15 Signs You Have A SaaS Metrics Problem (and How to Fix it) with Dave Kellogg

The 3 topics we're going to hit today are as follows. First, and don't be scared by that number, the 15 signs of a strategic SaaS metrics problem. And yes, we're going to spend a lot of time on the problem today. We'll spend some time on the solution, but we're going to spend time on the problem first because it's fun, right? We get to talk about the silly things we do and the silly things we see. And second, because by Understanding the problem, hopefully, that will lead us to a solution. So we're going to go through the 15 signs of a SAS metrics problem. Then I'm going to introduce a SAS metrics maturity model I built. It's a little 5-level, 5-layer model designed to help you go, oh, I think my company is roughly here. We're at level 2. And if we want to get better, we should work towards level 3. And then we'll look at the things you do in each level to try and move the ball forward. So that's our agenda today. Let's jump right in and start talking about the 15 signs of a SaaS metrics problem. The first sign, one of my personal favorites, is what I call bludgeoning, where SaaS metrics are used not as a tool of understanding, right? We're not trying to make the business better. We're not trying to understand our business. We're trying to club people with numbers, right? And boards do this to CEOs. They club. You're behind plan, you're over budget, COGS is too high, right? We can club the CEO. CEOs in turn can pass along the favor, right? They can club the CRO. Quota attainment isn't where it's supposed to be, productivity's too low. And you know it's really a case of bludgeoning when somebody goes, you know, we missed the lead number. And the CMO's like, we haven't counted leads in 2 years. We don't care about leads, we're all about pipeline, right? So you're literally clubbing me just for the sake of clubbing me with a number that isn't actually relevant. To my business. So the first sign that there's a problem is kind of this abusive use of metrics to kind of bludgeon people, to use them as a tool of almost oppression, not as a tool of enlightenment or understanding. Now this can lead to our second problem, which is misleading, where people use SaaS metrics to mislead people in numbers of ways. One might be because they're afraid of being bludgeoned, right? The other might be because they don't have what I would call a functional relationship with the board, and they see the board meeting as something to survive. Right? We don't actually want to have an open, honest conversation with our board and look at the numbers together and try to understand what's happening. We just want to survive that meeting, feed them some BS, and go back to work. Live to fight another day. This is a reason that some people do misleading. Right? There's lots of different motivations behind it, but those are two popular ones I've seen. I'll give you the world's simplest example of misleading. Which is this table. It's a perfectly good table. Nothing wrong with that table. Actual and planned for last quarter. It's a little ARR bridge. Starting plus new minus churn equals ending. But somebody stands up and says, we had a good quarter last quarter. We ended at 96% of plan. How do you feel about that? I feel angry when I see that, right? Because we didn't end at 96%. We used a $12.5 million stabilizer called starting ARR to damp out the fluctuation in plan performance. That's what we actually did by saying we're 96% of plan. The CFO says that, I'm maybe a little angry. The VP of sales says it, I'm super angry, right? Because the CFO gets to talk about results sometimes more than drivers. So, you know, what we should be talking about on this slide is this, that we're 60% of new logo ARR, dramatic problem. We're 93% in expansion, so not bad. 73% on new ARR. We're actually 100% of churn ARR Thank you, customer success. But overall, we're 60% on net new ARR, meaning our burn multiple for people who like that SaaS metric, and it's increasingly popular in this day and age, our burn multiple will be almost double plan, right? Because our net new ARR is almost half plan, right? So this is a kind of semi-catastrophic quarter. And you can simply— and look, the most common form of misleading is what I call cherry picking. You just don't put up the bad numbers, right? You put up different numbers every quarter. That's the most common form. And this is not cherry-picking. This is putting them all up there, but simple through this kind of sleight of hand of just talking about the 96%, that story that we tell about the numbers is kind of a lie. So that's our second sign of a problem, where people use metrics to mislead. The third, déjà vu. So you go to the same board meeting every quarter. You're like, does anything ever change at this company? I feel like we had this conversation last meeting, and I feel