SaaS Metrics 2.0 – A Guide to Measuring and Improving what Matters

“If you cannot measure it, you cannot improve it” – Lord Kelvin

This article is a comprehensive and detailed look at the key metrics that are needed to understand and optimize a SaaS business. It is a completely updated rewrite of an older post.  For this version, I have co-opted two real experts in the field: Ron Gill, (CFO, NetSuite), and Brad Coffey (VP of Strategy, HubSpot), to add expertise, color and commentary from the viewpoint of a public and private SaaS company. My sincere thanks to both of them for their time and input.

SaaS/subscription businesses are more complex than traditional businesses. Traditional business metrics totally fail to capture the key factors that drive SaaS performance. In the SaaS world, there are a few key variables that make a big difference to future results. This post is aimed at helping SaaS executives understand which variables really matter, and how to measure them and act on the results.

The goal of the article is to help you answer the following questions:

  • Is my business financially viable?
  • What is working well, and what needs to be improved?
  • What levers should management focus on to drive the business?
  • Should the CEO hit the accelerator, or the brakes?
  • What is the impact on cash and profit/loss of hitting the accelerator?

(Note: although I focus on SaaS specifically, the article is applicable to any subscription business.)

What’s so different about SaaS?

SaaS, and other recurring revenue businesses are different because the revenue for the service comes over an extended period of time (the customer lifetime). If a customer is happy with the service, they will stick around for a long time, and the profit that can be made from that customer will increase considerably. On the other hand if a customer is unhappy, they will churn quickly, and the business will likely lose money on the investment that they made to acquire that customer. This creates a fundamentally different dynamic to a traditional software business: there are now two sales that have to be accomplished:

  1. Acquiring the customer
  2. Keeping the customer (to maximize the lifetime value).

Because of the importance of customer retention, we will see a lot of focus on metrics that help us understand retention and churn. But first let’s look at metrics that help you understand if your SaaS business is financially viable.


The SaaS P&L / Cash Flow Trough

SaaS businesses face significant losses in the early years (and often an associated cash flow problem). This is because they have to invest heavily upfront to acquire the customer, but recover the profits from that investment over a long period of time. The faster the business decides to grow, the worse the losses become. Many investors/board members have a problem understanding this, and want to hit the brakes at precisely the moment when they should be hitting the accelerator.

In many SaaS businesses, this also translates into a cash flow problem, as they may only be able to get the customer to pay them month by month. To illustrate the problem, we built a simple Excel model which can be found here.  In that model, we are spending $6,000 to acquire the customer, and billing them at the rate of $500 per month. Take a look at these two graphs from that model:



If we experience a cash flow trough for one customer, then what will happen if we start to do really well and acquire many customers at the same time? The model shows that the P&L/cash flow trough gets deeper if we increase the growth rate for the bookings.


But there is light at the end of the tunnel, as eventually there is enough profit/cash from the installed base to cover the investment needed for new customers. At that point the business would turn profitable/cash flow positive – assuming you don’t decide to increase spending on sales and marketing. And, as expected, the faster the growth in customer acquisition, the better the curve looks when it becomes positive.

Ron Gill, NetSuite:

If plans go well, you may decide it is time to hit the accelerator (increasing spending on lead generation, hiring additional sales reps, adding data center capacity, etc.) in order to pick-up the pace of customer acquisition. The thing that surprises many investors and boards of directors about the SaaS model is that, even with perfect execution, an acceleration of growth will often be accompanied by a squeeze on profitability and cash flow.

As soon as the product starts to see some significant uptake, investors expect that the losses / cash drain should narrow, right? Instead, this is the perfect time to increase investment in the business. which will cause losses to deepen again. The graph below illustrates the problem:


Notice in the example graph that the five customer per month model ultimately yields a much steeper rate of growth, but you have to go through another deep trough to get there. It is the concept of needing to re-enter that type of trough after just having gotten the curve to turn positive that many managers and investors struggle with.

Of course this a special challenge early-on as you need to explain to investors why you’ll require additional cash to fund that next round of acceleration. But it isn’t just a startup problem. At NetSuite, even as a public company our revenue growth rate has accelerated in each of the last three years. That means that each annual plan involves a stepping-up of investment in lead generation and sales capacity that will increase spending and cash flow out for some time before it starts yielding incremental revenue and cash flow in. As long as you’re accelerating the rate of revenue growth, managing and messaging around this phenomenon is a permanent part of the landscape for any SaaS company.

Why is growth important?

We have suggested that as soon as the business has shown that it can succeed, it should invest aggressively to increase the growth rate. You might ask question: Why?

SaaS is usually a “winner-takes-all” game, and it is therefore important to grab market share as fast as possible to make sure you are the winner in your space. Provided you can tell a story that shows that eventually that growth will lead to profitability, Wall Street, acquiring companies, and venture investors all reward higher growth with higher valuations. There’s also a premium for the market leader in a particular space.

However not all investments make sense. In the next section we will look at a tool to help you ensure that your growth initiatives/investments will pay back:  Unit Economics.

A Powerful Tool: Unit Economics

Because of the losses in the early days, which get bigger the more successful the company is at acquiring customers, it is much harder for management and investors to figure out whether a SaaS business is financially viable. We need some tools to help us figure this out.

A great way to understand any business model is to answer the following simple question:

Can I make more profit from my customers than it costs me to acquire them?

This is effectively a study of the unit economics of each customer. To answer the question, we need two metrics:

  • LTV – the Lifetime Value of a typical customer
  • CAC – the Cost to Acquire a  typical Customer

(For more on how to calculate LTV and CAC, click here.)

Entrepreneurs are usually overoptimistic about how much it costs to acquire a customer. This probably comes from a belief that customers will be so excited about what they have built, that they will beat a path to their doors to buy the product. The reality is often very different! (I have written more on this topic here: Startup Killer: The Cost of Customer Acquisition, and here: How Sales Complexity impacts CAC.)

Is your SaaS business viable?

In the first version of this article, I introduced two guidelines that could be used to judge quickly whether your SaaS business is viable. The first is a good way to figure out if you will be profitable in the long run, and the second is about measuring the time to profitability (which also greatly impacts capital efficiency).


Over the last two years, I have had the chance to validate these guidelines with many SaaS businesses, and it turns out that these early guesses have held up well. The best SaaS businesses have a LTV to CAC ratio that is higher than 3, sometimes as high as 7 or 8. And many of the best SaaS businesses are able to recover their CAC in 5-7 months. However many healthy SaaS businesses don’t meet the guidelines in the early days, but can see how they can improve the business over time to get there.

