Feb 5, 2023·edited Feb 5, 2023Liked by Reid DeRamus
Thanks for the great post!
The calculation formula for the Blended curve for month X is (Total retained in month X across 6 cohorts) / (Total customers acquired in 6 cohorts). Is this correct? Could you please explain a little on the usage of SUMPRODUCT in your excel?
Thanks for the kind words, Deepak!! And appreciate the question - it's a good one!
You got the math right for the blended curve. The key is to *only* look at subscribers that have had the opportunity to pay a certain amount of times. If we have 100 paid subs that have only been around long enough to pay 2 times, we don't want to include them in our month 3 retention rate (or any month beyond that).
The SUMPRODUCT is really just a shortcut. We can get to the same blended curve by adding up all the paid subs that have reached a certain month or successful payment, and then dividing that by the sum of all the paid subs that had the opportunity to pay the same amount of times.
I created a copy of the CLV workbook with the above broken out. The link below should take you to the right cell, but let me know if you have any questions — hope this helps!
I love the tools and step by step process you laid out to measure your CLV and retention rates. However, my mind immediately went to what happens when your curve is decaying. How do you pinpoint what is causing the churn and what do you do about it....you know an easy question ;)
This is all just foundational stuff — the actionable stuff comes from segmenting and getting the bottom of what's moving the curve in either direction.
We could focus on month-1 retention (where the curve is the steepest) and compare behavior for those that make it through vs. those that cancel. That can give us some signal on behavior we want to try to encourage (usually something around frequency, depth, or diversity of engagement).
We could also look at early-curve retention (month 1, 2, 3) vs. outer-curve retention (% making it past month 12, month 24). For earlier-tenure subs, we can explore whether we're doing a better / worse job of onboarding new subscribers and making sure they understand the value prop. For more tenured subs, we could look into whether we're moving further toward / away from the original value prop that brought in early subscribers.
There's also a big difference in voluntary cancel (people choosing to turn off auto-pay) vs. involuntary cancels (people that cancel due to payment issues). The latter can actually be a significant portion of cancels & requires different tactics (there's a whole cottage industry around this stuff: Zuora, Chargify, Recurly, Chargebee, etc.).
I'll be rereading all your posts in Growth Croissant, probably a third or even fourth time each, Reid. It is quite complex, eh? Especially for a small fry like me. I can only execute what you're presenting to us if I radically simplify it, which will take me some time. Any thoughts on streamlining it for those of us who have micro-businesses?
Thanks, Reid. I very much appreciate what you're sharing with us. It's outstanding information. I started New Florence on December 12 and am happy to have my first 50 subscribers, including my first 10 paid subscribers. I'm thinking long-term about it -- reaching 1,500 paid subscribers and 100 paid subscribers by the end of 2024. So modest goals. You're paradigm and approach will definitely help me get there, but you can see why it might be a little overwhelming for a guy with modest goals, eh?
I wouldn't consider 1,500 paid subs a modest goal. Tons of writers (and even media companies!) would love to reach that threshold.
Given you just launched, I think the most important thing is to get in a consistent habit of publishing, trying new things (I loved the "early support and feedback" email you sent), and figuring out your audience. (These are all things I'm personally going through too.) Those things are more important than modeling out CLV, but I do hope this framework is helpful. At some point down the road, perhaps you'll catch yourself making a decision and thinking all this CLV stuff helped in some way.
Yeah, definitely, I'm going to memorize things you're writing, kind of immerse myself in your approach, and then things will arise in my mind at various points on my Substack journey, and so my own scaled-down version of your complex paradigm will end up being very decisive for me in subscriber growth and managing what I'm doing.
Thanks for the great post!
The calculation formula for the Blended curve for month X is (Total retained in month X across 6 cohorts) / (Total customers acquired in 6 cohorts). Is this correct? Could you please explain a little on the usage of SUMPRODUCT in your excel?
Thanks for the kind words, Deepak!! And appreciate the question - it's a good one!
You got the math right for the blended curve. The key is to *only* look at subscribers that have had the opportunity to pay a certain amount of times. If we have 100 paid subs that have only been around long enough to pay 2 times, we don't want to include them in our month 3 retention rate (or any month beyond that).
The SUMPRODUCT is really just a shortcut. We can get to the same blended curve by adding up all the paid subs that have reached a certain month or successful payment, and then dividing that by the sum of all the paid subs that had the opportunity to pay the same amount of times.
I created a copy of the CLV workbook with the above broken out. The link below should take you to the right cell, but let me know if you have any questions — hope this helps!
https://docs.google.com/spreadsheets/d/1mZkS7mSfySwB1ZBvxkGnia27HDCgZapJyDvSc6iCcXM/edit#gid=463378505&range=AB6
Another helpful post - thanks Reid
I love the tools and step by step process you laid out to measure your CLV and retention rates. However, my mind immediately went to what happens when your curve is decaying. How do you pinpoint what is causing the churn and what do you do about it....you know an easy question ;)
Yes, absolutely!
This is all just foundational stuff — the actionable stuff comes from segmenting and getting the bottom of what's moving the curve in either direction.
We could focus on month-1 retention (where the curve is the steepest) and compare behavior for those that make it through vs. those that cancel. That can give us some signal on behavior we want to try to encourage (usually something around frequency, depth, or diversity of engagement).
We could also look at early-curve retention (month 1, 2, 3) vs. outer-curve retention (% making it past month 12, month 24). For earlier-tenure subs, we can explore whether we're doing a better / worse job of onboarding new subscribers and making sure they understand the value prop. For more tenured subs, we could look into whether we're moving further toward / away from the original value prop that brought in early subscribers.
There's also a big difference in voluntary cancel (people choosing to turn off auto-pay) vs. involuntary cancels (people that cancel due to payment issues). The latter can actually be a significant portion of cancels & requires different tactics (there's a whole cottage industry around this stuff: Zuora, Chargify, Recurly, Chargebee, etc.).
Thanks for the comments Rob! 🙏
I'll be rereading all your posts in Growth Croissant, probably a third or even fourth time each, Reid. It is quite complex, eh? Especially for a small fry like me. I can only execute what you're presenting to us if I radically simplify it, which will take me some time. Any thoughts on streamlining it for those of us who have micro-businesses?
Thanks, Reid. I very much appreciate what you're sharing with us. It's outstanding information. I started New Florence on December 12 and am happy to have my first 50 subscribers, including my first 10 paid subscribers. I'm thinking long-term about it -- reaching 1,500 paid subscribers and 100 paid subscribers by the end of 2024. So modest goals. You're paradigm and approach will definitely help me get there, but you can see why it might be a little overwhelming for a guy with modest goals, eh?
I wouldn't consider 1,500 paid subs a modest goal. Tons of writers (and even media companies!) would love to reach that threshold.
Given you just launched, I think the most important thing is to get in a consistent habit of publishing, trying new things (I loved the "early support and feedback" email you sent), and figuring out your audience. (These are all things I'm personally going through too.) Those things are more important than modeling out CLV, but I do hope this framework is helpful. At some point down the road, perhaps you'll catch yourself making a decision and thinking all this CLV stuff helped in some way.
Yeah, definitely, I'm going to memorize things you're writing, kind of immerse myself in your approach, and then things will arise in my mind at various points on my Substack journey, and so my own scaled-down version of your complex paradigm will end up being very decisive for me in subscriber growth and managing what I'm doing.
Would love to hear any questions that come to mind along the way!
As always, very much appreciate your support. 🙏
Oops, I meant to write 1,500 subscribers including 100 paid.