Most important for Software as a Service? Isn't it the financial statements, EBIT, Gross Profit, FCF? Well, no. Your accounts are important, of course, but it's difficult to manage your business if you only look in the rear-view mirror. The fact that you have money on the books today does not guarantee that you will have it in a year's time. So we need to put in place measurements and indicators that help us plan for the future!
(PS: If you're new to B2B SaaS as a business model, it's worth reading this article first)
The key metrics we are going to go through here are important for many reasons, for example:
Understanding future capital needs. Is there enough money in the coffers to get through the next sales cycle?
Adjusting market and product strategies. Which market segments are performing best and where should we invest further in developing our product portfolio?
Assess the effectiveness of our marketing efforts. Are we getting value for money from our current activities, or do we need to change how we prioritize time and resources?
Prepare budget and scale up organization. What will our revenue look like over the next 12 months, and are there parts of the organization that need to be scaled up to handle it?
Due Dilligence. You're bringing in a new investor, you say? If so, many of the key figures here will be requested in any DD process.
Below we ask ten questions and explain how these can be answered with well-known “SaaS Metrics”. The figures in points 1-5 are essential. Points 6-10 are for slightly more advanced SaaS businesses.
Does this sound almost too easy? In principle it's easy, the problem is often data quality. Good customer and subscription data is often difficult to put in place without good discipline around data quality in the organization.
1. What is the value of your current customer base?
ARR | ACV |
ARR (Annual Recurring Revenue) is the (assumed) total value of all existing customer agreements for the coming 12 months, excluding non-recurring revenue. | ACV (Annual Contract Value) is the annual (assumed) contract value for an individual customer, excluding non-recurring revenue. In other words, ARR is the sum of your customers ACV. |
ARR and ACV are the cornerstones of any SaaS analysis. ARR gives you the (assumed) total value of all customer deals for the next 12 months, while ACV represents the (assumed) annual value of a single customer deal. With value we mean recurring (subscription revenue), all one-off revenue should be removed from the equation.
These figures are important because they provide an early indication of both future revenue and company growth. Regular monitoring of ARR through monthly reports can reveal trends and provide the basis for important strategic decisions.
In the example below, you can see ten fictitious customers, ACV and how ARR has been calculated:
Customer ID | Company | ACV |
1 | CloudGate Solutions | 250 |
2 | DataNest Analytics | 450 |
3 | EcoStream Tech | 325 |
4 | NeuroLink AI | 500 |
5 | QuantumSafe Security | 750 |
6 | BrightFiber Networks | 600 |
7 | AgriGrowth Labs | 400 |
8 | Urban Mobility Solutions | 200 |
9 | DeepBlue Technologies | 675 |
10 | ConnectHR Systems | 300 |
ARR: |
| 4 450 |
A figure that tells us a lot in itself. If nothing changes, we expect 4,450 in revenue from these customers over the next 12 months.
But: for these figures to be of real value, it's important that we take regular measurements, for example once a month:
More about this below.
2. How has your customer base developed over the last 12 months?
NRR |
NRR (Net Retention Rate) measures the development of your customer base over time, usually over a year. It gives an indication of how your existing customer base will develop based on historical values. |
NRR tells you how well you are retaining and developing your existing customer base. By analyzing how the ARR for a particular customer group changes over time, you can find out whether the trend is positive or negative.
To calculate NRR, you take what the ARR was 12 months ago and compare it to the ARR for the same customers today. How has it changed?
In the example above, we see that we have an NRR of 122%, which means that among the customer base that existed 12 months ago, we have had 22% growth.
So we know that we have a positive development in our customer base. But what was it that contributed to this? There is a better way to calculate NRR, lets continue down the rabbit hole.
3. What factors contributed to the development of your customer base in the last 12 months?
Churn | Upsell | Downsell |
Churn means customers who have terminated their customer agreement. | Upsell is an increase in the value of an existing customer agreement. | Downsell is a reduction in the value of an existing customer agreement. |
Understanding churn (lost customers), upsell (value increase from existing customers), and downsell (value reduction from existing customers) is important for nuanced insights into your customer base. This enables you to design initiatives to manage customer loss and maximize the value of the customers you retain. By analyzing these factors, you can develop more targeted customer service and marketing activities.
In the example above, we can easily categorize customers into Churn, Upsell and Downsell. And by adding these together, we get the same NRR as above.
If we then add data on new sales for the last 12 months, we have a complete explanation (or bridge) between old and new ARR.
Let's end it all with a waterfall diagram. Not only is this a great way to illustrate the ARR development in the monthly management report, it also makes it easy to identify which factors are driving growth and where there are challenges.
