Keith Rabois - How to Operate
I am not sure where I had first discovered the summary to Rabois’ How to Operate talk. But whoever wrote the summary below did a great job, and I am just copy pasting the summary here so that I can share it with people who want to learn about how to be a better operator
- MAKE IT IDIOT PROOF
Build a company that if idiots ran, nobody would notice.
They used to joke that if Martians took over eBay, it would take six months before people would know.
“Invest in businesses any idiot could run because someday one will” - Peter Lynch.
- BE A DOCTOR
The key to successful operations is to identify when an issue is a cold, and when it is something more serious.
A cold = A little annoyance that will fade away.
However, sometimes something that appears like a cold can actually be fatal.
- BE AN EDITOR
COO’s should see the company like an editor sees the first draft of a novel. You should take out the red pen and start striking anything that is getting in the way.
The more you simplify, the better people will perform.
Complexity = Failure in teams.
- CLARIFY
Ask yourself, what are the 1-3 things that matter for this company?
ONLY focus on these things.
With each elimination of what’s not on this list, you can improve performance by 30-50%.
- ENSURE CONSISTENT VOICE
You can pick up one article in the Economist - and it always feels like it has been written by the same person.
You should try to make your company always have the same tone of voice in its output.
- DELEGATE
Remember, you’re an EDITOR. You must delegate most of the work to the WRITERS (i.e. product, marketing, sales etc)
- OWN EVERYTHING
Even though you delegate, you must take what Jocko Willink calls “extreme ownership” over everything that goes on. It’s your fault if it goes wrong.
- TASK RELEVANT MATURITY
How do you know when to delegate and when to micromanage?
Low confidence in your own opinion + Low consequences = Delegate. Let people make mistakes and learn.
High confidence in your own opinion + High consequences = Don’t delegate.
- MAKE SURE YOU NEVER GET THIS RESPONSE
Avoid situations where you state what to do and the employee responds with “Okay, you’re the boss”
You will burn social capital over time by doing this.
(Note - this is so true. The worst manager I have ever had did this)
- AMMUNITION VS BARRELS
Ammunition = Average employees. Barrels = Great employees.
Barrels always find a way to get a task done.
How to spot a barrel? Watch when people start going to an individuals desk who they don’t report to asking for help.
Give barrels equity.
- THIEL’S ONE RESPONSIBILITY
Peter Thiel used to insist at PayPal that every single person was only responsible for ONE thing.
He would not talk to you about anything else.
Everyone rebelled against this, BUT, It made people focus on the A+ problems…
- A+ PROBLEMS VS B+ PROBLEMS
A+ problem = The most difficult problem that will change the company B+ problem = Second and third most important problem
Employees avoid A+ problems, and take the easy route out by focusing on B+.
Thiel would only talk A+ problems.
- BE TRANSPARENT
Have glass walls for meeting rooms. People worry more when meetings go on behind closed doors.
Patrick Collison at Stripe does email transparency. Everyone can read every email sent within the company.
Steve Jobs did transparent compensation at Next.
- DON’T LET OBJECTIVES GO AWRY
If you measure only one thing, the company tends to optimize for that. And often at the expense of something else that’s important.
E.g. Insurance companies that just measure the reduction in fraud rate. Employees start treating each customer as a potential fraud case. You may have reduced fraud, but you created the worst customer service in the world.
Make sure you have PAIRING METRICS that oppose one another.
Measure the fraud rate AND the customer service.
- PONDER ANOMALOUS DATA
At LinkedIn, 30% of new users clicked on their own profile.
This baffled the team.
Upon further digging, they realized this was people checking out their own profiles. Vanity clicks!
Anamolous data can give you key insights about your customer.
- HOW YOU DO ANYTHING IS HOW YOU DO EVERYTHING
Bill Walsh took the 49ers from the worst team in the league to winning super bowls.
What was the first thing he did when he took charge?
He wrote a 3 page memo on how the secretary should answer the phone.. His point was that if the organization does everything exactly the right way, then the receivers will start running their routes at 7 yards - not 8/9 yards. Exactly 7.
Steve Jobs insisted on immaculate circuit design despite the fact nobody outside of Apple would ever see it.
- SERVE GOOD FOOD
Serve your staff good food at lunchtime.
What do they do when they don’t like the food? Complain about the shit food rather than brainstorm ideas.
- CREATE A CULT
“Every good startup is a cult” - Peter Thiel
This is why you should never have shared office space.
It’s hard to start a cult when you are sharing your office with other people.
- BEST MANAGERS = BEST PERFORMERS
At PayPal, they only promoted people who were kicking ass at their discipline. They didn’t believe in general mangers.
The VP of engineering was the best engineer. The VP of design was the best designer. People can always learn to manage later.
When you don’t promote the best people, you demotivate other members of staff.
There’s nothing more demotivating than knowing you are better at your job than your boss is.
- AUDIT YOUR CALENDER (Best till last!)
Whenever Keith is working with a CEO who isn’t thriving, he asks them to write their priorities down on a piece of paper.
And then pulls up their calenders.
50% of CEO’s will say that recruiting is their number one priority, however, it doesn’t match in their calendar.
You can solve most problems by matching calendar with priorities.
If something is your number one priority, than 25% of your time minimum should be allocated.