2017: The Year First-Gen Cloud POS Starts Imploding?


We were going to publish this article at a later date, but with Revel already announcing the departure of their CEO after a Welsh Carson buyout we wanted to have something on record quickly so we could say, “told you so.”

When Revel was pushing an exit to IBM last August, we thought to share our reaction. Given how the deal was leaked leads us to believe it was Revel (ie Revel’s investors) who were looking for an exit.

But as that deal fell apart Revel’s largest investor, private equity firm Welsh Carson, likely acquired the remaining shares of the business. This was seemingly an acquiring transaction given the nature of the firm, but what percent is technically unknown.

This has us pondering what the state of first generation cloud ISVs looks like now, and if 2017 might be the year that sees the first implosion…

Revel has 30,000 POS terminals in-market. At an average of 2.5 terminals per store, that comes out nicely to 12K merchants. Let’s break down what this means for recurring and non-recurring revenue.

Revel’s current pricing is $1,749 for the first terminal and $1,649 for each terminal thereafter. The software licensing is $119/mo for the first terminal and $70/mo for each terminal thereafter (courtesy of POS Advice). If Revel ended 2016 with 30,000 terminals with a growth rate of 137%, math shows recurring revenue of $25MM for the software and $45MM of non-recurring on hardware for the year.

This hardware revenue number seems high. While I can’t support this with data, I’ve heard that Revel lowers hardware prices – sometimes below cost, like Toast – to win deals. The devil is in the details, like this comment about their Shell deal:

In September, Revel announced Shell Retail, which counts more than 43,000 global fueling stations in its fold, will use its POS system at certain locations.

With this recurring revenue, and speculating that Revel is earning maybe 10% margin on hardware – Revel has a serious cash problem. With 500+ employees at $100K/year annual costs plus a large number of offshore contractors, Revel is burning through > $20MM a year. If you start including churn and hardware discounts, the numbers could be much higher.

Welsh Carson – or any investor for that matter – is not going to stomach that for long. It’s really hard to dump an asset that’s losing that much cash unless you come up with a strategic epiphany and find the right buyer.

Our take?

Revel will see a 50% downsizing in 2017. Expect Welsh Carson to work on reaching profitability even if that means a pullback in growth.

We can’t talk about Revel without talking about Toast.

Our last analysis has Toast in a precarious place in 2017, with no more than 18 months of cash left in the bank. The interesting thing here is that Toast’s CEO mentioned they would be hiring an additional 200 employees in 2017, and Toast has 100 open jobs on their website.

How can one reconcile this?

When a company raises money, they don’t get it all at once. Usually there are performance triggers that catalyze a release of more funds. So if Toast raised $30MM, they might get $10MM on day one, and need to hit X in installs or processing revenue to release the next traunch of funds.

If Toast’s management is actively hiring this many people, they’re either crazy (which I doubt), or their investors are thinking this is a winner-take-all market. If the milestones were met to release all $30MM of their Series B, their investors might have told Toast’s leadership another round of funding would be right around the corner. “I don’t care what the economics look like – just grow market share!”

For the record, Google Ventures, who is an investor in Toast, took a similar approach with Flatiron Health, handing the founders $130MM to acquire healthcare data at any cost in a race to develop cancer treatments.

Of course, there’s always the industry gossip that Toast could be shopping itself. Micros built Simphony to offer a cloud product in the enterprise segment of the market but NCR has done nothing for their larger accounts; they’ve built NCR Silver for smaller merchants but have yet to propose a reasonable solution for more complex merchants.

Could Toast fill that void? It’s not as if NCR doesn’t have a long history of acquiring POS solutions.

Our take?

If you think Toast doesn’t get another round of funding, they cut their workforce by 50%+ in 2017. If they’re getting an investor lifeline, expect continued hiring. 

Lastly, we need to talk about Shopkeep.

2015 figures have Shopkeep processing $6B on 18,000 merchants. This means the average Shopkeep merchant does $333K/year. The end of 2016 showed Shopkeep with 23,000 merchants. At the same rate/merchant, this would be ~ $7.7B in processing.

So what does Shopkeep do in revenue?

Shopkeep charges $69/mo for each register. That means Shopkeep has a recurring revenue of $19MM on software. They also charge between $1,100 and $1,700 for their hardware, but like Revel and Toast that is pretty much at-cost, if not outright subsidized to win deals.

Shopkeep also has a credit card business that earns residuals. They charge 2.1% on credit and 0.75% on debit. Unlike Toast, merchants have a choice in their processor, so only some Shopkeep merchants use them for processing. Our guess is that Shopkeep earns the standard 0.10% on each swipe. If 100% of their merchants used Shopkeep for processing, that’s another $7.7MM in revenue… but this number is more likely ~$2MM since our guess is 25% of merchants use Shopkeep for processing.

If you add these numbers up, Shopkeep could be doing $21MM in ARR with anther $15MM a year in non-recurring hardware revenue, though possibly at a loss like the other cloud ISVs.

Unlike the other two, however, Shopkeep appears to have their personnel costs in relative order. On Linkedin there are 250 employees who have attached their profiles to Shopkeep. At an average annual cost of $100K per employee, Shopkeep might only be losing a few million a year. This is a far cry from where Revel and Toast find themselves.

Unfortunately for Shopkeep’s investors, this is also a far cry from an IPO (which requires $100MM ARR), and since investors have had money in for a long time by venture standards, they’re going to need to make some adjustments.

If Shopkeep went from 18,000 installs to 23,000 over 18 months (which appears to be the case based on the dates of their releases), that’s a compound annual growth rate of 16%. That’s not spectacular considering there’s likely no EBITDA, and “good” companies need growth + EBITDA margin = 40%. Shopkeep’s investors will need to find 24% EBITDA margins to find a ready buyer.

Our take?

Shopkeep will slim employee ranks by >25% to meet EBITDA requirements for an exit. But this might not happen until 2018. 

No doubt there will be many legacy providers cheering some of this news. But I have a feeling cloud POS, whether first generation or otherwise, is going to have the last laugh.