Wednesday, April 28, 2010

What should a box say?

Moosejaw likes stickers.

We're going to reprint our exterior packing box this year. There have been a lot of ideas floated on what we should do differently than the basics we have now — logo, telephone number, URL. One problem we're trying to solve is when recipients don't open the box soon enough and perishable items, well, they perish.

Digging around I came across these folks who bought from 137 different mail order companies and took a picture of every package that arrived. You can download the full list; most are frightfully boring.

I'm not surprised to see that they praised our friends at the outdoor apparel retailer Moosejaw (they're not photographed in the report since they're not a UK company — they're from Michigan). They have loads of fun. 

Any ideas? Add a comment below.

Monday, April 26, 2010

Traffic Lessons: How Slower Can be Faster


In his book Traffic, Tom Vanderbilt tells a story that I think helps explain how doing things slowly and metronomically by takt — the measured rate of adding work to a queue — can actually make a process faster.

Imagine you are going to pour a bunch of rice through a funnel into an empty jar. You can pour it in all at once. Or you can pour it slowly, in a measured way. Which method gets the rice through faster?

The answer is the second one — because the rice doesn't get jammed. 

The analogy doesn't exactly translate to orders on a conveyor line. Still, it's an interesting visual way to think about how doing something slowly can be faster overall.

Bonus! From the for-what-it's-worth-department, here are two other things I learned from his book, answers to timeless questions we've all pondered and argued about. Or, at least, I have.

Is it faster for traffic overall if, when faced with a reduction in lanes, drivers merge early or at the last minute?

Answer: last minute.

Why? Because both lanes are used to their full capacity up to the last minute.

Why is talking on a cell phone — even if it's hands-free — more dangerous than talking to someone who is sitting in the same car?

Answer: primarily because the person in the car can see the traffic that you see. They change what they're saying to you -- how fast they talk, when they pause, whether they scream -- based on what they see.

Sunday, April 25, 2010

Some Pricing Confuses Me



I don't know about you, but sometimes the way businesses do things completely baffles me.  Take these pricing games. They probably make sense in some way I don't understand, but I am left confused so excuse me while I rant.

Why is almost every song on iTunes the same price? Also, why is every new CD at a music shop essentially the same price? It doesn't make sense to me. As a comparison, take another product like clothes,  Do we expect every pair of jeans — regardless of who made them or how good they are — to have the same price? Why would we expect the same of music?

Why do book stores put hardcovers on sale the day they are released? After all, they're new. Customers are excited. They want to buy them. They'll buy at full price. Why give them a discount? Isn't this like a movie theater selling a movie's opening weekend tickets at half price? Or a clothes store selling a newly released spring fashion line at a discount while last fall's collection sells for full price?

Why do publishers print the price on a book? Why not let book stores sell it for whatever price they want? After all, if a book store has better service, better trained employees, a nicer shop, shouldn't they be able to charge more? Isn't that how it works with virtually every other product we buy?

Why do airlines charge to check luggage? Isn't it in their interests to have passengers board the plane faster? Carry-ons slow boarding. Now every passenger takes more carry-ons because checking luggage costs dollars. Also, passengers now game the system: they bring a carry-on and when it can't fit it in the overhead bin it's checked anyway — free!

Friday, April 16, 2010

Thursday, April 15, 2010

Checking Your Change is Building in Quality


While I was at the bodega today, buying Bubble Yum and popcorn, I counted my $1.34 in change and thought about order checking and error proofing. Wow, do I sound boring. But listen.

Counting your change is essentially checking a process. It's the same as a check process for order taking, a check process for picking, and so on. It's checking the change-making process. The clerk tells you the price. You hand them the dough. The clerk tells you how much the change is. They put it in your hand. You count it. Is the amount correct? If it's correct you say Thanks and leave. If it is incorrect you say Whoa immediately and get it fixed on the spot. They key thing here is that the check is immediate.

Imagine another scenario. Instead of immediately checking your change you saved it to the end of the day. You did that with every transaction all day long. At the end of the day you reviewed all the transactions to check if they were correct. How much extra work would it take? How much paperwork and extra record keeping would you have to create? How much extra work would it take to fix the correction, to go back to the clerk and convince them that there was an error? 

Now imagine you saved it till the end of the week. Or the end of the month.

This is essentially the concept behind building quality into a process versus confirming quality at the end. When you build quality in, it's not that you don't check. It's that you check immediately. Reducing the time gap between the action and the check has the largest effect on quality. It also has a tremendous effect on cost since it usually requires a lot less paperwork and bureaucracy. 

Friday, April 9, 2010

New York City Butchers: Turning Operations into Marketing

Today, an article from New York Magazine on the city's prince of hamburger, Pat LaFrieda.  Pat's third generation family business wholesales meat to restaurants in New York. He's the man behind the meat for many famous burgers in town, from Danny Meyer's Shake Shack burger ($7.25) to Minetta Tavern's Black Label burger ($26).

