Trends in Retail

Like much in society today, the retail landscape is changing.  And I am not talking about what store chain is up and what chain is down.  Or what trends are newly hot and what trends have gone cold.  I am talking about structural changes in the industry as a whole.

And this is not the only upheaval in retailing I have observed in my lifetime. There have been plenty.  But, as is my wont, I want to begin at the beginning, or, at least close enough to the beginning so I won't miss anything of significance.  So let's start a thousand years ago.

At that time, most people were tied to the land and saw little commerce of any kind, let alone retailing.  The exception was China, where things were far advanced over pretty much the entirety of the rest of the world.  I am going to ignore China.  The rest of the world eventually caught up to China so it will get covered indirectly.

Transportation at the time for most people consisted of foot power.  They were farmers who got from place to place by walking.  So typically there was a market or village (often the same thing) within ten to twenty miles of the farm, and at that time most people lived on a farm.  There was no money to speak of so what little commerce there was took place using barter.  And barter is very inefficient.  But at the time it was the only option available.

So people tried to accumulate a small surplus of something, anything, that others might need, so they had something they could trade for what they couldn't make or grow themselves.  So besides the market, the village hosted a small number of artisans, blacksmiths, carpenters, and the like, who provided what people couldn't make themselves.  The villages were small and served a small population so any one village hosted a very few artisans.  There wasn't enough local economic activity to host more.

But mostly people made for themselves.  So, they made clothing.  But they had almost no technology to help them out.  So they wove and they knitted.  And they used the skins of animals.  For weaving and knitting you need thread or yarn.  Both are difficult to make without equipment and the process is labor intensive.  So making the thread or yarn was where most of the effort went.  A good carpenter with good tools could make a spinning wheel.  That could turn fibers and hair into thread or yarn.  And a skilled carpenter could make a simple hand loom for use in weaving cloth.

But the these simple looms were best employed making rectangular pieces of fabric.  So a sarape was a good design.  You wove two rectangular pieces of cloth.  Then you could sew one edge of each together leaving a gap in the middle for the head.  A cloak, that article much beloved by purveyors of romantic fiction and epic fantasy, can also be made with a rectangular piece of fabric to which a couple of ties are attached.  Add ties to a rectangular piece of cloth and you have a wrap-around skirt.

But neither a sarape nor a cloak is actually very good at keeping you warm.  (The wrap-around skirt is a definite improvement over both.)  But complicated construction was nearly impossible and that limited peasants to simple garments, none of which worked all that well.  The button was invented late, not because buttons are hard to make, but because button holes are hard to make.  And the zipper dates back only to about 1900 and Velcro to about 1960.  Only rich people could afford elaborate clothes of complicated construction.

And that was generally true of almost the entire population.  If they couldn't make it for themselves, perhaps aided by a few simple tools that had to be well enough made to last decades, they had to find a way to do without.  All this made the exchange of goods very difficult so few goods were exchanged.

Things slowly got better.  Money, necessary to replace barter with more complex forms of exchange, was slow to arrive in quantities sufficient to make a difference.  And more complex tools also became available only slowly.  Steel was a luxury.  So, for a long time, if it couldn't be made from wrought iron, it couldn't be made.  This meant, for instance, that plows were very crappy for a very long time.  A wrought Iron plow wears out quickly.

But the situation slowly improved.  And specialists like harness makers started to arrive on the scene.  It was difficult but possible for a small farmer to make his own harnesses.  But a specialist could turn out a much superior product much more quickly.  But that specialist needed skill and tools. Cobbling, making shoes and boots, is similar.  Again, it is possible for a farmer to make sandals or moccasins but a specialist can turn out a much superior product much more quickly.

So, as money became more and more available and other improvements were made, the typical town or village slowly acquired more and more types of specialists.  And these specialists made the farmers more productive.  And that meant a village could afford to support more specialists leading to a virtuous circle.  And typically each specialist was installed in his own shop.  And at this point, advertising was not a problem.  There were still so few of them that everyone knew where everyone's shop was.

