How to Business case for SaaS
An example and foray into confidence & impact assumptions
How does an actual draft of an estimation look like I do them?
An actual example of a big opportunity from a semi-invented B2B SaaS platform with a product-led growth distribution model:
Aside from the numbers which I changed, this is exactly what it looks like in the real world.
It serves as a hypothesis that is quite visual and is a great discussion basis for the team before we dive into validation/invalidation.
Colour coding helps to see which assumptions are with high confidence (green, comparable features exist already, real data) or low confidence (red, completely made-up assumptions based on gut feeling)
Recurring themes for use cases or feature additions:
Direct impact on your existing customer base.
Retention -> LTV
Cost of Switching -> LTV
External Top of funnel impact -> How many additional users does this drive per month to the platform (SEO + paid)
Virality impact -> Loop impact
Churn prevention -> How many existing customers wouldn't churn that do churn today if the feature existed
The next step would be to rip the case apart by challenging assumptions and looking for opportunities how to validate with little effort. In general the more to the LEFT / (Start of a process) an assumption is the more likely you want to drive down uncertainty.
The goal to me is usually not accuracy but to spot extremely limited cases by thinking them through.
Understanding Confidence & Impact
2 factors matter when building a business case:
How confident are you of each assumption?
How big is the potential impact of each assumption?
We can infer confidence by comparing it to existing data points - like similar features. If those don't exist we can look at qualitative indicators.
Before those have been connected to quantitative exploration (experiments etc.) they are not in the high confidence bracket.
Another important indicator is *who* you ask. The big difference between customers and users is that we have proven in the first group that they want to pay for our solution.
Users have proven shown interest. Potential users who have never heard about us have no proof attached to them. This affects confidence.
For impact, we can have a simple look at the context of the solution.
Let's assume we want to see the impact of a batch processing feature (processing multiple files at the same time):
If there is a feature that can process 1 document it can also process 10 documents one by one. Or 100. Theoretically at least.
For each group of users the impact is different:
1 Document/day users: No impact. They don't need it
10 Documents /day users: Medium Impact, this makes their life easier. Instead of running the tool 10 times
100 Documents/day users: High Impact, these users can't do it at all.
For confidence, it's worth trying to find simple methods to move up. For impact, it depends on the stage of your company.
Early-stage companies should aim for High Impact solutions. Whereas typically scaling organizations are better at defending their current growth by optimizing.
The next posts will dive into more specifics and when you simply have to stop going into more detail. There is something like spending too much time assuming.
Thanks for reading Leah’s ProducTea Newsletter! Subscribe for free to receive new posts and support my work.
When I saw the first picture, I immediately wondered how you know to what extent you can trust those numbers?
Then I saw the second picture, and it all made sense. I like how you break down to what extent you can trust different the different kind of evaluations. Usually this kind of meta thinking is missing from our business cases.
What many Product Managers do, is they create the same kind of visualization as you did in the first picture (maybe not even as good), and then they don't perform the second step where they critically evaluate the kinds of evidence they are using.
And then they start building and in essence they are purely acting based on a spreadsheet victory containing a lot of noise and little signal.
The key thing is the second step, evaluate your business cases based on past experience and evidence, and do not act purely based on conjecture and speculation.
And when in doubt, you should resist the temptation to build, but gather a bit more evidence before building.