It's a really interesting article that you can read from many perspectives. 

From one angle it gives insight into the service levels, fulfillment obligations, and straight up wack madness it takes to run a wholesale operation for Manhattan restaurants.

From another, it makes most restaurateurs look like lemmings, which might not be far from the truth.

Third, it shows you how totally apeshit NYC is for hamburgers right now.

Fourth, and the reason I'm sharing it, is the business side. This article has a  Harvard Business case buried inside, deep below the intent of the story. There is a butcher who sells hamburger. What was once a completely undifferentiated product, a commodity, he has successfully differentiated. He's no longer tied to the standard market price for hamburger that every other butcher sells for. If you buy the prices in the story — I don't, but I bet they're not that far off— he's getting six times the price for his hamburger than normal. That's a tale in itself. 

But I kept waiting for the story to tell me why the price was higher in the terms I'm used to, the terms of provenance. Who was the rancher? What did the cattle eat? How long did they live? How were they slaughtered? Surely that must be the reason prices are higher? I got few answers there. It's not because they have secrets. It's because the answers are not special. Their source for meat sells corn-fed cows just like everyone else. Pat doesn't sell get his high prices because the cows are special. 

So how does he get to sell for such high prices?

The answer is that he creates custom blends for each customer. You want your restaurant's own blend of chuck, sirloin, cheek, jowl, dry cured, wet cured —whatever — for your hamburgers? No problem. LaFrieda meats will do it for you every day. Order tonight, you get it tomorrow. As Pat says, "“What we do, basically, is become a different supplier for every restaurant.” He even does it if you're a little guy who doesn't order much. In fact, he seems to like little guys. In the article two big guys call him — the legendary steakhouses The Homestead and Peter Luger — and he essentially tells them "Nope."

In other words, it's LaFrieda's operations that get him a higher price. It's his operations that perform his marketing. Operations drive the sales. How does he get to sell for such high prices, how does he get in all the best restaurants, how does he get this magazine article written about him? Basically, it is because he works in small batches with short lead times and customizes selection for all customers. Sound familiar?

Tuesday, April 6, 2010

Two Kinds of Checklists

Tabbed pilot checklist book for Cessna.

In The Checklist Manifesto, Atul Gawande discovers an interesting distinction among checklists. He divides them into two camps, based on studying airplane pilots.

One type of checklist is called Do-confirm. The other is called Read-do.

Do-Confirm checklists are basically what we call SOP's, Standard Operating Procedures. They are shorter than a full set of instructions — it would be hard to train yourself how to do something if you just read them — but they have sentences and detailed notes. They can be used by an auditor to watch someone performing a task. They can be used by an individual who is managing a series of tasks over a long period of time, checking off boxes as they go. The important point is that Do-Confirm checklists do not work well when a task is very time sensitive.

Read-do checklists are a different species altogether. Read-do checklists are the checklists pilots pull out when, say, an engine blows up while in flight. They are precise and only contain the most critical steps. For pilots, most Read-do checklists are less than ten steps long and some of the steps might be a couple words long, like "Fly plane!" They require an expert to use them because the expert must be able to perform the step without hesitation. The expert must also fill in missing steps — read-do checklists are never 100% complete. They are meant to be used in the moment for short tasks. These checklists are tested over and over in flight simulators to make sure they really work. There is an entire organization devoted to creating and re-testing these checklists so pilots won't refuse them for being too cumbersome. In other words if pilots say "The checklist is too long, it's not useful," the checklist maker considers it a fault of the checklist, not the pilot.

To me, this was a small revelation. A pilot will use both kinds of checklists. But they don't use a lot of Do-confirm SOP's while flying. At that point they use the Read-do checklists -- something very brief, something that doesn't distract them from the task of flying. The Do-confirm SOP is too long, too detailed. They use that before or after they are up in the air.

Pilots would say if you're trying to use a detailed SOP to manage a time-sensitive task —like checking an order or packing a box — chances are people won't use it regularly. It's not because it's broke. It's because it's too cumbersome. The pilot would say you are probably using the wrong kind of checklist — switch to something shorter.

Sunday, April 4, 2010

Good Signs: Potholes in NYC


I thought this was a nice use of a sign. It's a big sticker affixed to a pothole on my street in Brooklyn. Well, more like a potcrater. It lets us know several things:
  • Someone who can fix this knows about it 
  • Don't bother them because they already know about it
  • You can rest assured something is happening
  • If you want to let them know about more things like this there is a phone number
That it's attached to the thing it's informing on is key. There's no confusion about what the problem is, even though the sign doesn't say what it is they're "looking into." That allows the sign to be generic. NYC city services doesn't have to make a new sign for each problem. They just drive around with a stack of these stickers in their trunk.