like we had this conversation the meeting before. I think we had this conversation the meeting before that. Right? And while this is not a direct symptom of a SaaS metrics problem, I would argue it's an indirect symptom. That the companies I work with that kind of have their metrics act together don't have the déjà vu problem, because the numbers are telling us What's not working? And look, it may be very hard to fix it. Sometimes these problems take quarters to fix, but at least we can say, hey, we're having trouble in mid-market. We tried this new messaging. We thought it would increase conversion rates. It didn't. So we have something to talk about, right? It's not the same board meeting. It may have the same problem, but it's not the same board meeting. So if you have kind of board déjà vu or QBR déjà vu, right? If you're on the E staff, same thing can happen in a QBR. Maybe you have a SaaS metrics problem. Rat-holing. I was pretty famous for this one, personally. I'm vulnerable to it. That you're trying to have a conversation about the business using metrics. That's the goal of this stuff. We want to talk about the business, and we want to use metrics to do that, but sometimes we end up talking about the metrics instead of the business. So the CMO is up, and they say, "Hey, we generated 1,200 MQLs last quarter." Sales feels like they're starving. They're like, "What do you define an MQL as?" Right? You ever been to this meeting? Half an hour later, people are still arguing about it. The CMO is probably gone in 2 weeks, and they were counting everybody who came to the last user conference as an MQL. Right? But the point of the matter is, first, we had a bad definition of MQL, but second, because we didn't have it defined in advance, because we didn't agree to it, Somebody did something stupid and counted the wrong thing, and we spent an hour talking about the metrics instead of the business, and that's a problem. Rat-holing. Next, semantics. Semantics is almost the opposite case, where you think you're having a really good conversation, and then you realize some of the words don't mean anything. Right? Like, for example, here's one of my favorite conversations with the VP of Sales. Hey, what's the forecast for new ARR this quarter? 4.3 million. Oh, okay. What's the most likely outcome? 4.5. What do we have here? Noob alert. Right, like any professional CRO will say 4.3 again, even if it's not mathematically correct. And then I say, oh, okay, so I got him to move once. Well, what are we really going to do? 4.7. I think if I keep asking this guy questions, I'll have him up to 5. With like 2 more questions, I can get him up to 5. So there's two things going on here. One is the kind of noob alert, but the actual problem is in what's worst case, 3.5. If I say, "What does most likely mean? What does best case mean?" I get gobbledygook as answers. Well, my favorite definition of best case, "If the stars align." How often do the stars align? Is that like a frequent occurrence, or is this something, you know, happens once a month, once a year? Right? I don't understand that answer. Right? So what I encourage companies to do, and most CROs hate this, I think more velocity, more quantitative CROs are okay with it, but we're really talking about the bottom edge of a confidence interval. And when I give the forecast to the board, I want to give a 90% confidence interval, so I'm giving them the lower bound of the 90% confidence interval. Now people hate that, and if people get scared when you use those words, I have advice. Just say you get to miss it once every 2.5 years. That's what it means. Right? That's exactly what a 90% confidence means. You get to miss without getting in trouble once every 2.5 years. And by the way, since you missed it last quarter and the quarter before that, you're not giving me a 90% number. Right? So, try and adopt this probabilistic view on forecasting. And more importantly, just try to define what these things mean. Because we have these conversations and some of the most important words are undefined. So lack of semantics. Recalculation. So here's a joke. How do you— what's the difference between a venture capitalist and a private equity investor? There's a lot of them. We could have a lot of fun with this question. Here's my answer. We'll see if we're really metrics geeks. A VC asks you your CAC. A PE tells you your CAC. Very good. Glad we got a laugh on that. But it's true, right? 