In this episode of The Orbit Shift Podcast, we sit down with SaaS veteran Dave Kellogg. Dave is an advisor, independent board member, and consultant for various SaaS companies. He is the author of the popular blog Kellblog and has 10+ years experience at each of the CEO, CMO, and board levels across ten different companies ranging in size from $0 to over $1B in revenues.
We spoke to Dave about a topic that leaves many confused – SaaS metrics. Dave speaks to us about why ARR (Annual Recurring Revenue) and ARR growth are the two most essential SaaS metrics, why founders should pick a big enough problem to solve. When to prioritize Net Dollar Retention over Customer Acquisition Cost, why Dave does not put too much faith in Net Promoter Scores, and why startups shouldn’t be metric slaves.
Edited Excerpts of the podcast
Q. How do you look at ARR (Annual Recurring Revenue) and its relation to the valuation of a company?
Dave: There are two good questions in there. Let me get the implicit one first. If you’re a million-dollar size ARR SaaS company, one of the big questions you should consider is- are we MRR focused or ARR focused?
A lot of people think ARR is MRR (Monthly Recurring Revenue) multiplied by 12. But put it this way, if you do annual contracts that are prepaid 12 months in advance, and that’s all of your business, I don’t understand why you track MRR. Conversely, if you do month-to-month contracts where people renew every month, I don’t know why you track ARR because you’re just taking MRR and multiplying it by 12. In some ways, it’s fake because you have a chance to churn every month.
If you’re at a million dollars, now’s the time to ask, what’s the natural cadence of our business and how should we think about and track our metrics? And I would say if you’re an enterprise, there’s a strong argument to lean towards ARR and annual contracts. I know with product-led growth, you may get a certain number of trials and stuff. But in general, if you think your steady-state business model will come from lots of annual renewing contracts, be ARR-focused. That’s the first piece of advice.
The second question is- why if I could only know two numbers to value a SaaS business, would I pick ARR and ARR growth? The answer is because that’s the way most investors look at SaaS companies. Particularly private company investors because they get to see ARR. Public company investors have to look at revenue, which is a trailing result of ARR. But a private company investor just asks how much ARR do you have?
The typical formula for the value of a software company is just ARR times growth multiple equals value, and the growth multiple is a function of how fast your growth is. So if you’re growing at 150%, year over year, you might be worth 15 times revenues. If you’re growing at 10% a year, you might be worth four times revenues. Basically, the faster you’re growing, the higher the multiple, the multiple of what? – ARR, so one time, the other equal your value.
Q. What else is a companies valuation dependent on?
Dave: Having a big TAM (Total Available Market) is an important factor. Companies that are seen as having large TAMs can have higher multiples. There’s also a big premium for being the market leader in space.
I’ll pick one example in the enterprise software space. At one point in time, the leader was worth 5 billion, the second company 1 billion, and the third company 200 million. So the first place company was worth five times the second-place company, which was worth five times the third-place company. Those are massive winner take all returns in markets, and Silicon Valley companies are often valued in that way.
In the example I just gave you, number one was worth 25 times number three; that’s what we call leadership premium. There are many different metaphors; Jeffrey Moore calls them gorillas, chimps, and monkeys, and you could use other metaphors. But the basic idea is that the way to get the highest value as a SaaS business is to grow quickly in a space that’s seen as large and be the leader.
And let me say large enough because it’s going to be hard if you pick too big a space because then you won’t be the leader. And that’s really the art of being a small startup. If you say we’re going to go tackle the database market, it’s a $50 billion market. Everyone’s going to be like you’re, nobody in the database market. But just to pick an example of a company I know, say you’re Neo4j, and you say we’re going to be the leader in graph databases. That company has just raised $2 billion of an (I guess) $100 million ARR. You get that premium because it’s seen as a big market, and you’ve seen as the leader.
Q. You place a lot of importance on NDR (Net Dollar Retention). Why is that?
Dave: Let me do a quick basic of SaaS. You can think of a SaaS company as a leaky bucket of ARR. We have this bucket. It’s got water in it. Water is ARR. Every quarter sales pour more water into the bucket. Every quarter, customer success tries to prevent water from leaking out of the bucket, and the change in the bucket’s water level is ARR growth.
To use the metaphor, if I wanted to know how much your company was worth — the first-order valuation — I’d ask how high the bucket is and how fast is the level is going up? What’s your ARR, and what’s your ARR growth? That’s the first set of metrics.
