Well, if you divide the business into two kind of sub businesses, right? You've got the Food Service Technology side, which is a big labeling market, right? We have one customer mix that uses a 2x12 label. So as long as that business keeps picking up, the label sales are going to be wonderful because it's a very expensive label, and they use a lot of them. On the restaurant side, Mitch, as we close more restaurants, there, the label sales will be less, but the software will be higher. So it will be a lot more predictable because they're going to eat by contract, they will subscribe to a three, four or five-year contract with us. So there, the software is going to be a lot easy to project because we'll have that by contract. The only real issue that we faced in 2020 was the ups and downs of the economy as things opened up in the summer and then kind of crashed in the winter, we saw the label sales kind of follow that kind of pattern. Now what we're seeing, Mitch, is we've had a very good software quarter. We've had a very good label quarter. And I can tell you that Q2 is starting out much stronger than Q1. So if that continues, we'll have a very good Q2. And now what Steve was trying to explain about the terminals is even if we put a terminal out February 15. We have to count that whole terminal in that annual revenue. But at best, it's going to give us six weeks of software revenue and six weeks of label, even though we're counting in all 12 weeks. So eventually, we have the incremental terminals, let's say, we get to 30,000 or 40,000 terminals. If we add 1,300 in a quarter, it's not going to matter that much. It's just - if you think about it, Mitch, we went from, call it, 5,600, 5,700 terminals at the end of Q4 to 7,000 in Q1. That's a 25% increase. But a lot of those terminals didn't go in January 1, right? So it's just a function of timing. And what we were trying to do, what Steve was trying to do and saying, look, if we just look at the end of Q4 those terminals will have been in full-service for Q1, and we did $1.2 million in recurring revenue and bam, there's your $847 recurring revenue. So as the base of terminals grow, the incremental won't have as big an effect on our ARPU because of this issue of timing. Of when the unit goes into the market and when we start collecting our software, our service and our labels. It's pretty simple.