How To Change The World? Get The Small Things Right – Dalton Caldwell and Michael Seibel

Y Combinator

Tree Uprooting Analogy

Changing world = uprooting big old tall tree
Two founders: one knows trees have roots (advantage), other has no idea
Founder without knowledge experiences horrible on-the-job learning
"We're pushing it and it won't fall over" - math says it should work if tree ends at ground level
Point: much exists under surface that must be understood

YC founder comment: "Built product because we thought world should be certain way, not because we had customer who actually wanted to buy the thing"
Common hero worship: Elon Musk, Steve Jobs - forcing universe to their will
Elon example: electric cars of future, making it happen
Reality: hard for "mere mortals" to force people to want what you're selling
Difficult to debate/convince someone they want something they don't actually want
Most people lack organic sales experience

Vision: "Walked onto used car lot, genius salesperson convinced me to buy car"
Reality: doesn't work that way - nobody randomly walks on lot if they don't want car
Classic ideas: "buying car sucks, dealing with car salesman sucks, buying house sucks - let's build website to eliminate dealers"
Problem: haven't done research on why these systems exist
Can't just "wave magic wand" or "sprinkle software" to make intermediaries disappear
Realtors example: "why is it so hard to buy house?"

Simplistic engineering efficiency arguments don't always work when applied to real world
Hiring startup pattern: "I'm genius engineer being assessed by non-technical recruiter"
Thesis: "crime against humanity, dehumanizing experience"
Goal: empower hiring manager with software, eliminate all recruiters
Reality check: if massive recruiter teams exist at companies, must be some efficiency

Core learnings about hiring:
- Engineers better at evaluating other engineers
- But: don't tend to enjoy doing it
- Not greatest way to extract value from great engineer
- Not best use of their time
Hiring companies run against reality: "you aren't going to defeat HR"
Recruiters have more power than founders think
Hiring manager = "talk friend but not action friend" - complains but never uses your solution

Can make huge changes in human behavior if you understand subtle pressure points
Must understand motivations of ALL key parties and cater to those motivations
People who really change world got lot of small things right in earliest stages

Initially bizarre concept: getting in stranger's car to drive to airport
"Here's my idea: stranger comes to your house, you get in back seat of Toyota"
Was insane at the time, now normal
Started with hired cars/limo cars (which was already normal)
Many little things had to happen to normalize ride sharing
Required understanding all core constituents
YC funded ride sharing ideas before Uber that didn't work - was common bad startup idea
Key: understand small motivations of all people involved in ecosystem

YC founder comment: "Did not do enough research before deciding what to work on"
"More research you do in your space before you write line of code, the better"
Founders go through effort of starting company, raising money, applying to YC, making interviews
But haven't googled their idea or read articles about previous attempts
"Self-own" - free to research, doesn't take much time
Spending years/months on idea without basic cursory research is "really dumb"

Commonly young founder issue
Internet knowledge doesn't go way back - if haven't heard of company, assume never tried
Ask friends instead of googling (friends have similar limited history)
Psychological: don't want to learn something that would defeat idea before starting
Want to "live in illusion/dream for little longer"

Dot-com bubble 90s: people tried pretty much every idea you can imagine
Every idea today has analog from 90s
Same in 2000s, 2010s - patterns repeat constantly
Instacart example: people tried before, didn't work, but Instacart founders knew that and intentionally designed to be better
"Stood on shoulders of those that came before them"
Webvan existed - not knowing about it doesn't help

Platforms change faster than human problems
Lived through: Web 1.0, Web 2.0, mobile, current era
Human problems remain constant:
- Want grocery shopping without going to store
- Want good food delivered
- Want to get job
Almost guaranteed you're not solving new problem, just using new technique/technology
Should study how everyone else tried to attack old problem

Question: isn't expert advisor/angel/seed/Series A investor validation enough?
Answer: best to do your own work
Ultimately responsible for your company
"You have to live your life - if startup doesn't work, it's not really my problem"
Experts can be too biased positively/negatively based on small data set
Get wide variety of opinions, synthesize like listening to different music

Spoke to every person who tried challenger bank before
Got two things: useful information they wrote down, and "don't do this, are you crazy"
Ignored discouragement but used useful information
Need to know what bits to draw from expert advice vs blindly accepting as facts

Counterintuitive: investors who know slightly less about space can be more optimistic
Those with more expertise may be less optimistic due to past hits/failures
People assume expertise equals optimism - not always true
Example: asking DoorDash founders about food delivery vs asking someone else
Sometimes nice about not knowing much - easier to see as simple problem

Try for free