Probably the first type of business where some kind of advertising became important was the tavern or public house.  If a town was going to host more than one of a particular kind of business it would be a tavern.  Taverns started hanging up signs.  And, since most people still couldn't read, they went with easily identifiable symbols.  And as taverns multiplied in large metropolises like London or Paris, they started using with two symbols on a sign.  So you had the "rose and crown" or the "pig and whistle".  The sign would have a stylized but identifiable rose and a stylized but identifiable crown carved into it, as an example.

Probably the first standardized signage was the three gold balls that identified a pawn shop / money lender.  The second was the striped pole that identified a barber.  And the combination of signage and for a long time word of mouth was the entirety of an establishment's marketing campaign.

And for a long time all roads were crap.  So all larger cities were located on rivers or the seashore.  Back then ships were the only way to reliably and cheaply transport large volumes of goods.  The Romans (and the Chinese) had built good roads.  But the Roman roads stopped getting maintained when the Empire fell so they slowly decayed.  This means that outside of China the volume of commerce increased but it was still at a very low level.

If we move forward a few hundred years the high technology device of the day was the ship.  And better technology meant better ships, be they warships or merchant ships.  So there was money in the ship business pretty much before anything else.

And the complexity of an oceangoing ship, say of the fourteen hundreds, demanded a lot of specialized labor and equipment to build and maintain.  And that resulted in the first example of what we would now recognize as a mercantile operation, the ship's chandlery.  This was a retail establishment that specialized in selling all the bits and bobs that large, complex, ships needed.

And by this time more money was in circulation.  So we saw all kinds of specialty retailers following in the footsteps of chandlers.  And, for the most part, they followed the tavern model.  They would put a sign up outside with some kind of icon carved into it that indicated their area of specialization.

Later, when good quality and inexpensive window glass became available, a window display was added.  But their marketing campaign consisted primarily of "word of mouth".  This is when the phrase "by appointment to the King" came into general use.  This all changed with the advent of the newspaper.

It took a while for people (both the publishers and the readers) to figure out that placing an advertisement in a newspaper acted like word of mouth on steroids.  People read newspapers to find out what was going on.  It is a small step from learning the news of the day to learning where to find the best (or cheapest or closest) merchant selling a particular item.  The world has not been the same since.

To state the obvious, newspapers depend on a high level of literacy in the population.  Once a significant percentage of the population could read and write (and a process for making paper cheaply came into widespread use), trends in retailing started to be strongly influenced by trends in communication.  The other strong influencer is one I have already talked about, trends in transportation.

I noted that for a long time the siting of large cities depended on access to rivers or seashore.  This was because, as I noted above, for a long time volume transportation over substantial distances was only possible using ships.  If a merchant dealt in goods originating a distance away those goods traveled in ships.

So towns in the interior tended to be small.  And retailers in these interior towns tended to mostly to depend on the skill of their labor rather than their access to goods from a distance for their livelihood.  This severely restricted the types and amounts of commerce that was possible.

Roads slowly got better but the big revolution in land transportation came about with the introduction of the railroad.  It now became possible to transport large volumes of goods over long distances quickly and inexpensively.  Railroads came into existence in the 1800's.  And that roughly coincided with the introduction of the telegraph.  Telegraph messages were quick but they were also very expensive.  So their impact on retailing was limited.  It was cost effective to deal with suppliers via telegraph but it was not cost effective to deal with customers that way.

What railroads, and a reasonably quick postal system (perhaps a week to send a letter a thousand miles) enabled was the catalog retailer.  The most famous example of this was Sears, Roebuck, and Company.  Sears set up shop in Chicago, which had excellent rail connections to the rest of the country.  They sent out their "wish book" catalog and received orders by mail.  Then they used rail freight to deliver the goods ordered.  A wide variety of goods was available in a large city like New York City.  But Sears was able to provide a similar wide variety of goods to anywhere that had good rail service.

