"How to Operate" is one of my favorite lectures
Keith Rabois (@rabois) breaks down OPERATIONAL EXCELLENCE
The principles are taken from the experience helping build Square, LinkedIn and PayPal
They are timeless
Here are 6 mental models from the talk
THREAD...
Keith Rabois (@rabois) breaks down OPERATIONAL EXCELLENCE
The principles are taken from the experience helping build Square, LinkedIn and PayPal
They are timeless
Here are 6 mental models from the talk
THREAD...
RABOIS MODEL 1 - BUILD A HIGH-PERFORMANCE MACHINE
"We used to joke that if Martians took over eBay, it would take 6 months for the world to notice." - Rabois
"Build a company that idiots could run - because eventually, they will." - Buffett
"We used to joke that if Martians took over eBay, it would take 6 months for the world to notice." - Rabois
"Build a company that idiots could run - because eventually, they will." - Buffett
Rabois states that building a high functioning team or company is like building an engine.
It has 3 steps:
Step 1 - Build a beautiful conceptual whiteboard drawing of your engine at peak performance.
👆
This looks like the Ferrari engineering team making a systems diagram
It has 3 steps:
Step 1 - Build a beautiful conceptual whiteboard drawing of your engine at peak performance.
👆
This looks like the Ferrari engineering team making a systems diagram
Step 2 - Take that theory into the real world to build the engine.
👆
This looks like an engineer's worst nightmare.
It takes heroic effort, duck tape and thinking on your feet to keep it together.
"In theory, theory and practice are the same. In practice, they are not."
👆
This looks like an engineer's worst nightmare.
It takes heroic effort, duck tape and thinking on your feet to keep it together.
"In theory, theory and practice are the same. In practice, they are not."
Step 3 - Accelerate your idiot-proof high-performance machine
You have created systems and reduced all friction.
👆
This looks like a Ferrari on the left-hand lane of Germany's Autobahn.
You have no speed limit.
You have created systems and reduced all friction.
👆
This looks like a Ferrari on the left-hand lane of Germany's Autobahn.
You have no speed limit.
RABOIS MODEL 2 - CLARIFY, SIMPLIFY & ELIMINATE
"The more you simplify, the better people will perform." - Rabois
People cannot keep track of a complex set of initiatives.
Rabois says you need to distill it down to 2-3 things that they can repeat to themselves and friends.
"The more you simplify, the better people will perform." - Rabois
People cannot keep track of a complex set of initiatives.
Rabois says you need to distill it down to 2-3 things that they can repeat to themselves and friends.
Refuse to accept complexity.
People will argue that they or their company are different and are more complex than others.
Refuse this.
You can change the world, build the most important companies, and market great products in 30-140 characters.
People will argue that they or their company are different and are more complex than others.
Refuse this.
You can change the world, build the most important companies, and market great products in 30-140 characters.
"Force yourself to simplify every initiative, every product, every marketing - everything you do." - Rabois
He argues that you should see yourself as an EDITOR with a red pen:
1. Cross out everything that is unnecessary
2. Ask for clarification on anything unexplained
He argues that you should see yourself as an EDITOR with a red pen:
1. Cross out everything that is unnecessary
2. Ask for clarification on anything unexplained
RABOIS MODEL 3 - A+ PROBLEMS
"You need to spend a lot of time focusing people" - Rabois
He tells the story of how Peter Thiel "insisted at PayPal that every single person did exactly ONE THING."
Thiel refused to talk to staff about anything other than that ONE THING.
"You need to spend a lot of time focusing people" - Rabois
He tells the story of how Peter Thiel "insisted at PayPal that every single person did exactly ONE THING."
Thiel refused to talk to staff about anything other than that ONE THING.
"The insight behind this is that most people will solve problems they know how to solve." - Rabois
If given the option, people will solve B+ problems rather than A+ problems.
B+ problems are problems that we know how to solve and are in our comfort zone.
If given the option, people will solve B+ problems rather than A+ problems.
B+ problems are problems that we know how to solve and are in our comfort zone.
A+ problems are high impact problems
We don't wake up in the morning with a solution to A+ problems so we procrastinate with B+ problems
If you have an entire company focusing on B+ problems, this cascades and prevents growth
Thiel insisted on focusing on ONE A+ problem each
We don't wake up in the morning with a solution to A+ problems so we procrastinate with B+ problems
If you have an entire company focusing on B+ problems, this cascades and prevents growth
Thiel insisted on focusing on ONE A+ problem each
RABOIS MODEL 4 - METRICS FOR SUCCESS
"You need to simplify the company's success metrics... and create something that is very intuitive for everyone to use" - Rabois
Rabois has 3 core principles for a dashboard:
1. Output > Inputs - Revenue is more important than time spent.
"You need to simplify the company's success metrics... and create something that is very intuitive for everyone to use" - Rabois
Rabois has 3 core principles for a dashboard:
1. Output > Inputs - Revenue is more important than time spent.
2. Meta metric - The metric of success for your dashboard's metrics is what % of people check the dashboard per day.
If the dashboard doesn't get checked by the majority of the company, it's a failure.
