I like the point on growth models helping an organization make relative tradeoff decisions. I have seen individuals who can intuitively understand the tradeoffs but putting them on "paper" brings everyone to the same level of understanding.
Love the work you and Lenny have done in visualizing what a *real* growth model looks like – and having the spreadsheet you shared in this post is particularly useful.
I was wondering if you have a similar template for the other types of startups mentioned in the post with Lenny – in particular the 3 types of SaaS?
I'm a big fan of your writing. A very common flavor of pitfall #3 is adding too many or too granular dimensions to the model: separating geographies, customer acquisition sources, customer platforms, etc.
It often starts innocently enough - a stakeholder requests an extra dimension that seems reasonable. The change is relatively easy to make - adding a dimension is usually mostly copy/paste. But model complexity grows exponentially with each new dimension, so after only two or three requests, the original simplicity and utility is lost.
The key is being judicious upfront about which dimensions truly matter and flattening dimensions as much as possible (e.g. modeling geo as a binary US/ROW instead of adding every single country). Because while more granularity sounds helpful, it quickly obscures the core insights that make these models useful.
At our firm we use a growth model as an outline to gauge if we’re roughly on track. However there’re just too many intangibles that you can’t possibly take into account, that our time is often better spent actually working on the business opposed to planing on working on the business.
Seen this in a lot of the consulting projects we've been brought into - even for some organisations that should know better. I think it's inherent when trying to make predictions in a complex system using primarily historical data.
One thing we try to do, where possible, is to acknowledge the degree of uncertainty in any model based on the variability of past data. This lets us - as you say above - create a range of scenarios.
Excellent write, Dan. Re: #4 I’ve also found this profile to be rare:
“They must have excellent business judgment and an intuitive sense for the company’s strategy. But they also have to understand the business on a molecular level, and be able to get the right data on each individual input”
And even more rare to find that in someone who is playing an unbiased role in the organization. Inevitably people who have this level of comprehension will gravitate to roles of strong influence & decision-making i.e. they must be “biased” to be effective.
My main point was to avoid people who have an ingoing bias on what to work on simply because of where they sit in the org or what products/channels they own.
I like the point on growth models helping an organization make relative tradeoff decisions. I have seen individuals who can intuitively understand the tradeoffs but putting them on "paper" brings everyone to the same level of understanding.
Absolutely, it’s important to have a shared common currency for making trade offs
Love the work you and Lenny have done in visualizing what a *real* growth model looks like – and having the spreadsheet you shared in this post is particularly useful.
I was wondering if you have a similar template for the other types of startups mentioned in the post with Lenny – in particular the 3 types of SaaS?
Thank you! I’m not planning on sharing anything on SaaS, but Lenny may be, and would also check out Reforge Artifacts.
I'm a big fan of your writing. A very common flavor of pitfall #3 is adding too many or too granular dimensions to the model: separating geographies, customer acquisition sources, customer platforms, etc.
It often starts innocently enough - a stakeholder requests an extra dimension that seems reasonable. The change is relatively easy to make - adding a dimension is usually mostly copy/paste. But model complexity grows exponentially with each new dimension, so after only two or three requests, the original simplicity and utility is lost.
The key is being judicious upfront about which dimensions truly matter and flattening dimensions as much as possible (e.g. modeling geo as a binary US/ROW instead of adding every single country). Because while more granularity sounds helpful, it quickly obscures the core insights that make these models useful.
Absolutely - I’ve seen this failure mode many times.
Hi!!
At our firm we use a growth model as an outline to gauge if we’re roughly on track. However there’re just too many intangibles that you can’t possibly take into account, that our time is often better spent actually working on the business opposed to planing on working on the business.
Seen this in a lot of the consulting projects we've been brought into - even for some organisations that should know better. I think it's inherent when trying to make predictions in a complex system using primarily historical data.
One thing we try to do, where possible, is to acknowledge the degree of uncertainty in any model based on the variability of past data. This lets us - as you say above - create a range of scenarios.
Excellent write, Dan. Re: #4 I’ve also found this profile to be rare:
“They must have excellent business judgment and an intuitive sense for the company’s strategy. But they also have to understand the business on a molecular level, and be able to get the right data on each individual input”
And even more rare to find that in someone who is playing an unbiased role in the organization. Inevitably people who have this level of comprehension will gravitate to roles of strong influence & decision-making i.e. they must be “biased” to be effective.
Thanks Waqas!
My main point was to avoid people who have an ingoing bias on what to work on simply because of where they sit in the org or what products/channels they own.
Good risk to call out.