'Cause the VCs tend to be trusting and say, "What's your CAC and tell me how you calculate it and isn't that interesting?" Whereas the PE people just tend to say, "Give me all your data, I'll tell you what your CAC is." Fundamentally, especially in the context of an acquisition or a fundraising, I don't really trust your numbers, I'll do it myself. So if you find yourself in a situation, like I sit on a couple of boards where one of the board members will say, "This number's that." And I'll be like, "Where'd you get that? I don't see that in the deck." And they'll say, "Oh, I recalculated it all." I have one of my MBAs recalculate everything. I'm like, this is a symptom of a very serious problem, right? We don't trust the management team to present the correct numbers. We don't trust them to do the right formulas, right? We don't trust them to update the templates. If I just say, please add this row and calculate it that way, I can't even trust them for that. So I'm going to use one of my MBAs to go recalculate everything. So this is a problem. Recalculation. Inexplicability, one of my personal favorites. The company's not performing, we're missing ending cash, we're behind on ARR, but when I look at all the other metrics, everything looks fine, right? Top of funnel looks pretty good, the conversion rates look good, the cost per opportunity is okay, stage 1 and stage 2 conversion is okay, stage 2 demo is okay, demo to close is okay. Jesus, the ASPs are constant, average sales cycle isn't increasing. Like, I don't know what's going on. And this does happen. And all I can say is you've got a problem somewhere. If the top 2 important numbers are not being achieved, and you can't see why in all the other numbers, I can't diagnose what the problem is, but I can tell you you have a problem. And it does happen from time to time. Slide 2. Okay, so we're about halfway through our list of metrics— signs of metrics problems. Here we go. Dissidence. The words don't match the music. Right? You ever know like a song that kind of sounds sad, but the lyrics are happy? Or it sounds happy and the lyrics are sad. Like, you know, "Some Charmed Life," upbeat number, very catchy chorus. Doot doot doot, doot doot doot doot. That song's about meth addiction, right? It's all about meth addiction. One of the lines is, "Doing crystal meth will lift you up until you break." Right? And it has all these incredible words in it that if you're not paying attention, it's like, holy shit, we need to look at the lyrics of this song. And that's what some board decks feel like. Right? I was flipping through a board deck with a friend of mine the other day, and it's like, "Oh, we're struggling a little bit on ARR, and we're not quite there on PMF. And, well, but, you know, employee satisfaction is really high, and customer sat's high. Most of the customers are still here, and the bug backlog is down, and story points are up. Burn rate was 7x." And I'm like, "What? Do it, Crystal Method! Lift it up until you break!" The burn rate is— the burn multiple is 7x? We're spending— $7 for every dollar of net new ARR we're adding, right? That's like a catastrophically bad burn multiple, right? And it'll just be mixed into the song and you're kind of not supposed to hear it. And the answer, what's going on here is I think it's actually a form of avoidance that people don't want to actually admit there's a kind of a dead moose problem on the table and they're talking about all the other things and we're just letting this huge problem sit there in the middle. And not address it. So if you have dissonance, I think it's probably caused by avoidance. The next is the numeracy, and this ultimately is a cultural issue. The symptom is very simple. You go to a board meeting or QBR or management meeting, and the only two people who ever answer numbers questions are the CFO and the head of RevOps. Anytime you ask it, and when you ask it, when they're out of the room, "Oh, we've got to wait for Phil. Phil's not here right now. When Phil comes back—" I can tell you that. I can tell you average sales cycle length in mid-market, right? So part of me wonders, well, let's give Phil your job then, because Phil can answer my questions and you can't, right? What value add are you doing here? And the other is, why aren't you numerate? Why don't you understand your business? And ultimately, I think this is a cultural issue. I have a friend who calls CEOs the chief reminding officers. I don't love that. I think of them as chief norms officers. You're setting the norms for this company, and it's either okay or not okay, right, for on a Monday morning marketing not to know this quarter's to-go pipeline coverage. There's only one answer. It's either okay or not okay to know that. And if it's not okay, you need to set that as a norm. And in my experience, the better companies do. Déjà vu. So you go to the same board meeting, and it never changes. Sorry, you guys are really switched on. You caught that. So I'll just— I put it in there mostly for a joke, but I will say I had a friend of mine quit his job and that was the reason. I said, "Ed, why are you quitting this job?" He goes, "Because I've been to the same meeting too many times." And I actually think it's a fantastically good reason to quit a job, right? If you're always going to the same meeting, it means