🎯 006 Resource Allocation

How do you cast a movie?

🎬 Imagine this. You work in film. You made a movie. You were nominated for an Academy Award last year. You earned a great salary.

You want to make another movie. It’s an idea you’ve had for a while. You watch a lot of scenes from your favorite movies. It’s gonna be like this. You start to edit your favorite film footage into montages of what’s in your head. Yeah, like this.

You need actors.

Actors come. Actors read. Actors leave, rather, they are dismissed. You’re demanding.

You have a list. Call these people back you say to your casting director Fred, I need to see them on film and I need to see them together. You three people.

Sure, Fred thinks. It’s a good list. He likes it. But not that much. Fred adds some names. There’s one guy Fred really wants. You already worked with him. He was good. But you don’t want all your movies to be with him. Fred wants him though. Fred hires him to be around and suggests that since the guy is here he can read with other people.

Sure you say.

The guy is fine. No, he’s good. The more he reads the more you like him (thanks Fred!). A woman that Fred added back is really good too. Really really good with this guy. He’s smooth. She doesn’t buy it. This is a good combo. Another guy that you dismissed and Fred added back is good too. This is a good trio. Is this the trio?

---

🧐 Your problem is an allocation problem. How do you allocate scarce resources (roles in your movie)?

Usually, we allocate by proxy. We do one thing that (might) map to the job we want. Interviews are like this.

“We know from years of data that job interviews don’t predict a whole lot, the main thing they predict is if I like you and that’s not necessarily job relevant information.” Cade Massey

Other times we use proxies that are easy to quantify. Any industry where analytics has risen in prominence over the last two decades is like this: investing, sports, etc. The easy measures (hits) have given way to more predictive measures (on-base-percentage).

A third way to allocate is under the no-one-gets-fired-for-buying-IBM approach. It’s where local incentives triumph over larger incentives.

A wildcard allocation idea from Rory Sutherland is to allocate in groups. Hire in groups, Sutherland says, and you’re bound to get diversity. It’s like shopping, if you’re buying two tops do you get identical items or do you change the pattern, colors, cut, etc?

Another allocation idea is to include someone like Fred.

---

🎦 This is a good trio. There’s another trio too though. The other trio is serious. This trio is goofier. The other trio is your choice. This trio is Fred’s choice.

We’ll go with this trio.

You were George Lucas.

One thing that’s great about Lucas’s experience is that he didn’t have to answer to anyone, and avoided the buying-IBM angle.

About that other trio: Han Solo would have been played by Christopher Walken.

---

🔠 Resource allocation comes up in a lot of places. If time is included it comes up everywhere. What’s the best person to hire for this? What’s the best project to add money to? What’s the best way to spend my time?

The lesson from Star Wars this week - a belated May the fourth be with you - is to have a Fred that sprinkles in secondary choices for a second look.