This revolutionized small town life.  It did it in two ways.  The first way is the obvious one.  People could order and receive a much wider selection of goods than local merchants stocked.  But the second revolution is often overlooked.  A customer could look in the Sears catalog and see what something cost there.  If the local merchant charged a lot more than Sears, this could be brought to the merchant's attention.  The merchant was forced to either lower his prices or see business decline.  A much higher level of competition was thus introduced into small town retailing.

But this increased level of availability and competition did not extend to everything.  There used to be something called a "dry goods" store.  Dry goods were the kind of thing you could ship via boat or train without damaging it.  So anything "perishable" still had to be sourced locally.  And most kinds of food is perishable.  Drying or pickling many kinds of foods ruins it.  Just ask sailors whether they preferred "salt beef" to fresh meat.

One way this played out has become a staple of American mythology.  If you built an insulated rail car and packed it with ice you could keep meat, particularly beef, from perishing while preserving its flavor.  The cheapest place to raise beef cattle was out on the prairie a long way from anywhere.  So, "beef cattle" were fattened up, say in West Texas, then they were "driven" to a railhead, say at Abilene, Kansas.  There, they were slaughtered and the beef loaded into chilled rail cars and shipped east.  This is the economics behind the "cattle drive" much beloved in Western movies and TV shows.

At about the same time Sears was getting into business (1886) there arose another phenomenon that I don't completely understand, the "department store".  These only existed in medium and large cities.  They were a large, single, unified, store that internally was broken up into a large number of departments.

Each department functioned along the lines of a specialty shop.  One department sold dresses.  Another department sold furniture. A single department store could host an amazing assortment of merchandise for sale.  And, other than the fact that it was all the kind of thing an average person would buy, there was little commonality involved.

I am old enough to remember when a number of these operations were still going strong in Seattle, my home town.  But why?  I can come up with two suggestions.  The first is management.  Not everyone knows how to run a successful retail operation.  And there is a considerable amount of "general retailing expertise" that applies across lots of kinds of merchandise.  My second suggestion is also the one I think is the important one, advertising.

A department store can afford to advertise regularly in a newspaper.  They can build up "brand loyalty" for the store.  "Shop with us and we will have everything you need and treat you right."  A department store can also feature goods from one department in one ad and goods from a different department in the next.

The consumer was continuously reminded of the existence of the department store due to the regularity of the appearance of an advertisement.  And over time each department in the store could be featured.  The customer was reminded of the many departments that could be found, all in one convenient place and clustered under a trusted name.

A stand alone operation  specializing in the goods from just one department could not afford the cost of duplicating the size and frequency of the newspaper ads the department store could afford.  They had to settle for ads that were either smaller or less frequent.

So the state of transportation at the time (railroads) and the state of communications at the time (newspapers, catalogs) shifted the retailing landscape dramatically in the late 1800s.  The advent of airplanes, radio, and later TV, didn't shift the retailing landscape as much.  But cars did.

The wide availability of cars made the shopping center possible.  Starting in about 1950 the shopping center was able to sell itself as a bigger, better, department store.  The shopping center could engage in the same kind of newspaper advertising department stores had been doing.  And they could offer more departments in the form of specialty shops.  And each specialty shop could offer a wider variety of goods than the department in a department store could.

But for a long time in spite of the fact that they were competitors, department stores and shopping centers coexisted.  In fact, a successful shopping center was "anchored" by one or more department stores.  The only change for department stores was now, instead of a single store in a single city, all the successful department stores had expanded into "chains".  At one time there was only one Macy's store in the world.  Now they are all over the place.

At the same time the catalog operations added stores to the mix.  You could order from the Sears catalog or you could go to the Sears store.  You could even order from the catalog but have it shipped to the store.  You would then use your car to pick it up.  J. C. Penney's evolved into another of these hybrid operations.  For a long time both Sears and J. C. Penney's (and others) had extensive catalog operations.  At the same time they had large chains of stores, often serving as an anchor to a shopping mall.