you're not making progress on your problem. So again, there's a difference between not making progress on hard problems that take a long time to fix, which is fine, and literally having the same meeting every time. So we talked about déjà vu all over again. Next, disconnect, where metrics are not tied to strategy. And, you know, this is an example of how a company might present the second quarter metrics. New ARR, 85%. Expansion, 101%. Churn, 18%. New logo, 25%. And I've got a lot of problems with how this is presented. Let's be clear. In that other talk I'll write for a month or two from now, we'll talk about everything that's wrong with how this is presented. But, but the biggest problem I have is imagine this company is in the middle of a mid-market push that we're doing a lot of SMB business. The churn rates are too high. The lifetimes are too low. The ASPs are too low. It's throwing off our CAC ratio. It's throwing off our LTV to CAC. And we think, because we have some customers in mid-market, that if we just retargeted our energy more to mid-market, that we'd do better. And that's one of our top 5 strategic goals for the year. Imagine this company presents that chart. What do you think? You go, I can see the strategy everywhere but in the metrics. Right? As you say, I can see computers everywhere but in the productivity statistics. Right? I can see strategy everywhere but in the metrics. Where are they? And the answer is here. Right? This is what that company should actually present, where we say, you know that mid-market push that is the number one thing we're trying to accomplish? The first metric I'm going to show you is new ARR for mid-market. We're actually 111% of plan. Then I'm going to show you the ASP for new ARR in mid-market, which is 140% of plan. Sorry for the typo. Mid-market sales cycle, we were worried it was going to be long, but it's turning out to be short. It's only 45 days. Right. Mid-market pipeline coverage. Are we generating enough opportunities in the market? Yes, we are. We've got 3.4x. Our win rate. Do we win mid-market deals at the same rate because maybe we have new competitors? Yes, we do. Right. It's actually a better rate if you don't have there, but it's actually, you know, very strong win rate, 35%. So to me, and yes, I put the overall numbers down because they matter. In the end, we are running that business on the bottom. But our strategy, one of our top 5 strategic goals is to improve that business on the top. So disconnect. Next, piecemealing. This is super common, where numbers are evaluated in isolation, not holistically. You know, a VC will look across their portfolio and think of the best CAC, the best CAC payback period, the best LTV to CAC, the best rule of 40, and ask you to do, you know, beat them on all four. It's inherently unfair. It's a form of bludgeoning. Right? Like you want me to be the most musical kid in my high school and the most athletic and the best at math and the champion of the chess club, right? That just doesn't happen. Right? So we need to look at these things holistically. So piecemealing can be dangerous because, for example, given you have a high customer acquisition cost, what do we think about your lifetimes? So if you've got really low churn and a really long lifetime, it might be worth it. Hey, we're getting people who are going to be sticky for 10, 12 years. We can afford a CAC of 1.9. CAC ratio of 1.9, right? I need to look at them together. Given your high growth rate, right? So you're growing at 120%. That's going to drive your CAC up because CAC is a function of growth rate in enterprise software because you're carrying unproductive/ramping resources, right? So we need to think of these things holistically. Think of them in little groups. And piecemealing in the end, when you just take them one by one, you're losing insight and you're probably in all likelihood just bludgeoning. It's a different budgeting technique. Next, misbenchmarking. I love benchmarks. There's lots of great benchmarks out there. I think it's really important to compare ourselves to others. I don't think we should just automatically try to be, you know, 75th percentile in every metric, you know, across the board. That's not wise. These things should be a function of strategy. But I think we should know them. I think we should know what those people do, and we should understand how they work together. And misbenchmarking happens for a couple of reasons. One is some benchmarks are just highly selection biased, right? Whoever answered the survey. Okay. And you may have bootstrap companies in there. You may have PE-backed companies that want to sell in 3 years for 18 times EBITDA. And you may have venture capital moonshots that want to grow at 250% all the way up to $300 million ARR. Right. And you're putting them together in a benchmark and saying the average CAC ratio is 1.4. I was like, hey, wait