If the dashboard doesn't get checked by the majority of the company, it's a failure.
3. Have pairing indicators - Make sure you measure the effect & counter effect.
If you give your risk team the metric of reducing the fraud rate.
Then your risk team starts treating EVERY user like a potential fraud.
You now have the worst customer satisfaction score.
If you give your risk team the metric of reducing the fraud rate.
Then your risk team starts treating EVERY user like a potential fraud.
You now have the worst customer satisfaction score.
You need pairing metrics to prevent this from happening:
- Reduce fraud AND customer satisfaction
OR
- Number of hires AND quality of hires
"You always want to create the opposite (metric)... and the people responsible need to be measured on both." - Rabois
- Reduce fraud AND customer satisfaction
OR
- Number of hires AND quality of hires
"You always want to create the opposite (metric)... and the people responsible need to be measured on both." - Rabois
RABOIS MODEL 5 - LOOK FOR ANOMALIES IN YOUR DATA
"One insight I've had over my career is that you kinda wanna look for the anomalies. You don't want to look for the expected behavior" - Rabois
At LinkedIn, Keith was stunned by a stat that made absolutely no sense...
"One insight I've had over my career is that you kinda wanna look for the anomalies. You don't want to look for the expected behavior" - Rabois
At LinkedIn, Keith was stunned by a stat that made absolutely no sense...
"25-30% of all clicks from the homepage was people going to their own profile"
👆
This made no sense in LinkedIn's UI at the time.
You had to go to an obscure part of your settings to find your own profile.
Keith couldn't wrap his head around what was going on.
👆
This made no sense in LinkedIn's UI at the time.
You had to go to an obscure part of your settings to find your own profile.
Keith couldn't wrap his head around what was going on.
After weeks and a meeting with Max Levchin he finally figured it out:
It was VANITY.
People were looking at themselves in the mirror.
"This really clarified what users of the product really wanted. And we wouldn't have found it without looking for anomalous data" - Rabois
It was VANITY.
People were looking at themselves in the mirror.
"This really clarified what users of the product really wanted. And we wouldn't have found it without looking for anomalous data" - Rabois
RABOIS MODEL 6 - CALENDAR-PRIORITY ALIGNMENT
When Keith meets with a new CEO that is struggling he asks 2 questions:
1. "What are your priorities?" - He gets them to write this down
2. "Can you show me your calendar?" - He contrasts the above priorities with their calendar
When Keith meets with a new CEO that is struggling he asks 2 questions:
1. "What are your priorities?" - He gets them to write this down
2. "Can you show me your calendar?" - He contrasts the above priorities with their calendar
Often CEO's will tell him:
"My biggest priority is recruiting"
But when he looks at their calendar - the largest chunk of time is going to something that isn't recruiting
👆 This is falsification at it's finest
"My biggest priority is recruiting"
But when he looks at their calendar - the largest chunk of time is going to something that isn't recruiting
👆 This is falsification at it's finest
Reminds me of Munger on Mozart being miserable his whole life - because he overspent his income
"If Mozart can't get by with this kind of asinine conduct, I don't think you should try."
If elite CEO's fail the calendar-priority alignment test, I'm 99% sure we are (I am)
"If Mozart can't get by with this kind of asinine conduct, I don't think you should try."
If elite CEO's fail the calendar-priority alignment test, I'm 99% sure we are (I am)
The talk is about building billion-dollar startups... but some of the principles are universal:
1. Make systems idiot-proof
2. Clarify, simplify and eliminate
3. Focus on A+ problems
4. Clear dashboard for KPI's
5. Look for anomalies
6. Calendar-priority audit
1. Make systems idiot-proof
2. Clarify, simplify and eliminate
3. Focus on A+ problems
4. Clear dashboard for KPI's
5. Look for anomalies
6. Calendar-priority audit
I highly recommend checking out the full talk by Keith Rabois
It has less than 130,000 views - it is criminally underrated
It has less than 130,000 views - it is criminally underrated
I sent this thread to everyone in my newsletter - Mental Models Breakdown (Return Of The Mack)
I compress the big ideas + mental models from different disciplines into consumable chunks.
100% high signal. 0% spam.
You can sign up here 👉 eepurl.com/g37gVL
I compress the big ideas + mental models from different disciplines into consumable chunks.
100% high signal. 0% spam.
You can sign up here 👉 eepurl.com/g37gVL
• • •
Missing some Tweet in this thread? You can try to
force a refresh
