In today’s issue, I’m going to take on modeling considerations for subscription-driven businesses, specifically SaaS (software-as-a-service).
On the surface, recurring revenue models seem superior to non-recurring (one-time) models. While this is mostly true, it isn’t always true.
Let’s dive in.
Traction Roadmapping - Direct Recurring Business Models
The inputs to modeling SaaS models are:
1. Minimum Success Criteria
This is the target ARR (annual recurring revenue) in 3 years. The typical rule of thumb for SaaS companies targeting venture capital is hitting $1m ARR in 2 years and marching towards $10m ARR in 3 years. We’ll assume the same.
2. Pricing Model
This is the monthly or yearly subscription fee for the product. We’ll use a $100/mo product.
3. Churn
This is typically reported as the average monthly or yearly subscriptions canceled. While churn is an easy concept, it’s hard to estimate as a percentage. What’s a good enough monthly churn rate - 1%, 2%, 5%?
For this reason, I prefer using expected average customer lifetime versus churn, which is easier to estimate and has an inverse relation to churn.
Monthly churn rate = 1 / (expected customer lifetime in months)
![](https://cdn.prod.website-files.com/65c4e268c9d10398295b1af5/66689e377914a4d8df3b8120_1.webp)
The average SaaS lifetime is 4 years. We’ll also assume the same.
4. Paid Conversion Rate
Typical SaaS conversion rates range from 0.5-3% on average. When in doubt, I start everyone with 1% and then refine once better data becomes available.
5. Referral Rates
As most SaaS don’t typically grow virally, referral rates tend to be amplifiers vs. significant growth levers. This is a topic deserving of a separate post.
It’s perfectly reasonable to use a 10-40% referral rate. I like starting this out as 0% in the tool and then ramping up to see the difference.
So here are the inputs to the model:
- Model Type: Direct
- Minimum Success Criteria: Generate $10m ARR (annual recurring revenue) in 3 years.
- Pricing model: $100/mo
- Lifetime: 4 years | 2.08% monthly churn
- Paid Conversion rate (lead to paid): 1%
- Referral Rate: 0%
- Growth rate: 10x/year
Here’s the traction roadmap with these inputs:
![](https://cdn.prod.website-files.com/65c4e268c9d10398295b1af5/66689e41ab33ab3810cc6061_2.webp)
Takeaways
1. Active users only tell part of the story
Many founders are quick to calculate the number of active customers needed to hit their ARR goal — 8,333 actives in this case.
This number, however, gives founders a false sense of security because it gives the illusion that once the founders hit this number, the model just works.
But it isn’t that simple.
First, it doesn’t account for the new customer ramp needed to hit the goal. Second, it doesn’t account for churn that, left unchecked, will continue to lose customers.
You can see the first number in the traction roadmap above at the year 3 point. To hit 8,333 active customers at the end of month 36, the customer factory would need to make 1,598 new customers that month.
This number accounts for both the target growth rate (10x/year) and customers lost due to churn:
![](https://cdn.prod.website-files.com/65c4e268c9d10398295b1af5/66689e4ba26d9cb62ab538f1_3.webp)
This number is significant once you account for conversion rates and translate it to top-of-the-funnel traffic (159,800 monthly visitors).
2. The story doesn’t end at the end of year 3
Left unchecked, the customer factory will continue to lose 174 customers per month. And so, the founder needs to set a new growth rate to determine what happens next.
To maintain the goal ($10m ARR), the customer factory must make 174 new customers/mo to account for the leaky bucket from churn. The factory must create more customers if the new growth rate exceeds 1x.
3. A customer lifetime of 4 years doesn’t mean zero cancelations for the first three years
I frequently get asked why the model shows any cancelations in the first three years if the customer's lifetime is four years. Remember above that I use customer lifetime instead of churn because it’s easier to estimate.
The bit that’s often forgotten is this is the average customer lifetime and not an absolute customer lifetime.
In other words, on average, customers will stay for four years, but there will be some who stay for only a year and others who might stay for seven years.
![](https://cdn.prod.website-files.com/65c4e268c9d10398295b1af5/66689e599748920f350cf56c_4.webp)
4. When can one-time payment perform better than recurring subscriptions?
It depends on the product’s value story. Think of it this way.
- Every product targets a customer lifetime value (LTV).
- Subscriptions realize that LTV over time, while one-time payments realize it immediately (better for cashflow)
- With subscriptions, you risk churn, cutting LTV short.
- While with one-time payments, you risk refunds cutting LTV short.
For short customer lifetimes, like 1-3 years, one-time models can outperform subscription models, provided the refund rate and refund window are smaller than the churn rate and customer lifetime.
Here’s a real-world example: Selling a $2,400 online course with a 60-day money-back guarantee versus charging $100/mo with an expected customer lifetime of 2 years.
Both target the same LTV, but more people will churn out sooner than request a refund. It’s a good exercise for you to work out on your own.
That’s all for today.
Until next time,
Ash
P.S. A Coaching Lean Mastery paid subscription includes the traction roadmap tool for personal use.