The second guideline (Months to Recover CAC)  is all about time to profitability and cash flow. Larger businesses, such as wireless carriers and credit card companies, can afford to have a longer time to recover CAC, as they have access to tons of cheap capital. Startups, on the other hand, typically find that capital is expensive in the early days.  However even if capital is cheap, it turns out that Months to recover CAC is a very good predictor of how well a SaaS business will perform. Take a look at the graph below, which comes from the same model used earlier. It shows how the profitability is anemic if the time to recover CAC extends beyond 12 months.

I should stress that these are only guidelines, there are always situations where it makes sense to break them.


Three uses for the SaaS Guidelines

  1. One of the key jobs of the CEO is to decide when to hit the accelerator pedal. The value of these two guidelines is that they help you understand when you have a SaaS business that is in good shape, where it makes sense to hit the accelerator pedal. Alternatively if your business doesn’t meet the guidelines, it is a good indicator that there is more tweaking needed to fix the business before you should expand.
  2. Another way to use the two guidelines is for evaluating different lead sources. Different lead sources (e.g. Google AdWords, TV, Radio, etc.) have different costs associated with them. The guidelines help you understand if some of the more expensive lead generation options make financial sense. If they meet these guidelines, it makes sense to hit the accelerator on those sources (assuming you have the cash).Using the second guideline, and working backwards, we can tell that if we are getting paid $500 per month, we can afford to spend up to 12x that amount (i.e. $6,000) on acquiring the customer. If we’re spending less than that, you can afford to be more aggressive and spend more in marketing or sales.
  3. There is another important way to use this type of guideline: segmentation. Early-stage companies are often testing their offering with several different uses/types of customers / pricing models / industry verticals. It is very useful to examine which segments show the quickest return or highest LTV to CAC in order to understand which will be the most profitable to pursue.

Unit Economics in Action: HubSpot Example

HubSpot’s unit economics were recently published in an article in Forbes:

You can see from the second row in this table how they have dramatically improved their unit economics (LTV:CAC ratio) over the five quarters shown. The big driver for this was lowering the MRR Churn rate from 3.5% to 1.5%. This drove up the lifetime value of the customer considerably.  They were also able to drive up their AVG MRR per customer.

Brad Coffey, HubSpot:

In 2011 and early 2012 we used this chart to guide many of our business decisions at HubSpot. By breaking LTV:CAC down into its components we could examine each metric and understand what levers we could pull to drive overall improvement.

It turned out that the levers we could pull varied by segment. In the SMB market for instance we had the right sales process in place – but had an opportunity to improve LTV by improving the product to lower churn and increasing our average price in the segment. In the VSB (Very Small Business) segment, by contrast, there wasn’t as much upside left on the LTV (VSB customers have less money and naturally higher churn) so we focused on lowering CAC by removing friction from our sales process and moving more of our sales to the channel.

Two kinds of SaaS business:

There are two kinds of SaaS business:

  • Those with primarily monthly contracts, with some longer term contracts. In this business, the primary focus will be on MRR (Monthly Recurring Revenue)
  • Those with primarily annual contracts, with some contracts for multiple years. Here the primary focus is on ARR (Annual Recurring Revenue), and ACV (Annual Contract Value).

Most of the time in this article, I will refer to MRR/ACV. This means use MRR if you are the first kind of business, or ACV if you are the second kind of business. The dashboard shown below assumes monthly contracts (MRR). However in the downloadable spreadsheet, there is a tab that shows the same dashboard for the second kind, focusing on ACV instead of MRR.

SaaS Bookings: Three Contributing Elements

Every month in a SaaS business, there are three elements that contribute to how much MRR will change relative to the previous month:

What happened with new customers added in the month:

  • New MRR (or ACV)

What happened in the installed base of customers:

  • Churned MRR (or ACV) (from existing customers that cancelled their subscription. This will be a negative number.)
  • Expansion MRR (or ACV) (from existing customers who expanded their subscription)

The sum all three of these makes up your Net MRR or ACV Bookings:


I recommend that you track these using a chart similar to the one below:


This chart shows the three components of MRR (or ACV) Bookings, and the Net New MRR (or ACV) Bookings. By breaking out each component, you can track the key elements that are driving your business. The one variation we would recommend making to this chart is to show a dotted line for the plan, so you can track how you are doing against plan for each of the four lines. This is one of the most important charts to help you understand and run your business.

Ron Gill, NetSuite:

This chart is really good. I also like to look at this data in tabular form because I want to know y-o-y growth rates. E.g. “Net new MRR is up 25% over June of last year”. The Y-o-Y % is a metric easily compared with increased spending, sales capacity, etc.

The Importance of Customer Retention (Churn)

In the early days of a SaaS business, churn really doesn’t matter that much. Let’s say that you lose 3% of your customers every month. When you only have a hundred customers, losing 3 of them is not that terrible. You can easily go and find another 3 to replace them. However as your business grows in size, the problem becomes different. Imagine that you have become really big, and now have a million customers. 3% churn means that you are losing 30,000 customers every month! That turns out to be a much harder number to replace. Companies like Constant Contact have run into this problem, and it has made it very hard for them to keep up their growth rate.

Ron Gill, NetSuite:  

One oft-overlooked aspect of churn is that the churn rate, combined with the rate of new ARR adds, not only defines how fast you can grow the business, it also defines the maximum size the business can reach (see graph below).


It is an enlightening exercise to build a simple model like this for your business and plot where your current revenue run rate sits on the blue line defined by your present rate of ARR adds and churn. Are you near the left-hand side, where the growth is still steep and the ceiling is still far above? Or, are you further to the right where revenue growth will level off and there is limited room left to grow? How much benefit will you get from small improvements in churn or the pace of new business sign-up?

At NetSuite, we’ve had great success shifting the line in the last few years by both dramatically decreasing churn and by increasing average deal size and volume, thus increasing ARR adds. The result was both to steadily move the limit upward and to steepen the growth curve at the current ARR run rate, creating room for increasingly rapid expansion.

The Power of Negative Churn

The ultimate solution to the churn problem is to get to Negative Churn.


