Most modern retail outlets (walmart, sams, costco, kroger, etc.) all use "Just In Time" (JIT) shipping and delivery. This means supplies arrive just before they're needed for the shelves. This keeps things fresh, minimizes the amount of "back-room" space needed to store things not on shelves, and generally helps to keep costs down across the board (they almost never have excess inventory they have to get rid of, for example). Remember in the old days if something wasn't on the shelf they could generally find it "in the back" ? Ever notice how that's never the case anymore? "Hey, if it's not on the shelf, we don't have it!" This system really does save money, and in turn, keeps the price of goods a lot cheaper.
Walmart has this one down to a science... Their system is REALLY fascinating (without a doubt, the most sophisticated supply chain management system ever built). They do a couple of interesting things...
First, every purchase you make is tracked (not necessarily back to YOU).. they want to know what you buy with what. What do they do with this data? Well, there's a 20% chance that someone who buys a Hersheys candy bar will also buy Graham Crackers and Marshmallows -- so it makes sense to put some of each next to the Hersheys bars, or put some Hersheys bars next to the graham crackers (or both)... If you walk around a walmart store, you'll find they have this down to a total science -- you'll see all these random items being placed next to each other (many seemingly have nothing to do with the other one -- but their computer tells them to do it)... This is no parlor trick. The system has a MAJOR impact on walmart revenue -- ever notice how you can walk into walmart intending to spend $10 and walk out having spent $40?... Yeah.. they're pretty good!
The other thing they can do with this system is connect it to predictive modeling systems... such as weather. People tend to buy more Popsicles on a hot day -- so guess what walmart does? 2 days before a heat-wave, their computer system AUTOMATICALLY ships extra Popsicles to stores in areas that will experience the higher temperatures. No people are even involved. The guys in the warehouse get a "packing slip" (generated completely by the computer).. so they load up the truck, and it shows up at the walmart store where it's unloaded. I've heard jokes from guys who've unloaded the trucks that they can predict all sorts of funny things based on what supplies the computer decided to put on their truck.
Think about what this means in terms of "Just in time" arrivals... In the summer, people drink more lemonade. During storms, people buy more batteries, water, and flashlights. On extra hot days, they buy more ice cream, on extra cold days, they buy hot chocolate, On Saturdays, people tend to buy more milk and bread. All of these events can be predicted with a great deal of accuracy now, thus the stores always have "just enough" inventory, despite seemingly strange fluctuations in demand. They get down to very granular levels with their predictions (these are very simple examples).
Also, this is a REALLY slick system for hurricanes.... A few days before a major storm hits an area, truck-loads of diapers, batteries, toilet-paper, flashlights, bottled water, etc. will arrive at stores.... Again, the computer does all of this (no human EVER needs to be involved).
Some parts of the country have different purchasing habits -- the walmart system is smart enough to "learn" this.. (Example: In Utah, milk and eggs are a popular saturday-night staple -- the walmarts in Utah are ready, while they may not care in other parts of the country -- this is completely automatic).
I also know that fuel companies (like Flying-J) use a similar system for predicting where fuel demands will be, and thus automatically deciding how much gasoline should go to their different service stations across the country. When big storms are predicted, trucks are usually arriving with extra fuel before the storm is ever a problem.
While this is really an amazing technological feat, it does expose a certain amount of vulnerability in our current distribution system. If everything is Just In Time, what happens if a major event occurs that could not be predicted? (Example: an earthquake). Also consider, if purchasing habbits in a given area suddenly change, it may leave shelves a little empty until the system can re-learn what's going on.
Food for thought.