The point of sale (POS) markets are awash in money. This makes the fundamental business model really tough to nail down. Should a POS cost $100 per month or are venture dollars underwriting it? Is that POS really only worth $3,500 or are payments dollars making up the shortfall?
This is where the market finds itself, and why it’s so difficult to really understand the economics at play.
We thought to take our hatchet to the business model of POS to see if we could give a fair estimate for what a POS should cost. We’re going to lay some ground rules before we get started to make the math as clear as possible.
First, we’re going to assume the POS company takes zero payments revenue. We would concede that just about every POS company is getting some sort of payments kickback these days… but that’s because they have to: hardware, services, and other traditional POS revenues have fallen so steeply they can’t survive on the legacy revenue streams.
Second, we’re going to assume the POS company takes zero outside capital. Outside investors influence the pricing model a POS company employs, and how aggressively they must earn that money back. We think too much outside capital can pervert this process at worst, and obfuscate the fundamentals at best. So VC dollars are out.
With that, let’s get started.
How does one start a POS? If you’re into that sort of punishment you most certainly take one of two approaches:
- A handful of engineers get together and dedicate themselves to making a POS, usually in an echo chamber without real world feedback. This gets to be pricey because the feedback loops for iteration are slow. A positive would be if the engineers had industry experience, or a business person on staff that could socialize the offering with potential customers to get responses.
- A customer wants a custom POS and they pay you to build it.
The cost models for these two approaches are obviously different. The cost model for #1 is higher because you have a large outlay of engineering salaries (theoretically) before you have any money coming in the door. How much? Well, a barebones POS could be built by 2-3 engineers over 12 months. You’re not walking into McDonald’s after a year but you are able to support your local food truck (i.e. you’ve built yourself a Clover). Let’s say a fully-loaded engineer costs $150,000 per year. With 3 engineers, that’s an investment of $450,000 to start with.
For #2, a good business person will get all of the expenses covered. They might agree to drop engineering rates to win the RFP, but at least there’s a revenue stream. We must be cognizant that the type of merchant who can afford a custom POS is usually one requiring more robust features, which means the POS would take 2-3 years to build. At the engineering rates from #1, that’s a cost of $450,000 x 3, or $1,350,000. Maybe you get this whole cost covered but we bet you’re going to take a 50% discount to win the contract. So you now have a $675,000 shortfall over three years.
In either case, you’ve now got to sell your POS once you have something worthy of selling. We’re going to guess the engineers are not going to pick up the phone to do sales, so you have another two options.
- Hire a sales person to drum up deals
- Use a reseller channel
Our stance on the reseller channel is not very optimistic. Yes, there are some great resellers out there, but at least 95% of them wouldn’t be mistaken as a reliable partner. There are clearly modern cloud POS companies who have built businesses on the reseller model but it’s taken them roughly 6 years to do it. At that rate, even at a constant 3 engineers, that’s $2.7M in investment. Not exactly pocket change.
So that leaves you with only one option: hire a sales person. This sales person will need to not only do sales, but all the support as well. We never said it would be a glorified sales position. This person probably costs $100,000 a year with base and commission and should be culled from the ranks of the payments industry (payments people can sell much better than POS people).
It’s reasonable that the sales person closes two accounts per week for a total of 100 deals per year (rounding down).
Let’s stop for a minute and look at our investment so far.
Under the first option we had an investment of $450,000 for our 3 engineers. Then we hired our sales person for $100,000 and racked up another year with our engineers. This brings the total investment after 2 years to $1M, with 100 small, paying customers (assume 1 terminal per location).
Under the second option we’ve got a POS that works in larger merchants but we had a huge investment of $675,000 over the first 3 years. We now need to add another $450,000 for our engineers, and $100,000 for our sales person. This bring the total at the end of year 4 (because you took 3 years to build the POS for the first customer) to $1.225M. However, you have a great reference customer so instead of 100 small customers you have 200 medium customers (assume 2 terminals per location).
Now we need to think about recouping our costs. How much should we charge for our POS?
Well, under the first option we need to recoup $1M over 100 small customers. This means each customer would need to pay us $10,000 that first year to get to breakeven. Seems a little ridiculous for small merchant, no?
Under the second option we need to recoup $1.225M over 200 medium customers, which gets us to $6,125 per customer. This is closer, but still seems high.
At this point we’re going to bring in an excel sheet. These are the quick assumptions: 20% annual churn in the customer base, the sales person keeps adding the name number of customers per year based upon scenario #1 (100) or scenario #2 (200), and your costs stay the same (i.e. support costs don’t balloon, which is kind of a ridiculous assumption to be honest).
Breakeven Price is what we need to look at. Under Model 1, knowing you want to breakeven in 5 years, each merchant needs to pay $2,770 in annual fees to the POS company. If you wanted to earn a 20% profit just add 20% to the $2,770 figure ($554) and bill it to your customers annually.
Now we need to dissect this one step further.
Under Model 1 this POS company is targeting small merchants with 1 terminal. This means churn is undoubtedly higher (above 30% annually) and the customer cheaper. Getting $2,770 annually for 5 years from a food truck is going to be really difficult. You might make 20% on their one-time hardware purchase (probably $200 net profit), but the rest has to come from software/services.
How do you do it?
You move upmarket where merchants spend more on POS, you get an initial customer to fund your development costs so you’re not deeply out of pocket, or you do what most POS companies do: tap payments revenues.
We could spend forever modeling assumptions but you should see the economics of POS systems. And if you look at Model 1 you see precisely why so many cloud POS companies can’t make any money: their merchants are too small, their churn too great, and their recovery costs too high.
Moving upmarket to serve larger merchants takes time. Doing this organically is also really expensive. Now you know why we laugh when POS companies think they can tackle POS and every possible bolt-on they can conceive.
How about you make your POS viable before proving that you can fail at multiple things, mmkay?
Add comment