Lightspeed's Nnamdi Iregbulem: What is Weighted ACV?s

What is Weighted ACV? We came across this metric in an article that Nnamdi penned, Introducing a New and Improved SaaS Metric: Weighted ACV. It immediately grabbed our attention and we knew he’d be an excellent guest on the Metric Stack podcast to take a deep dive into this.

Nnamdi Iregbulem is a Partner at Lightspeed. He’s a coder, economist, and venture investor. Nnamdi recognized that SaaS companies need a different monetization metric and with that, weighted ACV was born.

In this episode, Allan and Lauren sit down with Nnamdi to get an inside look at this metric: What is it? How do you measure it? What’s the difference between ACV and weighted ACV? Where does weighted ACV come into play compared to net dollar retention and other revenue metrics?

If you’re short on time, here are a few takeaways:

• Weighted ACV is the weight of the revenue that a customer represents—if a customer is 5% of your overall revenue, they get a 5% weight in the calculation, whereas someone who is half a percent gets a half percent weight.
• Often, revenue distribution is across different contract sizes. If you have a handful of customers that represent a very large share of your overall business, say 70% of your revenue is from 20% of your customers, you can see where your revenue is concentrated. 
• Use weighted ACV to pinpoint where the revenue risks are in your business and where growth is likely to come from. 
• You should calculate both standard ACV and weighted ACV as a measure of revenue concentration and to tell the full story, picking one or the other omits valuable information for your company and your investors.
• Weighted ACV is an important metric to have ready to talk through with your investors. For early stage companies, if you only show your standard ACV metric it can dilute your ACV because you have smaller customers or free usage.

Allan: Welcome back to the Metric Stack podcast. Today we're joined by Nnamdi Iregbulem, a partner at Lightspeed Ventures. Nnamdi is a coder, economist and venture investor with an interest in technical tools for technical people. He works with companies like Vectorized, Materialize and Snorkel.

Previously, he was an investor with ICONIQ Capital where he backed companies such as GitLab, Fastly, Alteryx, and Epic Games. I'm joined by my colleague, Lauren Thibodeau and my name is Allan Wille. 

Lauren: So today is all about a new metric for many folks: Weighted ACV. We'll ask you lots about that Nnamdi, but before we do, could you just kind of set the stage a little bit? What's the context that we should have in mind as we're sort of having this discussion?

Nnamdi: The context of thinking about is sort of a typical software company selling to businesses that range in scale, from small businesses to mid-market to enterprise. And as you're selling to these businesses, you end up with pretty different contract sizes across these different organizations, putting on use cases, depending on your pricing model, et cetera. And so it's very normal to have a sort of distribution of different contract sizes and investors will frequently and operators calculate the average of all of that. While a simple average can tell you a lot it doesn't tell you everything. And so, the question is, is it more information in the data that you can kind of extract, as an operator or an investor, that could sort of tell you something a little bit different?

Allan: And you're definitely right about that. I mean, these simple averages, we always sort of caution people that averages lie, you've heard that, everybody's heard that time and time again, but yet people do keep coming back to these simple averages. So tell us briefly, what makes it weighted, how are we calculating this? And how do you get that more accurate view of where your distribution lies? 

Nnamdi: So one point to start off with, people often miss, is that it's normal kind of traditional average is also a weighted average. It's just that the weights are all the same. And so we don't think of it as a weighted average is thinking of that as the average, so to speak, but it is a way to, and it's just every way, it's the same.

And so I like to think about the standard ACV metric as being effective with customer weighted metric, wherever you weigh the customer. And since every customer is just one customer, everybody gets the same weight. What's different about weighted ACV is that, the weights are, instead of it just being one for every customer, the weights are the revenue that that customer represents. So if a customer is 5% of your overall revenue for that customer, we get a 5% weight in this weighted ACV calculation, and someone else who is, half a percent, we get half a percent, and so on and so forth. You take those percentages, you will divide by the ACV that that customer represents. And then you add those all up and that gives you the weighted ACV. 

“What’s different about Weighted ACV is that the weights are the revenue that that customer represents. So if a customer is 5% of your overall revenue, we get a 5% weight in this ACV calculation. And someone who is half a percent, we get a half a percent and so on.”

Lauren: That's fantastic. And can you just help us apply that? How do you personally use this metric and how could operators use this metric? 

Nnamdi: I think the right way to think about it is, a sort of analogy that I think about that  mentioned in the piece, which is sort of thinking about the center of gravity. Which is different than the literal center of an object, or if an object has different density across it, then its center of gravity or center of mass will be different than its actual physical center, geometrically. And so you can think about it in the same way for your business.

“Think about the center of gravity which is different than the literal center of an object, or if an object has a different density across it, then its center of gravity or mass will be different. You can think about it in the same way for your business. The standard ACV calculation is sort of the geometric center of your business.”

