Larger Merchants Crave Cloud POS – They Just Don’t Know It Yet


Right off the bat I’m going to admit this article will get very definitional. It’s important that we understand the difference between de jure and de facto. De jure is what the law says, whereas de facto is what society has adopted in reality. Let me give you an example: de jure law says jay walking is illegal. In de facto reality, everybody walks across the street when convenient. Got it?

Now: what is cloud point of sale (POS)?

There is no real definition of cloud POS. Novelty of the category is partly to blame, as is the fact that everybody and their dog thinks they have a cloud POS. This is like Big Data all over again…

Maybe we should go back to the beginning: what is a point of sale?

For the purposes of our understanding, a point of sale (POS) is software that captures item-level purchase data from a merchant. Often this is called SKU data – short for stock keeping unit. The software can be wrapped in a number of hardwares, but is fundamentally different from payment terminals, which are separate pieces of hardware running their own software to capture, encrypt, and process payments. Terminals do not capture item-level data, but capture information from the card transaction: customer name, merchant ID, transaction time and gross amount. The point of sale software will see some of the payments data but not all of it. The POS has also grown into the merchant hub, capturing employee clock-ins/out, discounts and other operational data from the business.

When a payments company says they have a mobile POS, or mPOS, they are referring to a mobile terminal, NOT a point of sale. Here’s a screenshot of mPOS according to Global Payments

This device is not going to capture all the data that a true POS will, so it’s a little disingenuous to suggest that this phone dongle can replace one. What this device can do is extend the reach of a terminal – for accepting card payments – to a mobile device.

To confuse the issue further, some people define mPOS as an untethered mobile tablet used to replace traditional POS software. This is just needlessly confounding. mPOS = mobile terminal, NOT point of sale on a tablet.

If we are now in agreement that a POS is not a terminal, we can attempt to define cloud POS.

Cloud POS is marketed as a consumer-grade, mobile tablet that can replace your conventional (legacy) point of sale. It is not mPOS: it’s a POS that runs on a tablet. Cloud POS providers boast that cloud has the features necessary to yank out a conventional POS system without losing sleep: employee management, menu curation, etc. The Apple tablet, as we know, has no connectivity ports and instead syncs via wireless communication to merchant peripherals (printers, scanners, bumpers etc.). This is the image most people conjure when thinking of cloud POS.

Now our personal definition of cloud is anything that communicates above-store to the internet. But that’s not how cloud is being sold. Cloud is being branded as a new hardware and software experience. In fact, you can’t get the software experience without the corresponding, untethered tablet to go with it. I partly blame Square for this, who raised and then dumped hundreds of millions into marketing sleek iPads as point of sale replacements. But the point is, you do NOT need a tablet to be a cloud POS.

The merchants switching to cloud POS today mostly reside in the long tail. That is, they are smaller independent stores (think mom-and-pops) and regional chains. As time progresses we see larger chains – the 20-50 store variety – adopting cloud POS. But when will the large, public retailers make the move?

There are a few things keeping the larger merchants from executing the inevitable upgrades.

First is retraining. Large retailers have tens of thousands of employees. Implementing a new POS system means establishing a training program, which by extension means paying hefty sums in overtime or accepting a drop in operational efficiency as employees learn the ropes.

Second is data. Merchant accounting, operations and marketing is tied to POS data… or at least it should be. We’ve yet to find a merchant run decently well, but that’s not to say POS data isn’t used in their business systems, however crudely. Ripping that out interrupts the data feed and, depending how their systems are architected, could mean a loss of historical transaction data.

Third, and probably most critical, is sophistication. Like EMV, merchants have no idea what cloud POS really is. Rumors start – no doubt perpetrated by their legacy POS providers – and before long it’s gospel. At every conference you hear information that’s worse than false: it’s downright dangerous.

To clear the air on cloud POS, we tapped someone whom everyone should acknowledge as an expert. Bill Draper is the founder and CTO of Gusto POS, a cloud POS system. Prior to founding Gusto, Bill served as the Director of Advanced Enterprise Technologies for Micros.