There are two ways to get this expansion revenue:

  1. Use a pricing scheme that has a variable axis, such as the number of seats used, the number of leads tracked, etc. That way, as your customers expand their usage of your product, they pay you more.
  2. Upsell/Cross-sell them to more powerful versions of your product, or additional modules.

To help illustrate the power of negative churn, take a look at the following two graphs that show how cohorts behave with 3% churn, and then with 3% negative churn. (Since this is the first time I have used the word Cohort, let me explain what it means. A cohort is simply a fancy word for a group of customers. In the SaaS world, it is used typically to describe the group that joined in a particular month. So there would be the January cohort, February cohort, etc.  In our graphs below, a different color is used for each month’s cohort, so we can see how they decline or grow, based on the churn rate.)

In the top graph, we are losing 3% of our revenue every month, and you can see that with a constant bookings rate of $6k per month, the revenue reaches $140k after 40 months, and growth is flattening out. In the bottom graph, we may be losing some customers, but the remaining customers are more than making up for that with increased revenue. With a negative churn rate of 3%, we reach $450k in revenue (more then 3x greater), and the growth in revenues is increasing, not flattening.


For more on this topic, you may wish to refer to these two blog posts of mine:

Defining a Dashboard for a SaaS Company

The following section should be most useful for readers who are interested putting together a dashboard to help them manage their SaaS business. To this, we created an excel file for an imaginary SaaS company, and laid out a traditional numeric report on one tab, and then a dashboard of graphs on a second tab (see below). These represent one view on how to do this. You may have a very different approach. But hopefully this will give you some ideas. I would recommend adding a dotted line with the plan number to all graphs. This will allow you to quickly see how you are doing versus plan.

There are two versions of the Dashboard: the one shown below, which is designed for companies using primarily monthly contracts (focused on MRR). And a second version that can be found here which is designed for companies using annual contracts, focusing on ACV (Annual Contract Value).





Brad Coffey, HubSpot:

At HubSpot we obsess over these metrics – and watch many of them every day. Each night we send out a ‘waterfall’ chart that tracks our progress against our typical progress given the number of business days left in the month. Here is an example of what we look at to ensure we’re on track to meet our net MRR goals.


By looking at this daily we can take action immediately if we’re tracking towards a bad month or quarter. Things like services promotion (for churned MRR) or sales contests & promotions (for new & expansion MRR) are adjustments we make within a given month in order to nail our goals. (In this model we combine expansion and churned MRR into one churned MRR line).

Detailed definitions of the various metrics used

Detailed definitions for each of the various metrics used can be found in this reference document:


Revenue Churn vs Customer Churn – why are they different?

You might be wondering why it’s necessary to track both Customer Churn and Revenue Churn. Imagine a scenario where we have 50 small accounts paying us $100 a month, and 50 large accounts paying us $1,000 a month. In total we have 100 customers, and an MRR of $55,000 at the start of the month. Now imagine that we lose 10 of them. Our Customer churn rate is 10%. But if out of the ten churned customers, 9 of them were small accounts, and only one was a large account. We would only have lost $1,900 in MRR. That represents only 3.4% Revenue Churn. So you can see that the two numbers can be quite different. But each is important to understand if we want a complete picture of what is going on in the business.

Getting paid in advance

Getting paid in advance is really smart idea if you can do it without impacting bookings, as it can provide the cash flow that you need to cover the cash problem that we described earlier in the article. It is often worth providing good financial incentives in the form of discounts to encourage this behavior. The metric that we use to track how well your sales force is doing in this area is Months up Front.

Getting paid more upfront usually also helps lower churn. This happens because the customer has made a greater commitment to your service, and is more likely to spend the time getting it up and running. You also have more time to overcome issues that might arise with the implementation in the early days. Calculating LTV and CAC

The Metric “Months up Front” has been used at both HubSpot and NetSuite in the past as a way to incent sales people to get more paid up front when a new customer is signed. However asking for more money up front may turn off certain customers, and result in fewer new customers, so be careful how you balance these two conflicting goals.

Calculating CAC and LTV

Detailed information on how to calculate LTV and CAC is provided in the supplemental document that can be accessed by clicking here.

More on Churn: Cohort Analysis

Since churn is such a critical element for success in a SaaS company, it is an area that requires deeper exploration to understand. Cohort Analysis is one of the important techniques that we use to gain insight.

As mentioned earlier, a cohort is simply a fancy name for a group. In SaaS businesses, we use cohort analysis to observe what happens to the group of customers that joined in a particular month. So we will have a January cohort, a February cohort, etc. We would then be able to observe how our January cohort behaves over time (see illustration below).


This can help answer questions such as:

  • Are we losing most of the customers in the first couple of months?
  • Does Churn stabilize after some period of time?

Then if took some actions to try to fix churn in early months, (i.e with better product features, easier on-boarding, better training, etc.) we would want to know if those changes had been successful. The cohort analysis allows us to do this by comparing how more recent cohorts (e.g. July in the table above) compared against January. The table above shows that we made a big improvement in the first month churn going from 15% to 4%.

Two ways to run Cohort Analysis

There are two ways to run Cohort Analysis: the first looks at the number of customers, and the second looks at the Revenue. Each teaches us something different and valuable. The example graph below simply looks at the number of customers in each cohort over time:


The example graph below looks at how MRR evolves over time for each cohort. This particular example illustrates how the graph would look if there is very strong negative churn. As you can see, the increase in revenue from the customers that are still using the service is easily outpacing the lost revenue from churned customers. It is pretty rare to see things look this good, but it is the ideal situation that we are looking for. For those wondering if this can be achieved, one company in our portfolio, Zendesk, that has numbers that are even better than those shown in the example below.


In the situation above, you will need a more complex formula to calculate LTV, as the value of the average customer is increasing over time. For more on that topic, you may want to check out the accompanying definitions document.

Predicting Churn: Customer Engagement Score

Since churn is so important, wouldn’t it be useful if we could predict in advance which customers were most likely to churn? That way we could put our best customer service reps to work in an effort to save the situation. It turns out that we can do that by instrumenting our SaaS applications and tracking whether our users are engaged with the key sticky features of the product. Different features will deserve different scores. For example if you were Facebook, you might score someone who uploaded a picture as far more engaged (and therefore less likely to churn), than someone who simply logged in and viewed one page.