It would be interesting to have access to the books of these hybrid operations like Sears and Penney's.  I think there was a slow shift.  Catalog volume dropped off slowly.  But for a long time it remained profitable (I opine).  And for a long time the anchor stores at malls were also profitable.  But I suspect that the trend of business moving away from catalog and away from department stores in malls moved slowly enough that a case could always be made to continue on as before.  But the trend has finally caught up with them.

And coming up along the inside was a different retail model, the big box store.  "Big box" is often associated with Walmart.  But Walmart is more accurately seen as a department store chain.  They were just the first to figure out how to adapt the department store model to rural and suburban-fringe areas.  The secret to success was the combination of the interstate highway system and the car.  The customer density might have been low.  But Walmart figured out how of how to get people to drive long distances to one of their stores.

And at the time Walmart was growing quickly, rural and other low density areas were underserved by the retail industry.  So, Walmart didn't make a fundamental change to the industry.  They just figured out how to expand an already tried and true formula into a new market.

As I noted above, the size of the average specialty shop in a mall could be substantially larger than the department of a department store.  That trend got taken to extremes.  All of a sudden we started seeing very large stores dedicated to a single line of products.  A classic example is Best Buy, an electronics store.

But the grocery store, which started out pretty small, kept getting bigger and bigger.  It is now a "supermarket", and it is truly "super" in terms of size.  We also have big box book stores, furniture stores, sporting goods stores, etc. now.  Most often they are in malls.  But sometimes they are like Walmart's, a free standing operation,

And, of course, what makes the big box store feasible is the same thing that made Walmart stores feasible.  People can now drive long distances to find what they want.  You just have to give them sufficient reason.

And then coming up fast on the rail you have the internet and Amazon.  The internet is a communications revolution.  The equivalent of a Sears catalog can now be delivered to your home on demand.  And the catalog entries can be updated instantly.  And the cost of creating a "catalog" entry on the Amazon web site is far less that the equivalent cost for Sears to create an entry on its catalog.

The "Amazon story" is now well known.  It started out selling books.  Then it branched out by adding this and that.  It has since added third party sellers.  So now you can get almost anything through Amazon.

And this whole "Amazon is killing retail" story is now well known.  My local mall has a big Barnes & Noble bookstore but B&N is reputed to be hanging on by a thread and most of it's competitors are already gone.  Video rental stores are gone.  And so on and so on.  The rapid expansion of the big box trend was supposed to be an "Amazon beater" strategy.  It worked for a while.  But it looks like it lacks long term staying power.

Retailing at all levels has had to adjust to the internet.  Many companies quickly added an internet along side the rest of their operation.  But for most it was supposed to be an adjunct to their main retail business.  Barnes & Nobel, for instance got into the web in a big way.  They even developed the proprietary Nook e-reader.  The conventional wisdom on all of this is now "too little - too late" and "doomed to failure".

But here's where I think I have something fresh to contribute.  And that is this.  I think retailing is not doomed.  I just think we need to do a rebalancing.

Nordstrom's is a department store chain that people used to think was getting it right but those same people now think they are in trouble.  And right this minute they are.  They put in lots of large retail stores in malls and for a long time that strategy worked well.  They were one of the first companies to create an internet presence so people figured that they had done the right thing here too.  But now people are not so sure.

The "all internet - all the time" people figured that all of retailing was going to follow the move rental model.  But I think that is too extreme.  What I expect to see is a rebalancing.  Neither retail sales nor internet sales will go away completely.  But I see the internet side becoming more important and the retail side less so.  In the specific case of Nordstrom's, I don't see them going down the drain.  I do see them needing to downsize their retail presence (smaller stores) while continuing to enhance their internet presence.  And that will be difficult and expensive to pull off.

