The Connection Newsletter 18 - Quit

The Connection Newsletter 18 - How to quit like a boss

July 30, 2018

Hello!

This is edition #18 of

The Connection

, the weekly email I send family, friends, and future friends (hi!) Glad you're here.

(If you don't want these newsletters, click the unsubscribe link at the bottom. No worries, we'll still be friends.)

Make sure to hit "Display Images" above to see puppy pics. 

. I’ve worked a lot of jobs, about

. Which means I’ve quit a lot of jobs, too. I’ve learned there’s a wrong way and a right way to quit. The wrong way burns bridges while the right way allows you to keep your relationships.

.

Now, onto the reads this week:...

. In an article about

, I mentioned that Michelle Williams should certainly fire her agent and agency (WME) for the compensation fiasco revolving around ALL THE MONEY IN THE WORLD. Ms. Williams never did, and in this Vanity Fair article, she talks about why. 

Favorite quote: "

I believe in second chances.

" 

...

? AT&T acquired Time Warner and therefore, HBO, and they want to squeeze the golden goose (or dragon, as it were) a bit harder.

Executives are looking to Netflix’s model for inspiration and believe by pumping out more content they’ll get more data, which to in turn helps them create better content.

But HBO’s bread and butter lies in their extensive development process and curation of quality content, not churning out 2-3 lackluster Adam Sandler movies a year. Either they make the product fit the channel, or it’ll be like a square peg in a round hole. Related:

.

...

. With a new CEO in place, Chipotle is working overtime to master the digital world. That means a back of house “second line” (another crew and burrito prep area (guacamole’s still extra) that focuses on mobile pick-up orders). Tickets get printed directly to the second line and bypass the front of house staff. Plus, they’re working on a digital drive-through: you order online, drive by, and pick-up your food -- without having to talk to any pesky humans.  

...

. Stitch Fix has gotten a lot of press thanks to its status as a B2C company whose IPO didn’t fall flat on its face (looking at you, Snap and Blue Apron). But how does it work? Is it as simple as filling out the preferences form, and a human stylist choosing outfits for you?

In a word: no

Data science is used at every step of the way. A few aspects of how data science helps after a client fills out their extensive survey:

1.

“Where does the clothes come from?”

There are many warehouses that package the clothes, and not all warehouses carry all the same items. How do they get the right clothes to the client at the lowest cost, and on time?

2.

“How do customers get clothes they’ll like?”

Based on the survey, what outfits will they like? They solve this with a combination of predictive modeling (“you liked this shirt so you’ll probably like this pair of pants”) and collaborative filtering (“those who liked what you have liked have also liked”) AND a human stylist

3.

“Based on what they said, what outfits do they need?”

For example, the customer might say they need something for a wedding. By analyzing the text with natural language processing (NLP) they can score items based on the client's request note AND textual feedback from other clients about the same item.

This only scratches the surface. Highly recommend this beautiful article.

...

Finished book notes from

by Tim Wu. Some incredibly interesting, rich stories from the wars of the information age, when the battles were waged over radio, television, and film.

...

This week’s inspiration: this single tracking shot in

music video (0:15s -> 1:15).

Thanks for reading!

Any questions, comments, or recommendations, feel free to reply directly to this email.

Reply

or to participate.