Similarly if you sold your SaaS product to a 100 person department, and only 10 people were using it, you would score that differently to 90 people using it. So the recommendation is that you create a Customer Engagement Score, based on allocating points for the particular features used. Allocate more points for the features you believe are most sticky. (Later on you can go back and look at the customers who actually churned, and validate that you picked the right features as a predictor of who would churn.) And separately score how many users are engaged with specific scores.

Over time you’ll also come to discover which types of use are the best indicators of possible upsell. (HubSpot was the first company that I worked with who figured this out, and they called it their CHI score. CHI stands for Customer Happiness Index. It evolved to be a very good predictor for churn.)

Brad Coffey, HubSpot:

At HubSpot we had a lot of success looking at this metric – we called in Customer Happiness Index (CHI). First – by running the analysis we identified the parts of our application that provide the most value to customers and could invest accordingly in driving adoption in those areas. Second – we used this aggregate score as an early proxy for success as we experimented with different sales and onboarding processes. If a set of customers going through an experiment had a low CHI score we could kill the project without waiting 6 or 12 months to analyze the cohort retention.

NPS – Net Promoter Score

Since it is likely that customer satisfaction is likely to be good predictor of future churn, it would be useful to survey customer satisfaction. The recommended way to measure customer happiness is to use Net Promoter Score (NPS). The beauty of NPS is that it is a standardized number, so you can compare your company to others.  For more details on Net Promoter Score, click here.

Guidelines for Churn

If your Net Revenue Churn is high (above 2% per month) it is an indicator that there is something wrong in your business. At 2% monthly churn, you are losing about 22% of your revenue every year. That is nearly a quarter of your revenue! It’s a clear indication that there is something wrong with the business. As the business gets bigger, this will become a major drag on growth.

We recommend that you work on fixing the problems that are causing this before you go on to worry about other parts of your business. Some of the possible causes of churn are:

  • You are not meeting your customers expectations.
    • The product may not provide enough value
    • Instability or bugginess
  • Your product is not sticky. It might provide some value in the first few months, and then once the customer has that value, they may feel they don’t need to keep paying. To make your product sticky, try making it a key part of their monthly workflow, and/or have them store data in your product that is highly valuable to them, where the value would be lost of they cancelled.
  • You have not successfully got the customer’s users to adopt the product. Or they may not be using certain of the key sticky features in the product.
  • Your sales force may have oversold the product, or sold it to a customer that is not well suited to get the benefits
  • You may be selling to SMB’s where a lot of them go out of business. It isn’t enough that what you’re selling is sticky. Who you’re selling it to must also be sticky.
  • You are not using a pricing scheme that helps drive expansion bookings

The best way to find out why customers are churning is to get on the phone with them and ask them. If churn is a significant part of your business, we recommend that the founders themselves make these calls. They need to hear first hand what the problem is, as this is so important for the success of the business. And they are likely to be the best people to design a fix for the problem.

The Importance of Customer Segmentation

In all SaaS businesses there will likely come a moment where they realize that not all customers are created equal. As an example, bigger customers are harder to sell to, but usually place bigger orders, and churn less frequently. We need a way to understand which of these are most profitable, and this requires us to segment the customer base into different types, and compute the unit economics metrics for each segment separately. Common segments are things size of of customer, vertical industry, etc.

Despite the added work to produce the metrics, there is high value in understanding the different segments. This tells us which parts of the business are working well, and which are not. In addition to knowing where to focus and invest resources, we may recognize the need for different marketing messages, product features. As soon as you start doing this segmented analysis, the benefits will become immediately apparent.

For each segment, we recommend tracking the following metrics:

  • ARPA (Average Revenue per Account per month)
  • Net MRR Churn rate (including MRR expansion)
  • LTV
  • CAC
  • LTV: CAC ratio
  • Months to recover CAC
  • Customer Engagement Score


Brad Coffey, HubSpot:

At HubSpot, we started to see some of our biggest improvements in unit economics when we started segmenting our business and calculating the LTV to CAC ratio for each of our personas and go to market strategies.

As one good example – when we started this analysis, we had 12 reps selling directly into the VSB market and 4 reps selling through Value Added Resellers (VARs). When we looked at the math we realized we had a LTV:CAC ratio of 1.5 selling direct, and a LTV:CAC ratio of 5 selling through the channel. The solution was obvious. Twelve months later we had flipped our approach – keeping just 2 reps selling direct and 25 reps selling through the channel. This dramatically improved our overall economics in the segment and allowed us to continue growing.

We ended making similar investments in other high LTV:CAC segments. We went so far as to incentivize our sales managers to grow their teams – but then would only place new sales hires into the segments with the best economics. This ensured we continued to invest in the best segments and aligned incentives throughout the company on our LTV:CAC goals. It also allowed us to push innovation down to the sales manager level. Managers could experiment with org structure, and sales processes – but they knew that if they didn’t hit their LTV:CAC goals they wouldn’t be able to grow their teams.

Calculating LTV:CAC by segment can be challenging, especially on the CAC side. It’s relatively easy at the top level to add up all the marketing and sales expense in a period and divide it by the total number of customers (to get CAC). Once you try to segment down your spend you run into questions like ‘how much marketing expense do I allocate to a given segment’, ‘how much of the sales expense’?

We solved this by allocating marketing expense based on number of leads and sales expense based on headcount but it’s not perfect. For us the keys are: 1) Needs to account for all costs – no free lunch, 2) It needs to be consistent over time. Progress on improving the metric is more important than the actual value.

Funnel Metrics

The metrics that matter for each sales funnel, vary from one company to the next depending on the steps involved in the funnel. However there is a common way to measure each step, and the overall funnel, regardless of your sales process. That involves measuring two things for each step:  the number of leads that went into the top of that step, and the conversion rate to the next step in the funnel (see below).

In the diagram above, (mirrored in the dashboard), we show a very simple three phase sales process, with visitors coming to a web site, and some portion of them signing up for a trial. Then some of the trials convert to purchases.  As you can see in the dashboard, we will want to track the number of visitors, trials and closed deals. Our goal should be to increase those numbers over time. And we will also want to track the conversion rates, with the goal of improving those over time.

Using Funnel Metrics in Forward Planning

Another key value of having these conversion rates is the ability to understand the implications of future forecasts. For example, lets say your company wants to do $4m in the next quarter. You can work backwards to figure out how many demos/trials that means, and given the sales productivity numbers – how many salespeople are required, and going back a stage earlier, how many leads are going to be required. These are crucial planning numbers that can change staffing levels, marketing program spend levels, etc.