People like to shop at Nordstrom's.  They got the whole "shopping as theater / entertainment" thing well before most of their competitors.  So their stores are very pleasant places to hang out.  But stores cost a lot of money to operate.  When almost all of Nordstrom's sales came through their stores this worked fine.  Their stuff was a little more expensive that what could be found elsewhere.  But people figured the price difference was worth it and Nordstrom's bottom line reflected that for many years.

But let's say you are a Nordstrom's regular.  If you are a "fast fashion" type who always has to have whatever is trendy RIGHT NOW then time spent in the store is still well worth it.  But most of us, including the fast fashion types, also buy a lot of staples.  When it comes to a lot of what we buy, we know exactly what we want before we enter the store.  For these purchases the in-store experience is less valuable.  So why not just order it off the Nordstrom's web site?

Nordstrom's needs to hold on to their customer base, and expand it, if possible.  For new customers the in-store experiences is the best way for Nordstrom's to introduce itself.   And for the person who needs assistance, help finding the right size or style or color or accessory, you can't beat the in-store experience.  The problem is "how does Nordstrom's (or any other retailer) get the additional expense of their retail operation to pencil out".

I think the way Nordstrom's needs to do that is by downsizing their stores, thus reducing the cost of their retail presence.  The remaining retail side of the business would be much more tightly focused on the customer service aspects of their business.  Nordstrom's runs an extensive cosmetics operation.  Providing help finding just the right shade of lipstick or other cosmetic is something customers regularly need and appreciate.

Taking the measurements necessary to make sure clothing fits properly is also important.  But if a complete set of measurements is taken and entered into Nordstrom's computer system, that system can then apply those measurements to all the clothes Nordstrom's offers in a way that insures that all the clothing the customer orders will fit properly.  That would be especially helpful to women as it is harder for them to achieve the correct fit.

Finding the right store size and the right mix of products and services will be hard to figure out.  And getting from where store chains like Nordstrom's are, namely lots of big stores in lots of malls, to where they need to be, smaller stores, perhaps in more locations, perhaps in fewer, will be even harder to do.  Nordstrom's is trying to pull exactly this off.  I wish them luck.

And Amazon is coming at it from the other side.  Like Sears, Amazon started out with no retail presence.  But Amazon has been experimenting with physical stores for some time now.  Their recent efforts include purchasing the "Whole Foods" supermarket chain, and rolling out "Amazon Go" cashier-less stores.  Both of these initiatives have attracted a lot of attention.  And no one, Amazon included, is sure where either initiative will end up.

And taking the wide view of retailing, there is another trend going on.  Lots of malls were built.  The theory was "there is no such thing as too many malls".  That is obviously nonsense.  But it took a long time for this sector to get overbuilt.  Then, among other things, the internet came along.  And, as I have argued above, we need less retail than we used to.

But at the same time, cities like Seattle, changed their zoning so that pretty much every building had to include row of small retail stores at street level.  At one time there was probably a shortage of spaces for small retail.  But the zoning was changed quite a long time ago now so there is now a substantial surplus of these sites for small retail, at least as I see it.

This combination of a surplus of retail space, large and small, and a decreasing need for all kinds of retail, means I see trouble in the market for retail real estate.  Lots of malls went out of business in the past decade.  I think most of the damage has been done in that area.  But, here too the shrinkage will continue.  And I see particular trouble for anchor stores in malls.  I see the occupants of those malls staying in the mall but I see them downsizing into smaller spaces.  And I see building owners having trouble keeping the small retail the City forced them to build full.

And as long as we are talking about external trends that affect retailing, here's a final one.  The entire world has been shifting from a middle class based economy to a banana republic-style model.  We will have a few super-rich people and a lot of poor people.  The middle class has been shrinking as a percentage of the total population and I expect this to continue.  (For the moment, China is an exception here.  China has a growing middle class.  But they are coming from no middle class at all.)

