In itnig’s Podcast #31 Jordi Romero, CEO at Factorial speaks with Antonio García Martínez about his past experiences designing ad-targeting products at Facebook and his perspective on the current Cambridge Analytica controversy. Listen to our podcast on Youtube, iTunes or iVoox.
After a doctorate in Physics, I started working at Goldman Sachs as Quant Analyst in 2005. It was the time of the financial crisis in the US, an economic apocalypse. I thought the only thing that might survive after this crisis is the technological sector.
So I became a Research Scientist in an online advertisement company, a niche in the tech sector. There I got to know my co-founders, we applied to YCombinator and went through their Bootcamp as a startup. During this time, all problems that can occur happened to us. After 10 months Twitter bought us in a so called Acqui-Hire, when they buy a company but what they are really after is the talent and to get the founders and employees on board. After a bit of drama I went to Facebook.
At Facebook, my role was that of a Product Manager of ad-targeting products. At this time, in 2011, the whole team at Facebook were 20–25 engineers and 5–6 Product Designers. We all fit in a meeting room. However, Facebook had about 1 billion users and even with almost non existing business or monetization model the revenue was high. It was the time Facebook would transition from a startup with a crude business model to what it is today.
I was involved in product development for things such as the Custom audiences. With the IPO Facebook went public and started to turn its focus towards monetization models. These two years of developing products are what makes Facebook money now.
What is going on with Facebook now in terms of privacy?
The Cambridge Analytica crisis comes from the Facebook platform. You probably all remember the time when you could login on Spotify through your Facebook account and would receive some kind of spam to your Facebook profile. Facebook decided to take this step of creating a platform to make an integral product that could span and connect different sites. However, this was not a good product, a fail, as in the long run nobody was using the Facebook platform. As a user, you would use Facebook to log in, some of your personal data are shared with the product you are logging in to and this data might be used.
There really is not much Facebook can do to regulate the data flow and what happens with this information once it leaves the Walled Garden of Facebook. And that’s basically what happened as researchers at Cambridge University created a psychographic model, an app which by asking questions tries to make a psychological assessment of you as a person. They were creating a five dimensional graph based on the big 5 personality traits (Ocean — openness, conscientiousness, extraversion, agreeableness and neuroticism) and projecting the user’s personality in five dimensions.
Through a model they then correlate this with your political views, eg. pro Trump, pro immigration.. and combine your psychological profile and political stand with Facebook’s ad platform to be able to find a person on Facebook and specifically target advertisement to the believed preferences.
Obviously the story has some James Bond badness, researcher, an almost hidden secretive company, financing Breitbart, Bannon’s involvement as editor — It’s a compilation of different elements that makes this story.
If the problem was that Cambridge Analytica breached the terms of service of Facebook, why did Mark Zuckerberg hide?
This is a bit curious. I think it’s simply because Mark Zuckerberg is not the most social person, he’ll do a Q&A internally and answer all kind of questions but externally he seldomly shows his face.
The problem is the perception, not so much the actual impact it had in the elections but the perception of it.
What about Fake news?
I think fake news is a real problem but it’s hard because there is no obvious solution. Compared to ad content, organic content is much harder to control. People are used to having a Feed of content optimized to their likes by default. In the US a lot of media consumption happens through Social Media, that’s hard to change. Two months ago, Facebook made a change to the Feed, giving journalistic content less importance and in a way bringing the old Facebook- between friends- back again.
This reminds me of the last podcast in which we spoke about cultural fit. You once told me about the employee handbook, the little red book that you received at facebook. Can you tell us a bit about it?
In my book, Chaos Monkey I talk about it a bit. During the interviewing cycles, somebody will always ‘there is no cultural fit’ and this can be some hidden part of racism, sexism…it might mean ‘the candidate is not like me’. T
In regards to the Little Red Book, it was born in 2012 out of Facebook’s fear of converting into an old, structured company. Worried that they would not be able to keep the agility and aggressiveness of a startup, the Little Red Culture Book was one way of fighting against the corporate ageing that might creep upon us as an organization.
On the last page for example it reads:
If we don’t create the thing that replaces Facebook, someone else will.
That’s the tone of the Little Red Book, I still have one copy actually.
If you want to find out more about it, we recommend to read Antonio’s book with many more anecdotes from his time as Product Manager at Facebook: Get the book here and listen to the whole podcast on Youtube.
Listen to our podcast to learn more about Antonio García Martínez takes on Silicon Valley. Learn more in this Podcast in Spanish on our Youtube channel, listen to it on iTunes or enjoy it through iVoox and subscribe to our newsletter to stay always up to date.