a minute, this is kind of crazy. And normally you can segment these by size or growth rate. That's good. But I actually think the best way to segment them is by aspiration. That if we could, and nobody offers these yet, but I want you to just consider this concept that we should segment benchmarks by aspiration, which is if I want to sell for 18 times EBITDA in 3 years, boy, wouldn't it be nice if I could compare myself to other companies shooting for that goal? Right? Analogy: if I want to go to Harvard, I should compare my SAT scores not to the average, but to the group of people who apply to Harvard. Right? In this case, Harvard is an IPO. Right? If I want to go public and be an IPO-track company, I can't— to get concrete example— take 104% ARR out of a blended benchmark when that Harvard cohort has 125% ARR. Right? I need to benchmark against people who are trying to do the same things I'm doing. Correlation and causation. I'll just give you a simple example. I think we all know what this is. But for example, the VP of customer success says, all of the customers who churned had multiple escalations. They're having escalations because we don't have enough agents. The agents are too busy. We need to reduce churn and hire agents. That might be correct. But you cannot by any means say that that simple correlation is causation. More likely in this particular example, maybe the product isn't working, and that's why the agents are getting all the calls resulting in all the escalations. And if you want to fix this, right, like ice cream deaths, temperature, and drowning, right, the underlying driver is temperature that drives more swimming and more ice cream consumption. That's why they correlate, right? In this case, the underlying driver is the product's not working. So we should fix that. So we have to be very careful because I watch companies do this. I gave you a super obvious example, but I watch companies do this all the time. They do some deep analysis of churn, and say, "We're going to focus on these 3," and they're acting as if it's causal, and there's no reason whatsoever to believe that. It's correlated, yes. Is it causal? We need to look more. Torture is the last one, so we made it through the list, where people, you know, basically— this is particularly bad when a board member does it, but they arrive with a hypothesis about what the company— what's wrong, and then they do what's called confirmation bias, right? They want to prove themselves right, So they will kind of selectively filter and torture the data in order to do so. And we all know the old expression, but just in case you don't, right, if you torture the data long enough, it will confess to anything, right? So we can prove almost anything with the data if we're determined enough, right? That's not the point. So I'll just give you one crazy example. This company walks in and says, hey, we've got a 5% churn rate. I go, wow, that's amazing. So you have a 20-year lifetime. Which means even if your CAC ratio is like 2, you're going to have an LTV to CAC of like 10+. This is amazing. God, fantastic company. And I say, can I see the data? And they put this up. And go, hey, wait a minute. Wait a minute. Your churn rate is 25%. You lost 500 out of 2 million in ARR. So your net churn rate is 25%. If you exclude expansion, you can check the math on me, the gross churn is 29%. Your logo churn is 5%. You've only lost one logo? But the reason you only lost one logo is what? You renewed 4 customers for a dollar. Right? You renewed 4 customers for a dollar to kind of juice your logo churn rate, to juice your lifetimes, to juice your LTV to CAC. Right? And the question is, so those 4 $1 renewals are hardly the problem. And, you know, do people actually do this, you might wonder. And the answer is, yeah, people do. This is an old Chevron ad, but people do. I competed against a company that did $1 renewals, and they would juice their churn rates so they could offset their SaaS metrics. So that's what I would call an extreme case of kind of twisting SaaS metrics. That's why you need to look at both logo churn and dollar churn. So let's now look at what's going on here. Let's talk about the SaaS metrics maturity model, but there's kind of 5 underlying problems here. One, there's no shared metrics foundation. No definitions, what they mean, how they're calculated. Two, there's no trust in the data calculations and formulas, that the story being told matches the data, and that we aren't using metrics to either bludgeon or mislead. There's no trust. Second root cause. Third, metrics are not integral to strategy. That— and I think this happens because people view metrics as a financing thing. Oh, we need to raise a round. Somebody go calculate all the SaaS metrics. 