This is a good starting point for a discussion in the management team. It looks like the company is good at value-added sales! But we have almost as much churn as we have new customers. What should we do about it?
(The figure does not include price adjustments or currency changes. It's important to adjust for these effects in order to separate operational and financial effects on customer development).
How well are customers adopting our services?
User Adoption | Feature Adoption |
How many users have adopted the solution as a percentage of what the potential is. E.g. a customer has 50 MAU (Monthly active users) but has purchased 100 licenses. UA is 50%. | How much a module or function in a product is adopted by users, or how advanced a customer's use is. For example, out of 1,000 active users, only 8% use the new report module.. |
Exactly how you measure adoption needs to be adapted to your services and your business model. If you (for example) sell monthly subscriptions, it will look different than if you sell annual ones. And then we need to measure over time. When we get historical data in place, it becomes easier to find correlations.
Does the customer relationship last longer with better adoption?
Can we see early signs that a customer is at risk from adoption data?
Is the customer more satisfied if the usage is more advanced?
Is it easier to upsell where adoption is high?
Below is a simple example where we have compared ACV and UA on four of the customers from the example above.
What can we deduct from the numbers above?
Cloudgate has stable and high usage, but no growth. At the same time, they are paying more for the solution. It's unlikely that this customer will need more licenses in the near future, and if the price continues to grow, it could become a challenge. Perhaps it's time for the account manager to have a chat with the customer?
DataNest's usage had already started to decline at the start of the year. The decline continues, and the customer gives notice in August. It's too late to do anything about this, but if you had entered into a dialog with the customer in January, you might have managed to keep the customer by renegotiating the terms of their agreement.
DeepBlue is persistently hitting the “ceiling” on its usage. The number of licenses is increasing rapidly. This customer is in a on a good path! Is there a good story here that could be turned into an article on the website?
Innovatech Solutions is a new customer from June. But as we can see from the adoption figures, they probably haven't managed to get started with the solution as they planned. Thee Customer Success team needs to get in touch to see if they can help! Customers who buy but don't succeed quickly disappear.
GreenOptima is also a new customer from June. Here we see that they have managed to implement the SaaS solution in the organization. Perhaps they are so enthusiastic about the solution that a conversation about the future and renegotiation of the agreement to a larger volume is in order?
As you begin to understand the relationships, you may want to create simple dashboards that give your organization quick insight into the “health” of the various customer relationships.
(You don't need to be a data analyst or AI expert to make this happen, but it helps 😊)
It's important to stress that adoption data isn't just about informing the commercial side of your organization! This is also key for your product team. Combined with qualitative data (from research), adoption data can also be used to identify areas for improvement in the service, which in turn can help with customer growth and lifetime value.
5. How satisfied are our customers?
NPS (Net Promoter Score) |
On a scale from 0 - 10, how likely is it that you would recommend us to friends or colleagues? |
NPS is a measure of customer satisfaction and loyalty. It's very important for SaaS companies, where the greatest value lies with your existing customers. NPS is calculated by asking customers how likely they are to recommend your services to others (also called the ambassador question).
Responses that rank as 9 or 10 indicate loyal promoters, while responses from 0 to 6 are considered detractors. A high NPS indicates that your customers are satisfied and likely to continue using and recommending your product.
Calculation: NPS is calculated by subtracting the percentage of detractors from the percentage of promoters among survey respondents.
Example: If you ask 100 customers if they would recommend your company, and 43 respond with 9-10 (Promoters), 30 respond with 7-8 (Passives) and 12 respond with 0-6 (Detractors), your NPS would be (43 - 12) = 31.
What is a good NPS score? Unfortunately, this is not a magic number that provides an objective truth. It will vary depending on the type of customers you have, the industry you work in and not least the country you are in. In a survey published by PE company Monterro in 2024 among 298 executives in B2B SaaS companies, the average was 46. Not surprisingly, it was the smallest companies that had the highest NPS.
But NPS isn't just important for comparing yourself with others! You can use the NPS results to engage with customers that are rating you low. And you can identify good ambassadors who can be used more purposefully in the marketing of your services.
By asking the same question year after year, you'll have a very good insight into the direction in which satisfaction is moving, and you'll find out whether initiatives or investments in your organization and your services have had an effect on your customers.
6. What is a customer really worth?
LTV (Customer Lifetime Value) |
Revenue generated from an average customer over the entire contract period. |
LTV is perhaps for the more advanced SaaS companies that have established a slightly larger customer base. But once you're in this phase, it's one of the most important key figures and can be used to identify measures that lead to further growth.
SaaS is not about charging the highest price, but having sustainable prices that keep the customer happy and provide the highest profit over their entire lifecycle. And the longer you retain a customer, the more you can invest in gaining that customer (more on that later).