Each individual rich person buys more stuff than poorer people do.  But there aren't that many of them.  So the aggregate demand they represent is small.  What drives a healthy retail economy is a large middle class.  Even more helpful, from a retailing perspective, is to put more money into the pockets of poor people.  Poor people spend what they have and they spend most of it on retail.  Middle class people spend almost as high a percentage of their money on retail as poor people do.  Rich people spend a far lower percentage of their money on retail than do people in either of the other two groups.

There are far more poor people than rich people.  Even with their reduced numbers, there are also far more middle class people than rich people.  If you add 1% to the income of every rich person there will be essentially no impact on retailing.  If you add 1% to the income of every poor person you will see almost 100% of that money flow through to retailers.  But poor people have little money so that 1% change will add up to a relatively small amount.  If you add 1% to the income of middle class people a lot of it will flow through to the retail segment.  And it will be a large amount in the aggregate.

What we have been doing is taking new income away from the people who underpin retailing, the poor and the middle class, and giving it to the rich, the people who do the least for retailing.  We have seen growth in retail.  But it has been almost entirely funded by the poor and the middle class taking on more debt.  That is unhealthy.  The best thing that could be done to improve the lot of retailers is to direct new income away from the rich and toward the middle class and especially the poor.  But that does not seem to be in the cards any time soon.  And that is going to put retailing into yet another bind.

U. S. Mail - Number Please

I like to know how things work.  And in almost every case the current system for how a particular thing works evolved in stages over time from humble origins.  This post covers the evolution of Mail Addresses into numbers.  It turns out that every single place the United States Postal Service (USPS) is capable of delivering mail to has a specific unique number associated with it.  How can that be?  Well, it didn't start out that way.

The phrase "Number Please" conjures up the image of a telephone operator.  Back in the day a lot of the telephone system was operated manually.  An "operator" sat at an "exchange".  When you wanted to make a phone call you did something (cranked on a handle in the early days) to get the operator's attention.  This lit up a light on the "plug board" of the exchange.  The exchange was manipulated by the operator by plugging and unplugging "patch cords".  So the operator would connect herself to you, ask for the phone number you wanted to connect to, then manipulate some patch cords to connect you up.

In the very early days you could just say "connect me to Mable" and the operator knew what to do.  But phones soon became so popular that asking for someone by name soon became unwieldy and everyone who had a phone was assigned a number.  For a long time the "number" was a combination of letters and numbers but that was actually a con.  If you look at a phone "dial" you will see that the "2" key is labeled "ABC", the "3" key is labeled "DEF", etc.

So if my phone number was "Olive 7335", which it was at one point, that really meant OL7335, "OL" being the first two letters of "Olive".  And the "O" was really a "6" and the "L" was really a "5".  So, for all practical purposes my phone number was 657335, a number.  At some point the phone company decided that this "letters and numbers" business was not worth the bother and dropped it.  People are now used to phone numbers being a series of digits.

But in an interesting development, "phone numbers" are going back to being names.  Most people now use smart phones.  And to call somebody they look that person up in their "Address Book" and hit the "dial" button.  People only know another person's actual phone number just long enough to enter it into their address book.  After that, they go by their name.  (The same goes for texting, which is rapidly replacing the phone call.)

Back in the day (early '60s) there was a British TV show called "Danger Man".  It was syndicated in the US under the catchy name of "Secret Agent".  And for the US run, besides giving the show a new name, they slapped on a catchy new theme song.  That song contained the line "they've given you a number and taken 'way your name".  The USPS has given every mail delivery location a number but they haven't taken it's "address" away -- yet.  But that's the end of the story.  Let's go back to the beginning.

For a long time almost everybody was illiterate.  Before paper was invented writing material was very expensive so most people found a way to get along without learning to read or write.  Then paper was invented.  This was followed by "movable type" printing.  Now there was much more of an incentive to learn to read and write, and lots more people did.  And if two people can read, information can be shared by one person writing a "letter" to the other person.  The problem was how to get the letter from the party of the first part to the party of the second part.

