POS Has A Hardware Addiction And It’s Crippling Its Future


We’ve spent a lot of time with hedge funds lately, combing through filings of public POS companies. One thing that surprises everyone is how much revenue contribution still comes from POS hardware. Here’s a view of Micros’ revenue breakdown from 2014 as an example.

As shown, hardware contributed 25% to Micros’ revenue even though it had a COGS of 63% (which, by the way, is a ridiculous premium for hardware when we see that the TV industry hasn’t turned a profit since 2011.)

Meanwhile, Micros’ software comprised only 12.5% of their revenue even though its COGS were a measly 10%- i.e. a 90% profit margin!

Here’s another example. NCR’s last full year of reporting showed revenue of $6.37B. Yet looking up their share price this morning shows that the market only thinks NCR is worth $3.9B. Thats right: even though software companies get a 5x revenue multiple, NCR gets 60% revenue discount because it’s focused on hardware. 

Given the inevitable trend of commoditized hardware and a life preserver in higher margin software, you’d think more POS companies would be pursuing software aggressively, wouldn’t you?

Well, this is classic innovator’s dilemma: if something is making a lot of money now it’s hard to look past it and invest in the future. This puts POS companies in a tight spot as the future of the industry becomes even less about commoditized hardware and even more about the guest experience created from software. In this impending market, today’s POS hardware is nearly eradicated.

Let’s first approach this from the front of house (FOH), or customer experience.

Customers are increasingly demanding convenience. This usually entails paying a little more to get things faster – see the rise of online ordering and ecommerce as the poster children for this trend. As these sales channels represent more volume for brick and mortar merchants, let’s think about what’s needed to support them from a hardware perspective.

Right now you can order online from a restaurant or retailer on your own device. Your order is sent over the internet to a merchant’s system. If it’s a large chain, that order should be routed to a main data center and then forwarded to the specific store for execution. It’s very reasonable to believe that at least half of our purchases will begin on our own devices over the next 10-20 years. Fundamentally this means the hardware in-store will be used less, thus making a traditional POS register more expensive on a per-transaction basis.

Who has to see the order at the store-level? Not a FOH employee. Only the entity fulfilling the order needs to see the order. Thus there’s no need for FOH hardware to service online orders.

Okay, but what about in-store customers? How are they serviced?

Well, we haven’t talked about the big breakthrough coming with voice and natural language processing (NLP). In this environment a customer need only speak what they want and the order will be sent to the back of house (BOH) for fulfillment. (Eventually NLP will be replaced with telepathy, where a guest need only think of a solution and an order will be made… as a stepping stone, think of making orders with a augmented reality (AR) headset that tracks your eye movements for ordering confirmation). In this scenario a FOH employee doesn’t need to key anything into a POS either: all that’s needed is a device to record the guest’s voice (or brainwaves). In higher-end environments there might be a person with a mobile device to save you from the first world hassle of making the order on your own device, but you’re going to pay for that white glove service.

If you doubt this inevitability just look at recent kiosk implementations. These kiosks are nothing more than a convenient way to collect information directly from a customer. It’s just that our current technology interfaces are fairly limited. When the technology evolves, a faster and more accurate way to capture an order will be with interfaces that don’t require a customer to deal with a merchant’s unfamiliar kiosk. You can talk faster than you can type, and you can think faster than you can talk.

In these scenarios a POS company is not providing any of the familiar FOH hardware that they do today: it’s all being replaced by more sophisticated data sensors that gather customer information while the real value is in the software that converts the data into an order.

If you’re convinced that this has no effect on the back of house (BOH), think again.

When a future order gets sent to the BOH, what’s looking at it? In today’s environment it might be a warehouse clerk pulling something off the shelf, or a line cook firing off an appetizer. But warehouse clerks are being replaced by robots who don’t sleep, will work holidays, and aren’t striking over health insurance. Cooks are slowly being replaced with robots too, though these robots have limited abilities in the manufacturing facility that is the kitchen. Eventually though? Robots win because they’re cheaper, faster, and more reliable.

So this means that customer order doesn’t need to be printed off. There’s no screen for order fulfillment or entree firing. There’s no bumper system. All that the BOH needs is data. And like the FOH, that data can reside in a far-off server; as long as there’s good internet connectivity or cell network fail over, there aren’t any pieces of hardware that resemble today’s POS stack.

Here’s our (very short) list of needed hardware looking forward:

  • Wifi network
  • In-store order capture. Voice recognition, brainwaves, other novel interfaces. No onsite server needed!
  • In-store pricing screens (this finally allows for yield management)
  • BOH robotics

Everything else is done via software.

Do POS/payments companies end up buying these new hardware components? Do they own the robots and the in-store data collection hardware? Will they be capable of building the software glue that holds this ecosystem together? Does this just become an as-a-service licensing play for hardware since it’s so specialized?

We’re ultimately talking about entities that have really struggled putting out quality product, so we just don’t see them developing this technology organically. Maybe they buy these hardware innovators at a later date, but the merchant’s hardware needs are so different from how the POS functions today the incumbents might be entirely disrupted by a new entity that has this figured out. And as we said earlier, POS/payments companies are making so much on hardware they’re not incentivized to think about the next generation of innovation.

We’ll have to see what the future of POS holds. If today’s POS companies don’t figure out how to break their addiction to overpriced hardware, someone else will hold that intervention without them.

  • I swear tablet POS companies , while cheaper on hardware, are not providing anything near the setup services or networking that small businesses provide. Salesfolk are incented primarily on software because it’s MRR, while support staff gets to the customer after the sales and… oops, Mr Merchant, your sh*tty combo modem router or wifi your stealing from your neighbor is not enough to run POS hardware. So sorry, here’s a link to $500 more in hardware and … what’s that, you’re churning, and want to return everything?

    …months later, with the sales team unable to actually pause to sell hardware, the CS team getting beaten in the face over churn that’s rightfully the sales team’s fault, CS is looking to eat a bullet due to stress, and the company has fired at least 3 execs in CS who’re treated like redheaded stepchildren anyhow.

    • JT

      Totally believable and happening more than we care to admit. People are just gunning for sales number and don’t care about the merchant after the fact.