LTV can be calculated by multiplying the average revenue per customer by the average customer lifetime value (usually the number of years a customer remains with the company) and subtracting the variable costs associated with serving the customer. Here you may want to segment customers into different groups. LTV can be very different for large customers (who don't change solutions very often) and smaller customers.
A simple example of calculating LTV: If the average monthly revenue per customer is 100, the average duration of the customer relationship is 3 years, and the annual variable costs per customer are 200, the LTV would be (100 12 3) - 200 = 3600 - 200 = 3400.
A slightly more complicated example: In the example above, we assume that life expectancy is 3 years. But we can also use expected churn as a starting point (see point 3). Let's say that expected churn is 15% in our customer base. Then the LTV in the example above is ((100 *12 ) - 200)/ 15% = 6700.
I am sure a lot of economists reading this scoff at the lack of a present value calculation. That's absolutely right! If you expect your costs to rise faster and prices to grow more slowly than your discount rate, then the calculation needs to be adjusted. But then you have a problem with your pricing model and cost control, and the viability of your entire business model can be questioned.
Once you start to understand your LTV and have a good insight into the data, there are many interesting questions you can ask yourself and perhaps experiment with.
If we reduce prices by 5%, does the LTV increase or decrease?
If we make an investment of one million in the product, does the LTV grow sufficiently to cover the investment?
If we become more proactive with customer success, does that have an impact on LTV?
LTV in combination with other key figures can provide even more insight. More on that below.
7. What does it cost us to aquire a new customer?
CAC (Customer Aquisition Cost) | CAC-payback | LTV/CAC Ratio |
The cost of aquiring a new customer | How long you need to keep the customer to pay back the CAC cost. | How many times will the CAC be repaid during a customers lifetime? |
The CAC metric is easy to understand, but can be difficult to calculate. CAC is calculated by dividing all costs associated with acquiring new customers (marketing and sales costs) by the number of new customers acquired in the same period.
Simple in principle (if you have a good way of separating sales costs from the rest of your operations), but with long sales processes or varying sales cycles it can become more inaccurate. Either way, the most important thing is to choose a method and stick to it over a longer period of time so that you can see trends.
CAC payback
How long does it take for all the costs associated with the sale to be repaid? In other words, CAC payback is CAC / (annual revenue - annual variable costs associated with serving the customer). It provides an awareness of the time it takes to earn back the money it cost to land a deal.
LTV/CAC ratio.
But how long the CAC payback is doesn't necessarily tell the whole story. You can tolerate a longer payback period if the customer relationship is long-term. So in combination with LTV, it becomes even more interesting. And it is naturally calculated by dividing LTV by CAC. The result is simply a factor that says something about how many times the sales cost you get back from the customer.
What is a good LTV/CAC? Anything below 1 is of course a disaster. In that case, it costs more to get the customer than you get in return. Then it depends on how big an organization you need to serve the customer, and what kind of cost you have operating and maintaining your product. But you probably need to be somewhere between 4x - 8x for profitable growth.
7. How much does it cost us to build a pipeline of potential customers?
Pipe-to-Spend |
The relationship between (marketing) costs and potential revenue from your sales pipeline. |
Pipe-to-Spend measures the ratio of potential revenue in the sales pipeline to the actual marketing costs spent to generate that pipeline. It is calculated by dividing the total potential revenue in the pipeline by the total cost of generating it.
For example, if the total revenue potential in the sales pipeline is 500,000 and the marketing and sales costs are 50,000, the Pipe-to-Spend ratio would be 500,000 / 50,000 = 10. This indicates that for every £1 spent, £10 in revenue is potentially generated.
Here it is important to underline the word potentially. Of course, it depends on the conversion rates in the rest of the sales cycle (see next point).
It is impossible to give a general rule for what Pipe-to-Spend should be. It depends on what type of sales your organization is doing and what the criteria are for a lead to become a pipeline.
But if you measure Pipe-to-Spend over time, you can see whether the effectiveness of marketing initiatives is going up or down, and if you also measure it at campaign level (e.g. a major industry trade fair, or ads on LinkedIn), you can gain a lot of insight into how well your marketing efforts are hitting the mark and what type of marketing you should focus on.
Pipe to spend is a great KPI for your marketing organization. It's easy to have a high level of activity without it generating results. But when you measure the effect on the pipeline, you become better at focusing on the activities that create the best outcome.
9. How well are we converting leads to customers?
Lead Conversion Rate |
How successful are we at converting a lead to a closed deal through the different stages of the sales process? |
In its simplest form, a lead conversion rate can be created by dividing the number of successful sales by the total number of leads (as a %). So if a company generates 500 leads and 50 of these are converted into customers, the conversion rate would be (50 / 500) * 100 = 10%.