You know, the standard ACV calculation is sort of the geometric center of your business. It's like where the typical customer is. The way to ACV is telling you, where is the typical revenue you'll come from in our business, which can be very different than the customer centric view. And so in enterprise software, it's very common to have this sort of distribution across different contract sizes. And so you can have a couple of customers that represent a very large share of our overall business. I've done an analysis in a different essay on public companies, and it's not uncommon for public software companies for 70% of the revenue coming from roughly 20% of their customers. And so if you were facing a similar distribution, you can use this, that sort of points to where that revenue tends to be concentrated in what kind of customer. And so an example might be maybe by the standard ACV calculation, then you have, let's say 5,000 is like a standard, average kind of typical contract.

You may have a couple of customers that are much larger than that. And so when you'd run the weighted ACV calculation, maybe it's closer to 30K in which case that tells you that the typical dollar in your company is actually coming from a 30K revenue customer. Not a 5k revenue customer. And so while most of your customers might be smaller, most of your revenue is coming from the bigger folks.

And so if you're thinking about where's the revenue risk in our business, whereas growth likely to come from, et cetera, you can do a better job kind of pinpointing that, using the weighted version of the metric.

Allan: So you often hear about those stories where people will say, 80% of the revenue is coming from 20% of your customers.

And I think that's exactly what we're talking about here. We're really trying to identify, not by just the pure count of customers in your portfolio, but really where is that weighted money coming from? Where is that concentration coming from and are all businesses equal in that sense, in that, if we run an analysis of ACV and we run an analysis of weighted ACV, we're going to see a difference or are there some businesses that are actually going to have the same ACV as weighted ACV and are those businesses different? Are there different stages? Where does this really start expressing itself?

Nnamdi: It tends to express itself most when your pricing model scales with, either in the realm of seat based pricing or true consumption-based pricing, that's what you tend to see the widest sort of differences in contract values. If you're on the other hand, if you're selling a product that has this sort of one price for it, and you pay like on a monthly basis, and there's no way to pay more for it or less for it, that's just the price.

Then you're not going to see much of a variance across different customers. And so in that, your weighted ACV and your standard ACV might be exactly the same number or very, very close to one another. In most software businesses, you do see this to the revenue concentration, which leads to weighted ACV being different than the standard ACV.

“In fact, the difference between weighted ACV and standard ACV is itself a measure of concentration in your company. It’s a nice trick you can do, which is why I always advocate that folks calculate both numbers rather than one over the other.”

And in fact, the difference between weighted ACV and the standard ACV is itself a measure of concentration in your company. And so it's a nice sort of trick you can kind of do, which is why I always advocate that folks calculate both numbers rather than one over the other. 

Allan: So now Nnamdi, I might've had an aha! moment and maybe I'm reading too much into this, but if you're saying that it's the companies where there's there's expansion capability, so you come in, you have a base price, your customers are loving the product.

They're expanding either through usage or through a number of users or a certain feature attributes. Are those also the companies, the companies that have a high differential in their weighted ACV versus ACV, or else those are the companies that have a better net dollar retention, they have other unit economics that are healthier.

Or am I reading too much into that? 

Nnamdi: No, I think you're reading into it in the same way that I've read into it, which is to say I sort of had a similar intuition that there might be a relationship with weighted ACV and it's different from typical ACV and then net revenue retention, net dollar retention.

I haven't done the analysis to actually confirm this across either public or private companies. So I can't say for certain, but the intuition I think is right, which is basically the more room there is to end up with the sort of whale customers, then the more likely you are to see you have this kind of difference for companies that have very high net retention that is also by the way, a weighted metric.

It isn't, people don't think of it that way, but it is actually a weighted metric. It's also a dollar weighted metric. You know, if you have your largest customer has a hundred percent net revenue retention and your smallest customer has 300%. Your overall net revenue retention is going to be a hundred percent. It's not going to be the simple average of 100 and 300. 

As I was saying, the more room there is to expand in the customer base and where you're going to see that. And I also haven't done this analysis publicly, but I'm fairly sure that if you look at companies that have very high dollar retention, they also have a high distribution that high variants of dollar retention across the different customers versus companies that tend to have very close to let's say a hundred percent as a baseline, very close to a hundred percent net dollar retention. It tends to concentrate around that  hundred percent. That’s my intuition. I don't know that for sure, but that would also be kind of evidence for this relationship between dollar retention.

Lauren: That's very cool. So we'll have another podcast on that once that analysis is out, but that's fantastic.

Let's go back to the beginning, if we could, how early in a company's lifecycle should they start, could they start looking at this and have it be actually meaningful? 

Nnamdi: It's a good question. There's sort of like two sides to it. I think on the one hand, If you're a small company with only a handful of customers, it's very easy for one of those customers to represent a large share of your revenue.

