Case Studies from The Lean Startup

Do not always assume that if metrics are not good, solution is adding or fixing features
Learned to use Cohort analysis to be able to connect changes made to the effects on metrics
When they realized that their metrics were not going to improve enough they pivoted and started a new baseline*
Even though # of users and paying users were at records (vanity metrics)
Engine of growth metrics were not being met
Diminishing returns would eventually peter out
Do you see how they could have record numbers of new users and paying users and still not on a path to success?
- Vanity metrics
- Engine of growth
- Change of markets
Decision to pivot here is very confusing
- Even once you realize the problem
- Not at all clear what to do about it/

Selling shoes online
Hypothesis was that customers would buy shoes from a nice web site (value hypothesis)
Getting info about shoes - models, brands, photos, prices, was a major schelp
Simple experiment yielded great information about customers, value and actual names

300,000 employees, may volunteer 4 hrs / month
Instead of having lots of focus groups to try to figure out what to do
Start small: formulate hypotheses and test them.
So in this scenario, what is the product?
Value hypotheses: Does this product or service provide value to users
Growth hypotheses: Will usage, purchase, deployment, adoption of the product grow?
Concierge MVP could be used to simulate the service and measure referrals (see Zappos)
The essence: experimentation, produce data, and interpret it to see what it means

Don’t skip right to #4!!
When building a very bare-bones product: if customers don’t complain about a missing feature, it may well not be needed!
A sequence of experiments any few of which can branch off and suggest a few more experiments”
In each case you are creating more MVPs each time closer to a real offering. Kind of a branching tree/network

I thought it an odd idea: delivering customized recipes based on specials in the local grocery store.
An extended scenario of an MVP is described. Lessons:

Whats the point of enumerating all the kinds of pivot?
How is Votizen doing now?
What were their original set of four leaps of faith?
- Register -> Activate -> Retain -> Refer
- How would you measure the effectiveness of the MVP and the stages?
- How does “cohort analysis” apply here?
How would this work for a hardware product?
What tuning did they do before they pivoted? What did they try?
Causes acquired Votizen. What do you think Votizen’s biggest asset was?
- Do some classroom research
- Causes acquires Votizen
- Votizen Success
Leap of faith" assumptions are funnel steps
- Register -> Activate -> Retain -> Refer (Where was revenue?
- Register -> Activate -> Retain -> Refer -> Pay (platform for lobbyists etc.)
- Register -> Activate -> Retain -> Refer - Pay (self-service transactional service)