Sales Capacity

In many SaaS businesses, sales reps play a key role in closing deals. In those situations, the number of productive sales people (Sales Capacity) will be a key driver of bookings. It is important to work backwards from any forecasts that are made, to ensure that there is enough sales capacity. I’ve seen many businesses miss their targets because they failed to hire enough productive salespeople early enough.

It’s also worth noting that some percentage of new sales hires won’t meet expectations, so that should be taken into consideration when setting hiring goals. Typically we have seen failure rates around 25-30% for field sales reps, but this varies by company. The failure rate is lower for inside sales reps.

When computing Sales Capacity, if a newer rep is still ramping and only expected to deliver 50% of quota, they can be counted as half of a productive rep. That is often referred to as Full Time Equivalent or FTE for short.

Another important metric to understand is the number of leads required to feed a sales rep. If you are adding sales reps, make sure you also have a clear plan of how you will drive the additional leads required.

There is much more that could be said on this topic, but since it is all very similar to managing a sales force in a traditional software company, we will leave that for other blog posts.

Understanding the ROI for different Lead Sources

Our experiences with SaaS startups indicate that they usually start with a couple of lead generation programs such as Pay Per Click Google Ad-words, radio ads, etc. What we have found is that each of these lead sources tends to saturate over time, and produce less leads for more dollars invested. As a result, SaaS companies will need to be constantly evaluating new lead sources that they can layer in on top of the old to keep growing.

Since the conversion rates and costs per lead vary quite considerably, it is important to also measure the overall ROI by lead source.

Growing leads fast enough to feed the front end of the funnel is one of the perennial challenges for any SaaS company, and is likely to be one of the greatest limiting factors to growth. If you are facing that situation, the most powerful advice we can give you is to start investing in Inbound Marketing techniques (see Get Found using Inbound Marketing). This will take time to ramp up, but if you can do it well, will lead to far lower lead costs, and greater scaling than other paid techniques. Additionally the typical SaaS buyer is clearly web-savvy, and therefore very likely to embrace inbound marketing content and touchless selling techniques.

What Levers are available to drive Growth

SaaS businesses are more numerically driven than most other kinds of business. Making a small tweak to a number like the churn rate can have a very big impact on the overall health of the business. Because of this we frequently see a “quant” (i.e. a numbers oriented, spreadsheet modeling, type of person) as a valuable hire in a SaaS business. At HubSpot, Brad Coffey played that role, and he was able to run the models to determine which growth plays made the most sense.

Understanding these SaaS metrics is a key step towards seeing how you can drive your business going forward. Let’s look at some of the levers that these imply as growth drivers for your business:


  • Get Churn and customer happiness right first (if this isn’t right, the business isn’t viable, so no point in driving growth elsewhere. You will simply be filling a leaky bucket.)


  • You’re in a product business – first and foremost: fix your product.
    • If you’re using a free trial, focus on getting the conversion rate for that right (ideally around 15 – 20%). If this isn’t right, your value proposition isn’t resonating, or you may have a market where there is not enough pain to get people to buy.
    • Win/Loss ratio should be good
    • Trial or Sales conversion rates on qualified leads should be good

Funnel metrics

  • Increase the number of raw leads coming in to the Top of your funnel
  • Identify the profitable lead sources and invest in those as much as possible. Conversely stop investing in poor lead sources until they can be tweaked to make them profitable.
  • Increase the Conversion Rates at various stages in the funnel

Sales Metrics

  • Sales productivity (focus on getting this right consistently across a broad set of sales folks before hitting the gas)
  • Add Sales Capacity. But first make sure you know how to provide them with the right number of leads. This turns out to be one of the key levers that many companies rely on for growth. We have learned from experience how important it is to meet your targets for sales capacity by hiring on time, and hiring the right quality of sales people so there are fewer failures.
  • Increase retention for your sales people. Since you have invested a lot in making them fully productive, get the maximum return on that investment by keeping them longer.
  • Look at adding Business Development Reps. These are outbound sales folks who specialize in prospecting to a targeted list of potential buyers. For more on this topic, click here.

Pricing/Upsell/Cross Sell

  • Multi-axis pricing
  • Additional product modules (easier to sell more to existing customers than it is to sell to brand new customers)

Brad Coffey, HubSpot:

Turns out the pricing your product right can have a huge impact on the unit economics. Not simply by getting the average MRR right, or by providing upsell opportunities – but also by signaling what pieces of the product are most valuable.

At HubSpot we changed our pricing in 2011 to be tiered based on the number of contacts in the system – and actually saw an increase in adoption of the contacts application after we made the change. This is counter-intuitive but makes sense given that we sell through an inside sales team. After the pricing change, sales reps now could make a lot more money by selling the contacts. And they quickly become much better at positioning that part of the product, as well as finding companies with a contacts-based use case. Product quality will remain paramount – but it’s remarkable how much impact pricing, packaging and sales commission structure can have on product adoption and unit economics.

Customer Segmentation

Customer Segmentation analysis will help point out which are your most profitable segments. Two immediate actions that are suggested by this analysis are:

  • Double down on your most profitable segments
  • Look at your less profitable segments and consider changes that would make them more profitable: lower cost marketing & sales approaches, higher pricing, product changes, etc. If nothing seems to make sense, spend less effort on these segments.

International Markets

Expansion internationally is only recommended for fairly mature SaaS companies that already have honed their business practices in their primary market. It is far harder to experiment and tune a business in far off regions, with language and cultural differences.

Brad Coffey, HubSpot

  • One of the biggest challenges we face is the trade-off between growth and unit economics (specifically churn).  Many of the things that we have done to reduce churn have (potentially) come at the expense of lowering our growth rate. Those have been some of our hardest decisions:  e.g. requiring upfront payments, requiring customers buy consulting, holding sales reps accountable for churn, etc. We are always looking at things that give us growth without the tradeoff of lower growth. For example product improvement is an obvious one – a better product is easier to sell and provides more value to the customer. Services promotions actually work well too. Many of the options that SaaS companies have to adjust their business are not simply a win-win but are still worth exploring. Too many companies think that every problem is a product problem and every solution is that the product must get better.
  • The other thing that’s really important is that companies don’t try to spin these numbers.  There is so much pressure to dismiss a bad customer (who hurt your churn number) or exclude costs (only count marketing ‘program’ spend – not headcount).  If you can get the accounting close enough to right it actually frees management from needing to make every decision.  If the accounting is right management can obsess over setting goals (growth, LTV:CAC), hold people accountable to those goals and then give autonomy to their team on how to achieve those goals.