That's a great sort of natural thing to have happen for an early stage company. On the other hand, it's few early stage companies that are signing million dollar deals in the first place. Right. It's typically a larger company that would even have these whale contracts. And so you could see both sides of it.

It could be more or less relevant for an early stage group as a later stage or public company. I think it's great to get in the habit of calculating it as an early stage company, because I think, more so than just what the number literally is, there's like a mentality shift around taking this revenue-centric view that I think a lot of people miss, and I think it's an important perspective to have.

“It’s great to get in the habit of calculating it as an early stage company because more so than just what the number literally is, there’s a mentality shift around taking this revenue-centric view that I think a lot of people miss, and I think it’s an important perspective to have.”

And then you, especially if you're a company who is looking to raise venture capital money, I think this is a good metric to have ready to kind of talk through with your investors, because I can say I've met a lot of companies who if they could only show their standard ACV metric, it's not a very attractive metric for them because they have a lot of really small customers or even like a lot of totally free usage.

And so that basically dilutes the standard ACV calculation. So when you're talking to an investor, it's like, oh, this is low volume revenue because you're only selling to very, very small customers. When in fact most of the revenue could be coming from, you know, larger customers. And so I just think it can be very valuable for early stage companies. If only they were to calculate it. 

Allan: So when you're sitting around the Monday morning partner meeting, are you looking at both weighted and ACV? So somebody sends in a a pitch for a series A or series B or whatever it is, if you only had weighted would that actually tell you enough, or do you also need to see the standard simple average? 

Nnamdi: So I didn't really want to see both, part of the reason you were talking about this earlier, that folks tend to sort of overuse and abuse. Simple averages are because you only need two numbers to calculate a simple average. You see the numerator and denominator and you divide the two and you have your answers.

Versus this weighted ACV, you need a lot more data to calculate. In fact, you need the revenue of every single customer in the business to calculate it perfectly. And you can approximate it with, let's say, enough customers that represents most of the revenue, you could probably approximate it, but basically you need a lot of data.

“Simple averages are because you only need two numbers to calculate versus weighted ACV, you need a lot more data to calculate. In fact, you need the revenue of every single customer in the business to calculate it perfectly. As an investor, from that perspective, companies aren’t necessarily always sharing that data, so from the outside I can’t calculate the weighted ACV metric. You have to rely on companies providing it, and I hope they do. Hopefully this helps promote this metric and people start reporting it outwardly.”

As an investor, from that perspective, companies aren't necessarily always sharing that data. And so from the outside, I actually can't necessarily calculate the weighted ACV metric. And so you have to rely on the companies providing it, I hope they do. Hopefully this helps promote this metric and people start reporting this outwardly.

Even if they don't provide the underlying data to calculate it, I think having the number at least would be valuable that that the outfit of the calculation. But again, as I mentioned, I think it's good to have both because there's as much information in there, absolute numbers as the difference between the two. And so just having one or the other, doesn't tell you a full story.

Allan: Let's actually touch on that a little bit because you're obviously right. I mean, calculating the simpler average, I can do that on the back of a napkin, I can have a quick Slack message to my finance team or my sales team and say, “Hey, how many customers do we have? What's our current MRR?”

Boom, there it is. I think as soon as you start looking at every single individual customer and then dividing that by your total ACV or total MRR. You need to get system data from Stripe or from Zuora, or from your biz ops team, straight from your database.

And I think that definitely makes it more complicated, but it's certainly something that a company is tracking. Customer movement, tracking their MRR. They should have this, most systems will provide this. It just means that your formula is a little bit more complicated.

So every single customer you've got to do this pre-calculation, and then you figure out what the average is. So certainly having a good relationship with the biz ops team is probably not a bad idea

Nnamdi: Always a good idea. 

Allan: They are the holders of the data quality and the keys to all the systems.

Lauren: Yeah, the metrics masters. There’s a hashtag! And so you've talked about calculating both ACV and weighted ACV, the difference between the two being insightful. What are some of the other metrics around weighted ACV that would be good to look at in context? 

Nnamdi: So you have already mentioned net dollar retention and hopefully I will soon have better, empirical data to confirm the connection between those.

I think that's a great metric to have, there's a metric that I've been thinking about that maybe I'll write a piece on this too, which is basically weighted customer acquisition costs. So require you knowing what your per customer customer acquisition cost even was. So it's sort of like, I don't think that's actually possible for folks to calculate.

Allan: Super interesting idea, right? Because I mean,you can think of it more high level and you can say, well, let's segment our partner channel and the cost to acquire there and let's segment our direct channel. That's super high level, or maybe you can do that. High-level segmentation on geo. If you have a sales team and a marketing team for each geo, you're talking about bringing it down to the atomic unit, right. Which is a little bit more complicated, but I love that idea of being able to actually then move this MRR, this customer came through these acquisition channels and this is how much it costs. 

