The enterprise ROI/TCO argument for mobile is getting a lot more interesting

In 2019, executives need to look anew at mobile and figure out what technology displacements make sense. For example, do companies need to buy expensive dedicated barcode scanners?

apple iphone7 cameras
Susie Ochs

In 2019, executives need to look anew at mobile and figure out what technology displacements make sense. For example, how much sense do expensive dedicated barcode scanners make today when employees almost certainly have a smartphone in their pocket with the CPUs and cameras to do a far better — and far, far, far cheaper — job?

Let's set aside for the moment who owns that smartphone, whether it's a device purchased by the company at a volume discount or BYOD-authorized smartphone bought by the employee. Either way, that little computer has some serious number-crunching capabilities — Apple's mobile A12 chip's neural engine, for example, claims to handle 5 trillion operations per second — as well as some impressive photography chops, such as the iPhone's 7-megapixel cameras.

On the Android side, photography capabilities are even better. The Google Pixel 3, for example, boasts an 8-megapixel front camera and a 12.2-megapixel rear camera, along with very compelling night vision capabilities.

Having talked with vendors on both sides of this argument, I have found the points and counterpoints to be ones familiar to anyone who has had a version of this debate in the last decade.

The smartphone side — companies that make apps to sit atop the phone and do the magic — point to the investment dollars (lots of references to maintenance contracts and lots of other added costs) and a clean TCO position, not to mention the added flourishes the software can offer, such as superimposing additional data on the screen, such as inventory numbers right above the price.

The dedicated-device side — from the companies that still make and sell such devices — speaks to ruggedization and speed, arguing that dedicated devices were designed to perform just one task and to perform it well, as well as standing up to the strain of difficult environments, such as those found in loading docks and warehouses.

Presumably, any company executive — presumably the CFO plus IT folk and some operational people — have already made their decision on BYOD versus company-owned, which certainly colors the subsequent ROI/TCO discussion. BYOD brings far lower investment costs, though it also requires higher training costs since the different OSes, models and configurations (not to mention software conflicts with other installed apps) muck things up. The corporate-owned argument goes in the opposite direction on investment and training costs and — for a handful of firms — can even negate the wear-and-tear argument by purchasing hardened devices to distribute to employees.

Another factor to consider: inertia. In short, most (if not just about all) of the large companies that need barcode scanning already have dedicated devices for the task and have had them for a ludicrously long time. Although this will be a fresh decision for startups, for most, the general-purpose mobile device has a substantial burden of proof. General-purpose can't win this argument if it's as good or even a little bit better than single-purpose scanners. To win, it needs to show benefits so strong that it merits a change in training and device handling.

That said, the general-purpose mobile devices (especially BYOD) do seem to have a powerful hardware-cost advantage. Let's drill down. To help, I spoke with a vendor pushing general-purpose mobile devices — with its app installed, of course, for a fee — and a pair of dedicated-device vendors, and let them make their ROI/TCO arguments. These points are relevant for mobile ROI/TCO discussions far beyond barcode scanners, but it seemed a good place to start.

Christian Floerkemeier is CTO and Samuel Mueller is CEO for Scandit, which makes an app that delivers barcode-scanning but arguably takes it to the next level by both superimposing additional information — such as inventory levels — and by claiming to be able to read barcodes in almost any environment.  

This is how Scandit pitches itself as being able to handle difficult lighting and other environmental challenges: "To achieve the same barcode scanning robustness with smartphones [that] dedicated barcode scanning handhelds [do], we leverage the computing resources over modern smartphones and the advances in real-time, on-device machine learning. When the autofocus struggles to adjust due to low light, scene changes or small objects with barcodes in the foreground, we leverage real-time machine learning-based image processing to compute a sharp barcode from a blurry barcode. When the barcode is far away or very small, we use the CPU/GPU resources to do super-resolution image processing. When the scenes are dark and lack contrast, our deblurring techniques can deal with very little contrast in the image. When there is glare in the image, our machine learning approaches localize glare regions in the barcode and we stitch the decodable barcode regions together."

Scandit makes its ROI/TCO argument based on the cost of dedicated hardware plus the cost of required accessories (for things such as charging), along with annual maintenance as well as repair contracts. Against that, Scandit charges vary based on deployment. But Mueller said, "In the case of company-owned smartphones, depending on smartphone type, volume, use case and feature and capability set, customers can typically expect to pay between $50-$300/year per scan-enabled smartphone. In the case of a pure BYOD deployment, annual costs for a scan-enabled enterprise application would typically be between $5-$50/year per device for an unlimited number of scans."

Floerkemeier makes an additional argument, that a general-purpose mobile approach can extend the barcode-scanning capabilities to far more employees. That could make a difference in a retail environment, for example. Today, Floerkemeier says, employees are typically sharing a small number of dedicated scanners, which are often "so clunky it won't even fit into their pockets." This could be useful for a store associate who may only need occasional scan capabilities, but it would be helpful when a customer makes an inquiry.

As for that superimposed data — a.k.a. augmented reality — Floerkemeier makes a compelling case. Beyond inventory data (how many of that SKU are on the shelf, how many are in the backroom, how many are on order and when they will likely be delivered, etc.), he points to superimposing sales data (trying to decide whether this item is worth reordering?) and visually flagging a specific desired item in a sea of similar-looking products on the same shelf. "This is augmented reality that actually provides value," Floerkemeier said.

He also argues that the latest consumer-friendly mobile devices are more than fast enough for industrial and corporate needs. With any iPhone X, Floerkemeier said, "in 150 milliseconds, it can scan 20 packages."

For the other side of this ROI/TCO debate, I spoke with two executives with top-selling vendors of the dedicated barcode scanning devices: Alan Melling, senior director of data capture product management for Zebra Technologies, and Karen Bomber, director of retail strategy for Honeywell.

Melling's case is that consumer-level mobile devices simply won't be able to keep up. "The choice really comes down to, 'How critical is the application to your business?' Simply slapping plastic around an iPhone doesn't make it a rugged device," he said. "Can it read a damaged barcode or a poorly printed barcode? Dedicated devices just do that better [because] they are engineered to do that."

Specialized software such as Scandit says that it can read problematic barcodes, but there's no way which device handles it better without putting both devices through testing in an enterprise's actual environment and see how each does.

Melling also pointed to an iOS and mobile concern that this column addressed last month, namely that IT "can't have the software go away without any notice."

Honeywell's Bomber said she has seen some of her customers trying to use general-purpose mobile devices and "about half have tried and failed using personal devices." She cited one key problem was that the devices ran out of power more often and needed to be recharged.  Also, if the company is using a BYOD approach rather than buying corporate devices in volume, the mix of device types can cause problems far beyond needing more training: "A lot of employees don't have the latest and greatest devices. If I drop it and use it for work, I'd expect the company to buy me a new phone. [Consumer devices] break. It's not ruggedized."

Whether or not a consumer phone will work for any specific company's operations will depend on the environment and needs of that environment. I'm not suggesting that the ROI/TCO will work all of the time, but it is worthwhile for the idea to be examined periodically and testing the latest devices — both dedicated and general-purpose — periodically. Even if the ROI argument doesn't work today, the answer might be quite different six months later. A little extra testing never hurt.

6 tips for scaling up team collaboration tools
Shop Tech Products at Amazon