Plan ahead

It takes time for most initiatives to have an impact. We’ve learned from some tough lessons that planning has to be done well in advance to drive a SaaS business. For example if you are not happy with your current growth rate, it will often take nine to twelve months from the point of decision before the growth resulting from increased investment in sales and marketing will actually be observed.

The High Level Picture: How to Run a SaaS Business

Hopefully what you will have gathered from the discussion above is that there are really three things that really matter when running a SaaS business:

  1. Acquiring customers
  2. Retaining customers
  3. Monetizing your customers

The second item should be first on your list of things to get right. If you can’t keep your customers happy, and keep them using the service, there is no point in worrying acquiring more of them. You will simply be filling a leaky bucket. Rather focus your attention on plugging the leaks.

SaaS businesses are remarkably influenced by a few key numbers. Making small improvements to those numbers can dramatically improve the overall health of the business.

Once you know your SaaS business is viable using the guidelines provided for LTV:CAC, and Time to recover CAC, hit the accelerator pedal. But be prepared to raise the cash needed to fund the growth.

Although this article is long and occasionally complex, we hope that it has helped provide you with an understanding of which metrics are key, and how you can go about improving them.


I would like to thank Ron Gill, the CFO of NetSuite, and Brad Coffey & Brian Halligan of Hubspot for their help in writing this. I would like to thank the HubSpot management team without whom none of this would be possible. Most of my learnings on SaaS have come from working with them. I would also like to thank Gail Goodman, the CEO of Constant Contact who also taught us many of the key metrics in her role as board member of HubSpot.

Be Sociable, Share!


  • David Pethick

    Thank you David and the other contributors for such an amazing resource. I can’t wait to read version 3.0 at some point in the future!

  • Suresh

    more than i expected. good insight.

  • Sean McAllister

    Hi David
    Great article, a great article that clearly explains the levers involved. I have heard of companies just simply running out of cash- if only they had this.

    Our company is a Startup and we don’t have any real data as yet on the expected churn but can only guess. We have one 2 high value contracts before we have finish our Saas platform. I am estimating what our cost will be for sales and marketing.

    At a high level based on what I believe the customer needs are: our LTV for a customer will be $40000 over the 60 moths. 60 months is a standard term in this industry for supplying the software solution. We will be getting approximately $250 per month under the subscription but I believe the revenue that will make up the remaining LTV will come from service as up sells, which compliment the product giving more functionality.

    What is the best way to incorporate the service upsell into the into the spreadsheet so that I can get an indication of the cash low point that will be realistic ?


  • David Skok

    Hi Sean, the spreadsheet that I included with this post was not intended to be a model that you could use as is. The formulae were just used to create some sample data, and therefore wouldn’t make sense is a real world situation.
    For your own situation, I would look to figure out a way to model these additional service payments for a typical client and then create a model that includes that as a second line item alongside of the true recurring revenue. Be sure to also take into consideration the Gross Margin for that service component. Any people involved in delivering the service should be included in the Cost of Goods Sold line, thereby reducing your gross margin. So instead of $40,000 being your Lifetime value, it would be $40,000 x Gross Margin %.
    I am sorry I can’t provide you with more help in constructing that model. But it should not be too hard to figure out.

  • Ben

    One question I had: Why do you model new customers as an absolute number gain (2+ per month) while model the churn as a % of existing customer base? To me, it would make more sense to put both new customers and churn (opposites of each other) on a relative scale.

  • David Skok

    Hi Ben,

    The reason for that is that you will find that the number of customers that churn at any time is directly proportional to the total number of customers that you have. So if you lost 20 customers out of 1,000 on month, and changed nothing, you would likely see yourself losing about 40 customers when you reach 2,000 customers.
    I hope that makes sense.
    Best, David

  • Ben


    Thanks for the reply. That makes sense. And are you then assuming the addition of new customers is not generally related to the total number of customers you have but rather an absolute value with a fixed marginal increase per month (2+ in your example).

    PS, I found this post to be one of the most useful and valuable I’ve read on the internet. Thanks for taking the time to write this and replying to the numerous comments!


  • David Skok

    Hi Ben, That is correct. The number of new customers added is usually directly related to spend on Sales and Marketing, and in this case for the model, I assumed they could grow the number of new customers booked by 2 each month. That is purely a guess to have a number in the model. It can vary widely by company.
    I would not suggest that you project forward a “fixed marginal increase per month (+2 in your example)” as my experience shows it will vary depending on how well marketing is doing at lead generation, and how much investment is put into adding sales resources to help close the leads (assuming the business requires some level of sales touch to close a customer). When I am working with a specific company, I usually build a model of how the funnel works (or based on a best guess of how it is hoped it will work). The funnel will have different conversion rates. For example, a very simple funnel might involved getting visitors to a web site, converting visitors into a free trial, and then converting the free trial users into a customer. Based on those conversion rates, it is possible to take any forecast for bookings that management would like to see, and compute the number of trials, and the number of web site visitors in each month. If these look realistic and achievable, they can then be used to drive the appropriate sales and marketing spend.
    I hope this is helpful. Thanks for your kind comments!
    Best, David

  • Mike Viney


    What’s the biggest problem you have with calculating your SaaS metrics using this method? ie. Time sink to input data to get updated metrics?

  • T.J. Westerhaus

    David – our commercial model is a yearly subscription plus monthly user fees. I’ve used the ACV spreadsheet to model our biz. Not sure for is I should divide these monthly fees by 12 or enter as is for Expansion ACV. Thanks for your work here. Look forward to your reply.

  • David Skok

    Hi TJ – sounds like you should multiply them by 12 to get the expected annual value of them. ACV = Annual Contract Value.
    Best, David

  • T.J. Westerhaus

    Thanks David. What has been your rule of thumb for the calculations in the spreadsheet that depend on a Churn %, when Churn % is zero? Also, to follow up on my previous question; the number of users fluctuates month to month (upward trend thankfully) as the majority of the users are sub-contractors who come and go based on the customer’s project requirements. When I multiple by 12, I get huge ARR that does not seem correct. When I enter as is (for example $60k in one month vs. $720k for 12 months), the ARR seems reasonable to what we are currently experiencing on a cash basis. Do I still have this wrong? It seems the user fees are more like MRR and the annual license fees are ARR. Any suggestions on how to model this combo play?