Nnamdi: I think iit'd be super cool to do it. And actually the point you mentioned about looking at channels as like a higher level for segmentation, you can do the same thing on the weighted ACV side.

Instead of weighting each customer, you could weight the segments. It doesn't give you the same results, but if that's a level of granularity you have, that's still a useful calculation that could be done. So whatever, however deep you can go, if you could calculate a weighted ACV.

We did customer acquisition costs and compared those. I think that would be an interesting thing for companies to do. Then there's sort of classic stuff. Like this is a magic number metric, which I actually don't like very much, and I wrote a whole piece about why people shouldn't use it. I don't know if it's worth going into here, but I don't think it actually measures sales productivity, measuring something else.

Lauren: What are your thoughts on customer revenue concentration? There's some benchmarks out there saying it's maybe not a great thing. If you're an enterprise customer, you have more than 10% of your revenue coming from one customer, things like that. What's your perspective? I mean, maybe it doesn't matter as long as you're aware?
Nnamdi: It's interesting because I think the standard idea is to think about concentration as a bad thing or a thing to be avoided.

What I've generally found in the data is that very rarely is all else equal. In other words, it's not that you can kind of just choose to have whatever you're concentrating on and want to have. Typically you were using your average revenue concentration in the first place because you've actually probably been very successful as a go to market organization.

You've landed these big contracts. It just so happens they take a huge chunk of your revenue, but that's not a bad thing. You'd rather have the contract. And so I think one has to be careful thinking about revenue concentration as purely a bad thing. It exists for a reason.

In fact, I think, and this is another thing that I don't fully have the data to be able to say, but I wonder if success sort of leads to concentration in the first place and there's sort of a subtle connection between the two, because when I did look at the public companies, it was shocking how consistent the concentration was across a bunch of high flying, highly valued SaaS companies. It wasn't that there were ones that was had lower concentration and were trading out a better multiple or something like that. It wasn't really any relationship like that. 

Allan: You've got my mind spinning again, it is interesting. If you don't have any concentration, you run the risk of not having any. If you have too much concentration, are you running the risk of one customer dictating the future of your company? Right, but there's probably a very nice place in there that is more skewed towards having concentration that tells you, you found that best fit cohort.

You're satisfying that need, and you're able to extract good net retention, good dollar retention out of that group. And they're sticking around. That's probably what you should be looking for. So there's probably a range there. And again, maybe this is more research down the road, but, I think there's a lot that this weighted concentration, weighted ACV can actually tell us here.

Nnamdi: I totally agree. I think it's like a complex multi-dimensional thing. Thinking of any one of those dimensions as good or bad, it's probably not the full. 

Allan: Did you find the companies that you're working with, are they adopting it? Are they seeing the potential of it? 

Nnamdi: Yeah, so I have at least one company that is adopting it and reporting it, which I'm very happy to see. The rest, I have to beat them over the head with it a little bit more, but it is being adopted, which is actually really nice, And I've been actually very pleased with the attention that the article has gotten. Like I write a lot of essays and some do better, some do worse, but I would like to see how well this one has sort of resonated with folks.

I think people immediately kind of get it when they read it through. Everyone has faced this issue of an average not telling the full story. And so here is kind of a solution to that. I can see my traffic, the article being shared in people's Slack channels or Asana. So I know it's getting around within companies. 

Allan: I think that's great. I mean, we both read it. We'll share the link for everybody as well. It's a great article. I mean, it goes into a good amount of depth without going over your head. And again, that last little piece about concentration, it sort of nailed it, right?

Because that was sort of, if you don't believe this, here's the next thing that you should be really thinking about? So, as I thought it was a great article, but we'll share. 

Lauren: For sure. And on that note, Nnamdi, are there any last words of advice that you would offer listeners as they're contemplating crunching this number?

One thing I would mention is that and this is actually true of a lot of metrics that don't get too caught up on like what the number is in isolation. As I mentioned earlier, there's at least as much information and what it is relative to another, so relative to the standard ACV calculation, what it is relative to itself over time, you know, how has the way they see ACV changing over time?

There's a sort of mini analysis of this in the piece, but it'll tend to change in a very different way than the standard ACV metric will change over time. And that is information in and of itself. And so I'm just thinking about, if you're trying to figure out what this number should be, you're sort of asking the wrong question, sort of what it is relative to other things, how it's evolving over time.. That's kind of, at least, an important piece of this. 

Allan: Nnamdi amazing advice, lots of insights. Again, everybody Nnamdi Iregbulem from Lightspeed Ventures on weighted ACV, on a revenue culture, and sort of taking the bigger picture into context. Thank you so much for joining us today.

Nnamdi: Thank you for having me.