'Cause they're going to ask about them, right? And you haven't burned them into your day-to-day operations, right? Because they're not linked to your strategy. They may not even be linked to day-to-day ops. It's just something you do for financing. All the VCs are going to ask. The culture is not metrics-driven. We talked about this. Is, you know, is numeracy a prerequisite to have conversations about the business? Is it tolerated if you're not? And then metrics are not being used. This is more of an opportunity cost we haven't talked about. But metrics are not being used to define the trajectory and long-term goals of the business. Because when you get the other 4 taken care of, this is what you do next. You say, hey, wait a minute, what are we trying to accomplish over the next 3 to 5 years? And how's that going to impact our business? Because if we open Europe, that's going to hit our CAC ratio in the short term. Sorry, we're going to be opening a bunch of offices, hiring a bunch of salespeople. They won't be selling anything. That's going to be bad for CAC in next year and the year after. It'll be good for growth. In the years after that, right? Ditto for an R&D investment. Hey, we do something inefficient in the software. If we did a big R&D project, we could make it more efficient, reduce our COGS. It will hit R&D in years 1 and 2 and show up in COGS in year 3 and 4. And to have these conversations, you need kind of a long-term driver-based model, and you need to think about your strategic priorities. So towards this end, I built this SaaS metrics maturity model. Where we have those 5 layers, level— and we're going to talk about each in a minute in more detail. Level 1 is foundation, which is definition, calcs, and semantics. Level 2 is trust. And I'll say it a couple of times here, but templates build trust. Just if you can only get one notion from this entire talk, templates build trust. So we're going to talk about templates, cadence, how you use them, right? And then continuous improvement, how we always make them better. Level 3 is linking metrics to strategy. Level 4 is culture, and Level 5 is trajectory. And I can just say that improving SaaS metrics maturity is a journey. There is even Layer 1. Layer 1 has taken me, with certain companies I work with, quarters. And I walk in every time, I have this kind of entrepreneurial blind spot, going, "This is going to be easy. Me, head of FP&A, CFO, we'll get this nailed in 30 days, 60 days." Every time I do that, And most of the time, 3/4 later, we're still working on it. And it's good, right? I mean, if you ask founders, if you had any idea how hard it would be, would you start your company again? And most say no, right? So I actually view this as the endearing and valuable trait of someone who's entrepreneurial, is we radically misunder— misunderestimate. That's, say, a George Bushism. We radically misestimate the amount of work it is to do something. And it's actually in some ways very helpful, because we get fresh on it every time. So, but it is hard. You can't plan on having this done in a few weeks. The change management is hard. It may require staff upgrades, particularly key analytics roles like RevOps, FP&A. Sometimes those people are not that strategic. They're not that fast, right? They're good at making a budget, and they're good at getting comp plans done, but they're not really good at thinking about the business. Will require meeting changes to cadence and agenda. It requires, I don't want to get all touchy-feely here, but trust and the willingness to be vulnerable. If you're going to put these metrics out there and somebody clubs you to death with them, that's not going to feel real good. Right? So we have to be careful that we make sure the board doesn't club us, and that we don't club the exec staff by looking at them. So, let's now look at each of these 5 layers. One, the foundation. So we've talked about what this means before. Two, well, one, two new things on this slide. One, sometimes this requires the rework of underlying systems. I, most of the companies I've worked with that have this problem in Europe for some reason, but they don't produce a standard SaaS P&L. They produce a nature-based P&L, not a function-based P&L. And you can't do SaaS metrics off a nature-based P&L, right? They would literally go GL account by GL account, map it to sales and marketing. And, you know, once a year they'd calculate their CAC ratio because their P&L was not set up to do it. And boom, that's a 9-month delay. You want to know how it takes 9 months to fix SaaS metrics? Step 1, change your accounting system and your chart of accounts, right? But you have to do it. And there's no time like the present to start. The other thing I'd say on this, and this is again from hard-learned experience, offboard this to metrics committees. There are people like me who care, and rather than trying to do this with the whole E-Staff, or God help you, the whole board, say, I think we should get a committee of interested people