    Thank you!

  • Sean McAllister

    Thanks David

  • David Skok

    For the first question in your email, “What has been your rule of thumb for the calculations in the spreadsheet that depend on a Churn %, when Churn % is zero?” – can you let me know which particular calculations you are concerned about?
    For the second question, my apologies, I didn’t understand that these sub-contractors come and go, and were not a truly repeatable, recurring revenue stream. Given what I now understand, I would not consider these users as part of the recurring revenue. For the LTV calculation, you might be able to look across all of your customers and predict the average annual revenue from those subcontractors, and use that as a predictor of future value from each account. But if the amount varies greatly from one customer to the next, this could be too unreliable. In which case perhaps use a sub-contractor $ number that is conservative, based on the lowest performing accounts.
    I hope this time I have it right, and that the answer is what you’d expect. If not please don’t hesitate to ask again.

  • Scott Powers

    David – Thanks for the comprehensive article. How do you recommend companies report OEM royalty streams within the traditional subscription model. ACV bookings, churn, etc. work for the subscription piece of the business, but the OEM royalty streams are an outlier.

  • David Skok

    Scott, are the OEM royalty streams also recurring predictable amounts?

  • Scott Powers

    Thanks for the reply. No they are royalties based on a percent of the OEM’s quarterly sales. Further, the OEM is selling on a perpetual license model.

  • David Skok

    Scott, in that case they won’t fit into this framework. I would recommend tracking them separately from the predictable recurring revenues, and using this framework only for the recurring revenues. Does that sound like it will work for you?

  • Beckmania

    David – thank you SO much for this post, as well as all the others. I have learned more from these posts than from anything else I have read online! I have a perhaps down in the weeds question. In the spreadsheet, Bookings is calculated as #of new customers * ARPA for new customers, which makes sense, but then you multiply by Avg months paid up front. Why? I thought bookings would be the size of the deal. Months paid up front affects revenue and billings, but not bookings, no? Thanks! JB

  • David Skok

    Hi John, sorry for the confusion. You are 100% correct. The formulae in the spreadsheet were never meant to be used by readers. I had tried to make that clear, but obviously not well enough. I wanted to create some data that looked roughly realistic, and to do that I had to work backwards to get to some of the numbers. That will explain why this doesn’t make sense to you.

  • Beckmania

    Thanks David – that helps a lot. I am going to try to introduce this dashboard at my firm and was going over the spreadsheet cell by cell to make sure I understand it. I was already detecting that some cells that would clearly be inputs you had as formulas, so your response makes perfect sense!

  • Scott Powers

    Thanks. That works for me and confirms how I’ve seen it done. Thanks very much.

  • Betty

    David – this is a great article! For ACV, how would you recommend handling contracts that are less than a year? Our business is mostly annual with some longer term contracts, so ACV makes more sense than MRR, but there are some cases where we have contracts less than a year. For example, if you had a 6 month contract, would you recommend multiplying it by 2 to get an annual value, or would you recommend leaving it as 6 months?

  • David Skok

    Hi Betty, In the line where it refers to ARR, I would track it at the annual value (i.e. multiplied by 2) provided you have a good renewal rate. But for Bookings (a number that has lower value in a SaaS business), I would only track it at the 6 month value. Best, David

  • Fionn OKeeffe

    We sell different editions or license types of the same underlying product. Is there value in CAC by edition or license type? What advise do you have with regards to effectively allocating edition wide sales and marketing costs to each edition CAC?

  • David Skok

    I think it is most useful to split your customers into segments that identify people that have similar buying behaviors and product requirements. Then, if you have the resources, track CAC and LTV for those different segments. You will quickly discover where you are making the most money, and where you are spending the most money. That can help you allocate your marketing, sales and development resources. If that happens to also align with the different editions, then that would make things easier.

  • David Chen


    Thanks so much for the spreadsheet! It’s really helpful for us to set up our tracking system.

    Thanks so much for your spreadsheet. It’s really helpful. I do have a question about one particular calculation. I’ve a question regarding one of the calculations.

    The way you calculate trial-to-purchase conversion is by having “new customer of the month/trials in progress of the month”. What if there’s a 2 week trial period. Say someone sign up for the trial on Jan 25th, and wouldn’t convert to pay (or churn) until Feb 10th. The calculation by this model will effectively use customers who sign up for trial in January to divide the new trials in February. That seems to be a data source inconsistency, right? It would be great if you can give me some advice on how to resolve this. Thanks a lot!


  • David Skok

    Hi David, I had not intended the calculations in the spreadsheet to be used, as they don’t make sense in a real world scenario. I just used them to create some sample data, which means that some of the formulae work backwards.
    You are right that would create an inconsistency. You will want to look at a way to measure this that is consistent based on when the trial started, or when it ended. I hope that helps.

  • David Chen

    Thank you David! I think I found a way to reconcile this. Again, can’t thank enough for your wisdom. I look forwarding to reading all your future blogs!

  • Fionn OKeeffe

    Our new customers, from one month to another, can have very different MRR’s such that our LTV from one month to next can vary significantly. If we wanted to present the best view of our performance should have take an average over a period of time? We already are using a “group av” churn rate where we have no actual churn in a month.

  • Fionn OKeeffe

    When we sell our service the customer locks in into a license fee structure. On any given month the customer can have a variable bill by virtue of variable components within the license – think seats – and also a variable number of associated transactions charges i.e. sending messages etc. We consider the transactional charges element part of MRR, but it can go up and down – even after accounting for churn and new, so in effect our “expansion MRR” can go both ways! Your examples have positive expansion MRR… This variable MRR means than from one month to the next we have a diff APRA on the installed base which then gives varying LTV – and this mainly due to the variable transaction charges component. Should i take an average view over a period. Also how should i calculate LTV with growth …apply an average growth rate over the same APRA’s and then take an average over say the last 12 months? To provide some context, the transactional charges component of revenue is now greater than the license component so removing it would greatly reduce our MRR.

  • Fionn OKeeffe

    For the LTV with growth calculation where there is much fluctuation, is it OK to use longer term average of MRR Expansion %.

  • David Skok

    Yes – that is the best way to deal with a lot of fluctuation.