The next set of metrics is what does it cost to pour water in the bucket? Which is the CAC (Customer Acquisition Cost) ratio. And the second ratio is what happens to water once it’s in the bucket? Does it all leak out because you might be able to acquire customers really efficiently, but if they hate your offering, you’re going to have a fantastic CAC ratio but enormous churn and that business isn’t worth anything.
The opposite is much better; to pay a little extra to get a customer, but you hold on to them forever. Or even better than holding on to them, making them grow. This is what NDR is about.
There are several ways to measure this leakage rate or the leakage to expansion rate in the bucket. But in my opinion, and certainly, most investors are focused on NDR, sometimes also known as NRR (Net Revenue Retention), NDR says, okay, of all the customers we signed in the first quarter of 2019, take that cohort of customers, what was their ARR worth then, and then what’s their ARR worth now? Let’s divide the latter by the former. And that’s going to give us a one-year NDR rate.
Q. What’s a good way to look at NDR for early-stage companies? Anything below 100 is bad, but what’s a good thing?
Dave: We’d like the bucket to hold its water level. So 100%, NDR means once I pour some ARR in, it just stays there. I agree with you that below the 100 mark, people get nervous. Even if there are other people below 100, they get nervous.
I always think of the investors. If you say 120, they’ll say, great. If you say 160, they’ll be grabbing their checkbook. If you say 95, they’re going to go ‘let me ask some more questions. And if you say 80 or 85, they’re going to say, ‘it was really nice meeting you, I got to go.’ That’s the way I correlate SaaS metrics. Do they grab their checkbook? Do they go yes? Do they say I have a few questions, or do they say nice meeting you? Those are really the four ways to think about SaaS metrics.
The other thing we got to talk about here is that there are two issues. One is strategic, and one is tactical. The strategic issue is if you’re a million-dollar company, as an investor, I care about the CAC ratio. You’ve got a million dollars in ARR, which means you’ve proven some degree of product-market fit. You’ve built something that people are willing to spend money on.
Now the next question is, how many more can you get? I care a lot more about customer acquisition cost than I care about net revenue retention. It’s tough to measure the cost of sales in that first million, that might be the founder personally doing an amazing job selling 20-50K customers, and only the founder can sell it. That’s not a scalable model. What I’d be looking for would be. Do you have professional salespeople selling this stuff? What is their sales productivity? Ultimately, what is your customer acquisition cost ratio?
Then I start saying, Okay, what happens to those customers? You’ve proven that you’ve built something useful. You’ve proven that you can acquire them in a cost-efficient way. Let’s talk about NDR. Do they expand or not? Product Led Growth changes some of that outlook. But I like this systematic view.
Q. Talk to us about NPS (Net Promoter Score). Why is it important? When should you start tracking it? And what are some of the pros and cons of tracking NPS?
Dave: Net Promoter Score, most people know what it is. You basically ask people on a scale of 1 to 10 to say, Would you recommend this to a friend or colleague? One of the best tweets I’ve ever seen was from a person on Twitter, who got an NPS survey from Microsoft, and the response was, do you people actually believe that friends recommend operating systems to friends?
If you think about it mathematically, it’s what I’d call a tail amplifier. You’ve got this distribution. And rather than just get the median or the mean, the mean might be 8.2. Well, nobody gets excited about a move from 8.2 to 8.4. With NPS, by doing this amplification, you can get an NPS score of plus 60 or minus 40. It amplifies the difference between an 8.2 and an 8.4. In some ways, it’s a math trick on the distribution to make people pay attention. Now, the people behind NPS argued that they did market research and that people who give 9s and 10s actually tell their friends in the categories they studied. And people who give 3s and 4s actually do tell their friends not to use it. So they would argue there is a linkage between those scores and the behaviors.
As a metrics person, I didn’t particularly like it. I was anti-NPS because I felt like it was manipulating the distribution. There was very little proof that in software, at least the people would go recommend something. But the reason I’m not anti-NPS anymore is because now everyone uses it. So you have benchmarks. That’s the great thing about metrics. In general, you can know how you’re doing relative to other people.
The thing about NPS that gets people confused is they think it’s a predictor of renewal. You might give me a 10, and I put you down in the renew column. But if you just got acquired by a bigger company, and they use someone else’s planning system, you could be my happiest customer, and you’re not going to be renewing. Or, if you fill in the survey and then submit your resignation letter, you’re not going to be there to sign the renewal in two months. So I always say there’s a loose coupling between NPS and renewal. Happy customers sometimes don’t renew. Unhappy customers sometimes do for years.
The way I tried to get around this is whenever I ask the NPS question, I ask the question I actually care about after, which is- do you intend to renew? And I make sure I know their persona. And if it’s the buyer persona, and the answer’s no, that’s a big fire alarm.