who want to meet once or twice a month to work with FP&A, CFO, Ops, to make sure we have the templates you guys want to see in the meeting. And we put people on the committee and we let them work. This is the answer to how to improve your staff metrics. Delegate the problem to people who care. Let them work. Give them time. Have them check back in. Second, build trust. Templates build trust. Templates build trust. Why? Because there's no cherry-picking, and the most common form of misleading is cherry-picking. We just leave stuff out. We don't talk about expansion ARR being so far off plan last quarter. Somehow it just didn't make the slide. And when you have good templates, you can't leave stuff out. So we decide everything that we're going to show. We show it every time. It never changes. And we provide trailing 5 or trailing 9 quarters of history. And we provide context like the plan or the most recent forecast or last year, right? Or last quarter. So it's rich in context and there's nowhere to hide. And I honestly believe that templates build trust. It's the only way I know to build trust is to keep working on the templates and then keep making them better. The other thing is fix data at the source. As long as people are adjusting data out of systems in spreadsheets and putting it back in, no. You must become kind of a fascist about, we will fix the underlying systems, we will fix the transforms, we will get this stuff done in a systematized way. 3, link to strategy. So I love strategy. I love this guy, Richard Rumelt. He's written 2 good books. One's called Good Strategy, Bad Strategy. The other one's called the Crocs. Pick whatever one you want, but the point of all strategy is to identify your biggest challenges and make plans to overcome them. It's that simple. And the thing that operators get wrong is they don't understand, they don't have empathy for the board. So, for example, let's say we're back to that company that's trying to do an upmarket push, and we want ASP to be one of our key results. We're going to do this upmarket push because we think the ASP is going to come up. There's two ways to approach this. One is just make ASP an OKR, and we'll wait till the end of the year to see if it changed. And you're effectively telling the board, "Trust me. I got this. Trust me. It's going to change." And maybe it'll change on a quarterly basis, but that's not fast enough. And by the way, the last three guys, sorry, said, "Trust me." So as the board member, I'm not really in a very trusting mood right now because the last three COOs have said, "Trust me." It's not good enough. If you want me to trust you, give me some leading indicators, right? And I'm going to use a sailing metaphor here. I'm not a big sailor, but I did take sailing lessons. And when a boat tacks, the bow goes through the wind at some point, right? So I want to know in a metric sense, when is the bow going through the wind? There's actually signs before that. The sail starts luffing. There's a thing called telltales on the sails that will tell you you're starting to execute the maneuver. We want telltales. We want luffing. We want Val going through the wind. To make that less nautical and more literal, what if we looked at the ASP of deals in the pipeline? If that was bigger in mid-market, wouldn't that be a leading indicator of closing deals, right, with a higher ASP from mid-market? Yes, it would. Right. That's a leading indicator. That's a telltale. It's not the KR. The KR is to get ASP up for real. But the leading indicator is to look at the ASP of the mid-market segment. In the pipeline, right? So I don't have to just say, trust me. 4, metrics culture. This we've talked about. It's really a choice. And the goal is to not have conversation about metrics, but conversations about the business using metrics. My favorite quote on this, it's an old one, but I love this quote. If we have data, let's look at the data. If all we have is opinions, let's go with mine. And if you say that enough in your staff meeting, maybe people will get the hint, right, and start showing up with data. And then finally, agree on strategic trajectory. We haven't talked as much about this as I would have liked, but I do have a blog post on this called Glideslope Planning. It's the metaphor I use on how we want to take this company to its exit. And it's all about using models to do basically timing, sequencing, convergence. So we're going to do some Q&A. Let me just wrap up here with our last slide before we go to Q&A. But SaaS metrics can be used as tools of enlightenment, oppression, or obfuscation. I've seen all three. We've talked about 15 symptoms of a SaaS metrics problem. We've introduced a 5-layer SaaS metrics maturity model, which you can use to both locate yourself and make your roadmap. And we discussed things you could do at each level to build your SaaS metrics maturity. Thank you.