  • Troy Bourdages

    Hello David,

    I simply picked a comment with the most recent date as I haven’t found a discussion related to my inquiry…

    Regarding your SaaS salesforce economics spread sheet, specifically the “Sales compensation and overhead” section. The area for Variable Compensation (I assume meaning commission), what does “with 50% draw for first four months” mean exactly?

  • David Skok

    50% draw means that the new sales people are getting paid 50% of the commission that they would get if they were on target. It is not always necessary to pay this. But I do often see some level of financial help while new sales people are going through training and are still ramping.
    Best, David

  • Troy Bourdages

    Got it! Thanks David.

  • David Skok

    Hi Fionn, this definitely complicates things. I think you are best off looking at things as follows:

    1. Figure out the customer churn rate (not the dollar churn rate), and use this to calculate the average customer lifetime.
    2. Look at all of your customers using a period of time that averages out all these ups and downs. This might be six months, or possibly 12 months.
    3. For each 6 or 12 month period, look at the average of all your customer cohorts, and see where their revenue in that period is relative to the first period. Hopefully it will be growing. (Note: the cohort revenue should also include lost revenue from customers in that cohort that churned.)
    4. Now to calculate LTV, you can do a simple model that adds up the cohort revenue over the predicted average lifetime, using the data you got for how revenue evolves over time.
    I hope this helps.
    Best, David

  • Paul B. Silverman


    Excellent article- thanks for sharing.

    Several months ago I was invited to do a guest blog post and serve as an Advisor for Funding Profiles, a Santa Clara-based company offering a powerful suite of financial analytic tools that “integrates with existing business applications to continuously translate traditional financial metrics into the language of business strategy”. For companies with thousands of products, infrastructure, and processes spanning the globe, the ability to ‘drill down’, examine ‘what-ifs’, and assess how and if global LOBs meet KPIs and support the strategic plan, is a powerful planning tool. Your post reinforces this point.

    But markets and technology are moving quickly, consumer power is increasing, and external global factors will impact all global businesses which creates risk and uncertainty. In fact, one study shows macro-environment, competitive and corporate positioning factors account for about 80 percent of ROA variation among LOBs. So optimizing the company’s internal resources, processes, and KPI’s really addresses only 20 percent of the planning challenge based on these findings. My post “How Analytics is “Raising the Bar” for Corporate Strategy:

    Understanding the External Environment” talks about how new analytic tools can provide a competitive edge, creating what Tom Davenport (Author- Competing on Analytics: The New Science of Winning) defines as “analytic competitors”.

    Bottom line- while emphasis of the above are larger corporations, my
    view is entrepreneurs that also understand how to analyze markets, external
    opportunities and threats, and how to use analytics with Porter’s Five Forces
    Model, STEEP, and competitive benchmarking tools can achieve a competitive edge. Properly used, external market analytics provide a competitive edge for evaluating, strategy positioning, and managing entrepreneurial ventures.

    During the past six months, I have looked at ventures in areas of wearable healthcare monitoring devices, clinical analytics, legal analytics, analytics for fraud detection, and solar energy among others. To accurately gauge outlook and opportunity for these and others, venture evaluation must go well beyond the typical “size of market, expected market share” and ‘drill-down’ to understand
    external market threats and opportunities. We have a way to go yet in educating
    the entrepreneurial community but I believe today’s “hyper competitive” dynamic
    global markets will help accelerate the adoption of these new analytics
    capabilities. You can read my entire post at

    Paul B. Silverman is an Adjunct Professor in the R.H. Smith School
    of Business at the University of Maryland, former CEO of public and private
    companies, and Managing Partner Gemini Business Group, LLC. He can be reached at paul@PaulBSilverman .com or blog at, twitter at @globalbizmentor.

  • Thibaut Taittinger


    Thanks a lot for a great article! It is very useful to many of us. Two points:

    1) I have not figured out how I should treat a business model which includes both recurring and one off payments. Our One Off payment can happen at any time along the LTV (sales commissions + fully custom premium services). It is not logical to include it in Expansion MRR… So I guess I will have to do some kind of reconciliation by adding monthly non-recurring revenues per user X Customer Lifetime to the LTV? Do you have any advice there?

    2) You have a small typo mistake which makes a link non valid (directly below the % of MRR Retained over time – by Cohord graph)

    It is at the end of this paragraph:
    In the situation above, you will need a more complex formula to calculate LTV,
    as the value of the average customer is increasing over time. For more on that
    topic, you may want to check out the accompanying definitions document.
    (matric should be metric in link).


  • mbrooks

    Hi David,

    If my company is using the annual contract setup, would AVC calculations be average or sum numbers? Thanks!

  • David Skok

    Which calculation are your referring to?

  • mbrooks

    For companies who experience high upsell rates, where is the credit given in the Lifetime Value calculation?? It seems like there are two possible places we might “get credit” in these calcs. The first is using new business plus upsell in the MRR added component. The second seems to be using the AVC expansion formula — but if we experience negative AVC expansion, that would suggest an infinite customer lifetime. Thoughts?

  • David Skok

    I address this issue in the companion article “SaaS Metrics – Detailed Definitions”. What actually happens is that at some stage, the losses from a particular cohort due to churn will outweigh the expansion revenue. There is a formula in there that attempts to account for this phenomenon. Let me know if that meets your needs.
    Here is the link:

  • M

    Excellent write up. Dissects & explains the subtle intricacies of SaaS.

  • mbrooks

    How would you go about calculating “monthly growth rate in ARPA by account” ?

  • David Skok

    Good question. I would take a look at all accounts in your tracking system (either CRM or accounting) and for those that have remained with you, I would look at how much they have expanded and how long they have been with you. So for each account, you would have a monthly growth rate (even though the expansions won’t have happened every single month) computed as:
    Total growth % since sign up / Months since sign up

    Then to get to “Monthly growth rate in ARPA by account” I would simply create an average of all these, (across all remaining accounts).

  • mbrooks

    What if you have a growth rate that isnt steady? For example:

    Customer A: yearly contract, 12 months steady ARPA
    Customer B: yearly contract, 6 months steady, 6months 20%growth
    Customer C: yearly contract, 6months 10% growth, churn at year end
    Customer D: yearly contract, no growth, churn at y/e

    How do you account for the growth? In the formula in the article, we take a double hit for the churn here when we calculate the monthly account growth and when we calculate the churn rate. Using the M(1-c)/c2 formula wouldnt work.. what should we do?