Gusto was designed to ensure constant operability above store. Bill explains that this means critical data – like menus, printer configurations and pricing – run on a native application in the store. Some systems don’t do this and everything is run in a browser window, which means when the internet goes down, so do your most critical POS features. As Bill says, “POS systems must be designed so operations are always working. The business must be able to accept cash and push orders to a kitchen. If the system cannot do this you’ve asked your customer to take on another liability.”

Gusto’s transaction data is put into a virtual queue and pushed above store (i.e. cloud) when the internet is available. If there’s a latency in internet connectivity, the transactions will simply sit locally while an isolation layer monitors the network until connectivity can be reestablished. Once the systems are in communication again, the transaction data seamlessly syncs to the cloud.

Gusto also uses a peer-to-peer (P2P) technology. The best way to explain it is like this: legacy POS systems were an on-premise, client-server model. The POS data was all on one local machine – termed the “back office” – irrespective of the number of terminals in the store. If the back office crapped out, your data was gone, and you could do nothing but wait for a replacement.

In P2P, the data is shared around all the terminals on the network to avoid a single point of failure. You’ll hear this referred to as “distributed architecture”, but all you need to know is that it’s a night and day difference from what the majority of legacy POS systems offer.

Some legacy systems are trying to catch up, though, so we shouldn’t detract from those efforts. The legacy POS companies create a pseudo distributed architecture with the addition of an above store component. Some legacy POS providers (whose business philosophy is to bilk merchants for everything) charge for these features – think MyMicros. Others are creating the redundancy for free as they realize the architecture is an investment in their business.

Lastly, you hear complaints about performance for cloud POS. Ringing transactions and putting out kitchen tickets takes forever, or sometimes orders don’t happen at all. There are a few possible explanations.

First, it’s possible the software was not robustly tested and code has not been optimized. Bloated code can run infuriatingly slowly, so we understand this grievance. Gusto avoided this by testing heavy loads in 20+ terminal deployments – an ode to the large deployments from Bill’s Micros days.

Second, peripheral connectivity can cause all sorts of headaches. Those using iPads will note that there are no ports for scanners or printers; the POS is communicating to peripherals over bluetooth or wireless protocols. Gusto supports these connections but Bill notes that, “In restaurants, despite everyone’s best efforts, wireless protocols are just reliable enough to be compelling and just unreliable enough to be frustrating.” The takeaway for merchants is that if you want to use wireless peripheral connections, make sure your connection and router are business-grade. I shouldn’t have better connectivity in the bathroom at Starbucks than you have supporting your critical operations.

Merchants are wont to ignore technology since they don’t understand its increasing important to their business. Before long merchants will be using data to better manage labor planning, promotions, theft analytics, menu pricing and a host of other things I won’t bore you with right now.

99% of this data will originate from the POS. It’s time merchants learn the truth about cloud POS, and why they should be taking it seriously. Those that don’t won’t be around to regurgitate the falsehoods from extinct legacy providers.


  • Rob Finley

    Firstly please lets not confuse Cloud POS with any specific device or OS. Cloud POS should not be reliant on any particular device, it should be at home on an IPad running iOS but also then be able to transfer the sale to a standard Windows POS terminal seamlessly. This is what Cybertill has been doing for 15 years albeit in the general retail sector and yes we do run to a browser and yes the terminal still operates without a connection if there is a failure. What larger chains really like is true multi-channel integration, imagine a sale started on line and then collected from store utilizing the store inventory in real time, imagine easy updates and a single view of all customers and stock across all stores and channels again in real time. Oh and of course we have many retailers with not 10 or 20 stores but 50, 100 and even 600+ branches. Cloud POS is the future.

    • Jordan Thaeler

      What happens if the browser window crashes?

  • Wladimiro Bedin

    The Browser is not a pre-requisite for Cloud POS. The App is a better approach for a such mission critical function.
    Here I try to explain why

  • Thomas Benavides

    Firstly please lets not confuse Cloud POS with any specific device or OS. Cloud POS should not be reliant on any particular device. Thanks for the information.