Q. Talk to us about the culture of metrics. What are some of the best practices, and what are some things you should avoid while tracking metrics?
Dave: I have four kids. The average American family has 2.5 kids. I’m not aspiring to have 2.5 kids. It’s useful to know that I have an above-average family size. But it doesn’t automatically become my goal. First, in that case, because it’s nonsensical, literally. And second, it just gives me a way of understanding the world around me. I view metrics like an airplane cockpit. They’re just showing me the values. They’re telling me a lot about the state we’re in.
As a strategist, if you told me your average on every SaaS metric, my first answer would be that you have no strategy. For example, say you have a product-led growth strategy. I would expect to see a very low customer acquisition cost ratio, a very high net dollar expansion rate, and a very large investment in R&D. That’s what I’d expect. So for me, it’s about a coherent narrative.
If you told me you were an old school enterprise software company displacing Oracle with an enterprise sales model and a fairly standard product, but better customer service, I would expect to see high sales and marketing cost, I would expect to see high sales productivity, and I’d expect to see relatively low R&D invest.
Every founder wants to believe that they’re selling because they have the best product, but I’ve worked at companies where we won in the market, and we did not have the best product. People were buying because of our sales and marketing. In that case, if you’re really good at sales and marketing, you shouldn’t go take money out of sales and marketing and put it in R&D. You should double down on sales and marketing and then go acquire more products to sell.
Have a coherent narrative and be honest with yourself about what you’re doing well. Is it great innovation? Is it great customer service? Because your metrics are a reflection of your strategy. If I give you a benchmark report and your average on every metric, that’s a source of concern for me. That’s not a source of pride.
Q. What’s your view on churn? What’s healthy? Any benchmarks that you can talk to us about?
Dave: In terms of benchmark, the new way to measure churn is gross dollar retention. It’s the same cohort-based metric as net dollar retention, but you don’t include expansion. For example, if I had a cohort of 100 units a year later, it’s worth 87. Therefore, my gross dollar retention is 87%, also known as my churn was 13%.
An interesting point is churn classification, regretted versus non-regretted, or ICP (Ideal Customer Profile) versus non-ICP. You hear a lot of these words. Once again, I get very nervous when I hear those words. You don’t get the chance to ex post facto non-regret, which is the number one tendency. ‘Oh, they’re a bad customer anyway.’ If you’re going to classify churn on this basis, beware of this behavior.
If you have a rigorous ICP, you could look at ICP churn versus non-ICP churn. But that ICP has to be objective, not subjective. Regretted versus non-regretted, I don’t like as a term, I’d rather just say ICP versus non-ICP, because then you’re saying, Hey, we’re going to sign up some opportunistic customers, non-ICP customers, but we really do care less about their churn, because the purpose of signing up those customers is to maybe find a new ICP. Maybe we’ll find that our stuff is really useful for retailers, and we never knew it, and maybe next year, retail will be in the ICP, but we’re going to try 10 of these markets, and if one hits, we’ll move it. I like the spirit of experimentation without hurting you. Because otherwise, it’s going to hurt your overall churn.
Controllable versus uncontrollable churn. I do not like it because the classic example of uncontrollable churn is the company got acquired. Sometimes when a company gets acquired, the parent company says you need to use this new system. But sometimes they don’t. And sometimes they say, Hey, do you guys like what you’re using? And the people say it’s okay. And they say, well, if it’s just okay, then use ours, and they go, okay. Is that controllable or uncontrollable? It’s those nuances that make me nervous, and that would make any savvy investor nervous when talking about it.
Q. Do you take into account all marketing spend in CAC or only cost for campaigns that worked? Do you exclude the experimental campaigns from CAC?
Dave: In general, I don’t exclude anything. A lot of this is really about trust with investors. If you show up and say we only spent $1 on marketing, oh, just kidding, we spent two, but half of it was experimental. That undermines the trust.
Using trust as the first principle, I would always show everything in my CAC. And then I might do different cuts of the CAC after. For example, just say I was trying to sell in two markets, telecom and finance, and I had telecom salespeople and finance salespeople and telecom marketing and finance marketing. I could delineate it clearly. I would show the blended CAC. It would not look very good. And I would then say telecom CAC 8.3, finance CAC 1.2. We’re getting rid of the telecom division because it didn’t work. So our blended CAC doesn’t look very good, but the future is all about this finance CAC. In that way, you can tell the story you want to tell, which is very important because we do try stuff that doesn’t work.
Q. What’s a good way to follow your work?