But if you're doing B2B sales, you often have a slightly more complicated sales process! Below is a hypothetical example of what a simple process might look like for a B2B organization focused on inbound marketing, and what their conversion rate is:
Stage: | New Lead | Marketing Qualified Lead | Sales Qualified Lead | Pipeline | Won |
Description | Aquiring new leads | Leads that show interest for your services. | Leads that are in the marked to purchase. | Leads that are actively being sold to | New customer! |
Fokus: | Creating awareness | Build interest and engagement | Qualify the budgets, needs and timelines | Active prospecting/sales | Customer Success |
Example | 1000 | 300 | 100 | 50 | 10 |
Convertion rates |
| 30% | 33% | 50% | 20% |
The total conversion rate is then 1000 * 30% * 33% * 50% * 20% = 1%.
Again, a number that is impossible to say is good or bad - in itself. But it does highlight the importance of working on both lead acquisition and improving conversion rates. And combined with a little insight into the organization, you may quickly see where in the sales process there is most room for improvement or investment.
If you're in a “bottomless market” where it's easy to get new leads, you might want to focus on increasing the number of leads. In a mature market where there are limited customers, you would rather optimize the later part of the sales process.
And not least, good data on this (in combination with the next point) makes it easier to predict future sales revenue.
10. How long does it take to close a deal?
Time To Close |
Average time from a new lead to becoming a customer. |
Time to Close brings a time dimension to the conversion rates from point 9. We simply look at the average time a lead that leads to a sale spends in the different stages of the sales process.
Stage: | New Lead | Marketing Qualified Lead | Sales Qualified Lead | Pipeline | Won |
Example: | 1000 | 300 | 100 | 50 | 10 |
Conversion rate |
| 30% | 33% | 50% | 20% |
Time spent | 30d | 10d | 15 d | 60 d |
|
Total |
|
|
|
| 115 days. |
In the example above, the average time from lead to customer is 4 months. This can give you an indication of what your sales will be in the coming months. Let's say you have a lead database that currently looks like this:
New Lead | Marketing Qualified Lead | Sales Qualified Lead | Pipeline |
500 | 150 | 50 | 25 |
How many and when do these sales come? Let's say the calendar says January 1, and on average our leads are in the middle of the stages above.
| Conversion | No of sales | Closing in | Sales Jan | Sales Feb | Sales Mar | Sales Apr | Sales May |
New Lead | 500 x 1% | 5 | 15+10+15+60= 100 days | 0 | 0 | 0 | 5 |
|
MQL | 150 x 3% | 4,5 | 5+15+60= 80 days | 0 | 0 | 4,5 | 0 |
|
SQL | 50 x 10% | 5 | 7 + 60 = 67 days | 0 | 0 | 5 | 0 |
|
Pipeline | 25 * 20% | 5 | 30 days | 5 | 0 | 0 | 0 |
|
SUM |
|
|
| 5 | 0 | 9,5 | 5 |
|
Note that this approach works poorly after the first four months. Since the average time from lead to customer is 115 days, it means that there are leads you don't yet know about that could become customers in April.
By the way, this method does NOT replace a good sales forecast, especially when it comes to the later stages of the sales process. Here we only look at averages, and the variations in the customer base can be huge. Adding a layer of insight from the sales organization through a forecast will increase accuracy - if done correctly.
Typical pitfalls for B2B Software as a Service
Having worked with many different SaaS metrics in an international SaaS company for a couple of decades, I've seen most of the pitfalls you can stumble into when analyzing these metrics (and others!). The examples above are simple, but the world is complex. So please keep in mind:
1. Quality in you data sets does not magically happen. It must be incentivized and facilitated. The most important thing is that managers are good role models in the process and that they lead the way by using data as a basis for discussion and decisions, and are standard-bearers for discipline in data quality.
2. There is always an exception. It's often a new sales leader or account manager who wants to make changes to the process or data model for their market and customers. Changes that in themselves are sensible enough. But remember that this creates more complexity and makes it harder to compare new data with historical data. Be a gatekeeper against changes that can create unintended consequences.
3. Customers and products are not necessarily comparable. They come in different sizes, industries, sizes and perhaps nationalities. Make sure that your customer data contains a good segmentation model so that you can analyze data at segment level. If you sell different subscription products, you must of course also have segmentation for this.
4. Think carefully about possible challenges when setting KPIs for key data. Of course, many of these metrics should be KPIs for managers, salespeople and marketers. But be aware that metrics can be influenced without creating real value. For example, if you incentivize an increase in leads, you'll most certainly see an increase in numbers. But then the quality of new leads may also decrease, and the number of sales will be the same as before.
Good luck!