Carl and I had a chance to sit in on an hour of Q and A with DJ Patil at Greylock recently, alongside the Princeton TigerTrek. Under the mantle of Data Science fame, I was surprised and pleased to find a very authentic person – a huge practioner of Carol Dweck‘s growth mindset, who is open about his own struggles, and despite all of his successes, remains humble. At the same time, there is also a fierce streak of determination – and the belief for change that accompanies all great entrepreneurs. A lot of what he said resonated with me, and I hope, with you:
Start with needs that you identify at a gut level. Use data to refine those feelings into concrete product decisions.
Even the term “Data Scientist” itself was born out of a need and decided by data. DJ and Jeff Hammerbacher shared a mutual problem. HR was asking for a common term to describe the people that they wanted to hire, as they respectively built out Linkedin and Facebook’s Data Science teams. After some thinking and clever A/B testing, they found that the term ‘data scientist’ attracted the right profile of people they wanted – strong analytical skills.
Ask for Help.
This is something that I struggle with. Often I want to hide away and study it on my own, but I’ve come to appreciate that often the best way to learn is from other people. DJ talks about not being afraid of asking for help – “the worst thing that someone can say is: ‘You should know this by now?’ – to which one can just respond: ‘That’s why I’m asking now!’” It takes courage and humility to say: “I’m struggling with X.”
Reality is always malleable, but some parts are more malleable then others. Place yourself in areas of ambiguity and chaos.
That is where you can have the most freeway in defining reality. Like the title of Amy Jo Martin’s book: “Renegades Write the Rules.” Axioms:
There are no such things as rules, just guidelines.
There are no such things as no, just yes’s that have not been said.
////////////////////// Some snippets of thought:
On Finding Direction:
When starting a company – ask the smartest 100 people around you – “what do you struggle with?”
Establish a goal. Pursue the goal. Be relentless in the pursuit of the goal.
Use the inverse function: decide where you want to be in 20 years, and work backwards from there. Ex. If you want to be CEO of a Fortune 10 company, you want to be X in 15 years. … This allows you to look for unconventional roles that might get you closer to your goal, instead of directionless prestige projection-pursuiting.
If you embark on a doctorate program, you must finish. Otherwise, you will leave irreversible psychological scars.
On civil service:
“I was raised on government funds, and feel a responsibility to give back to my community.” DJ has a passion for education, and in the long term, wants to go back and reform government.”
On Change in Companies:
Surprisingly, DJ says that it was an incredible struggle to get LinkedIn and Fb to be data-driven. Why as the adoption so hard? Inertia and people’s own agendas. Fortunately, DJ had spent time previously working in academic and government agencies, so he was ready to deal with bureaucracy. Where other entrepreneurial types would have thrown their hands up and given up in the face of the red-tape and patience required to get buy-in, DJ hunkered down and worked to convince stake-holders. One of his favorite statements when met with rejection was to say: “Okay, I hear that you are telling me ‘No’. Can I ask you, as a personal favor, if we could just set that aside for 5 minutes, and have a conversation about ‘How’. I am not asking you to commit to anything, just humor me.”
The ability to deal with and resolve conflict is also augmented by a novel definition of the word itself. As DJ says, “Conflict is when someone is in the way of your objective. Most people have the same objective, but tension arises when people are out of alignment.”
On Dealing with Others:
In terms of difficult moments, when someone has wronged you – the best way to respond is to let the other party know what the impact of the negative thing is for you, and then embrace the silence.
On where to work:
Ask yourself: Is this going to be a place where you are going learn an amazing amazing amount? Who is your boss, and will they help you learn?
On the best education for a data scientist:
Mathematics (Probability, not necessarily statistics), Computer Science and something in the humanities, like literature or theatre. In order to understand product, you have to have a deep grounding in something human.
the idea is that you are constantly growing the organization so that you can fire yourself.
You are only as good as the people you can hire.
How to attract the best people? Passion. Authentic passion in what you are doing.
When DJ hires a new person, he sits down to make a list of expectations with them. On the top of the list is the visibility of their career goals. The best way to keep talent is not to make it hard for them to move, but to show them that you are sincerely interested in developing themselves.
Perhaps we are looking at talent wrong – instead of looking at talent as an innate thing, you can grow talent by growing yourself, and those around you.