncology service, I bet that those processes have changed. Your staff has developed shortcuts that you don’t know about, or are dealing with problems you didn’t imagine, and the process doesn’t run the way that it once did. By taking the time to really watch the work that’s being done—order entry, patient discharge, credit checks, etc.—you’ll either see how the new technology should support the work that’s actually being done, or you’ll want to redesign the work to remove the obstacles that will prevent you from reaping the benefits of the technology. I know of one hospital system that avoided buying $200K in new ventilators and wheelchairs—which everyone was dead certain they needed—simply by creating standard work and using 5S to ensure that they were returned to the right place.

2. Address behaviors first. Then add technology.
Technology might actually exacerbate problems if you don’t address bad policies and behaviors first. In the case of the client with the 18 months of slow moving inventory, the ERP system made dysfunctional policies even more harmful. Management let the software make purchasing recommendations based on sophisticated algorithms that didn’t account for the special sales promotions. That resulted in much larger purchases than a human in purchasing would have made. The company should have eliminated the practice of goosing sales with special promotions before rolling out the ERP—or at the very least, built a process that ensured review by the head of sales, finance, and operations to check the software recommendations. In the case of the Amazon stow line, it wasn’t until they created a supporting process to check, reload, and monitor the scanner batteries that the company achieved the promised efficiency gains.

3. Find the root cause of the problem.
Is technology and automation really the answer to the problem? While the siren call of technology is alluring, it may not address the root cause of the problem. If your customer service team task switches between entering customer orders and answering incoming phone calls every three minutes, you’ve got a problem that no Order Management System can fix. The reason the CSRs don’t have time to process orders is the interruptions, not the software they’re using. You’ve got to eliminate the interruptions before you add software.

Technology is not a panacea for your operational ills. But if you follow these three steps, there’s a much greater likelihood that you’ll be able to get real value from the cool technology you buy.