Well, most of the people the typical person knew didn't live that far away.  So talking a friend or acquaintance into delivering the letter worked reasonably well.  But governments and some business people found it valuable to communicate over longer distances.  And for a long time the "find a friend or acquaintance who is going that way" was the only option available.

Governments have the most money and the most need of long distance communications.  So governments started setting up regular messenger services.  Messengers would travel a fixed route on a fixed schedule carrying letters.  At first these messenger services only transported "official" letters pertaining solely to government business.  So business groups ended up setting up parallel messenger systems that handled commercial letters.

Eventually sanity intervened and the systems were consolidated into a single messenger system that handled both government and business correspondence.  But these systems were complicated, idiosyncratic, and expensive.  And most people still couldn't read or write.

Fast forward to the American Revolution and Ben Franklin.  Literacy in the "colonies" was increasing and Franklin recognized that an inexpensive, simple, reliable, Mail system was valuable to everyone.  So he made sure that the US Constitution (Article I, Section 8, clause 7) empowered Congress to "establish Post Offices and post Roads".  The US government has been in the Post Office and Mail delivery business ever since.

One idea Franklin pushed successfully was for a standard cost to send a letter regardless of how far it was going.  This drastically reduced the complexity involved in using the system.  You addressed a letter.  As long as you were in an area served by the USPS and you were sending a letter to a destination served by the USPS you paid the same low price ("for the low, low, cost of a stamp . . .").  This meant mail going a short distance subsidized mail going a long distance.  But initially very few letters went a long distance.

Unfortunately, junk mail has changed that.  We can now get bombed with catalogues, sales flyers, etc., from all over the country.  And this ability to send a letter from anywhere in the country to anywhere in the country demands a uniform system for identifying where the letter is supposed to go.

And, as with the case of telephones and operators, initially whatever the system was it would be operated by people.  And people are very flexible.  So early on locations had what amounted to a nickname.  There is the famous "Bleak House" of literature fame.   The house Winston Churchill lived in most of his life was called "Chartwell".

Once you got close, people were just supposed to just know where it was.  In order to facilitate the "get close" part of the process, the destination was additionally identified (in the case of Chartwell) as being near the town of Westerham in the district of Kent in the country of England.  And back in the days when only rich people in fancy homes got mail that system was sufficient.

About this time a standard hierarchical form was adopted.  The bottom line of the address contained the most general type of location information, typically a city, state, and perhaps a country.  The next line was more specific, typically a building number and street name.  And the line above was still more specific, typically the name of the recipient.

Thanks to Franklin and others, more and more people were availing themselves of the postal system.  And as the number of users increased, the stateliness of the typical mail recipient's abode declined.  Just going with a name like "Chartwell" stopped being tenable.  So, an effort was made to give all streets in a town a unique name and each building on a given street was given a unique number.

Addresses of the form "123 Elm Street, Anytown, Anystate" came into general use.  Around the World War II time period this was not always enough so large cities started getting broken up into "zones".  So it became "Anytown 3" to indicate that the address was located in zone 3 of Anytown.  This meant that letter sorters didn't need to be familiar with all of a large city.  They just needed to know their way around whatever zone they were working in.

But by World War II mail volumes were large and a lot of the mail traveled a significant distance.  By this time the USPS was organized into a hierarchy.  The country was divided up into "sections".  If a letter was not going to someplace local it was sent to one of a limited number of "sectional center facilities".  Internally within the USPS these facilities were assigned a 3 digit number.  This represents the third incursion (after street numbers and then city zones) of numbers into the business of delivering the mail.

Even after the War a lot of flexibility was permitted because the system was still a manual one, for the most part.  A person can easily determine that "Wa", "Wash", and "Washington" all refer to the same state so all three forms (and several others) were permitted.  And address lines that became too long were not a problem.  If it looked like there would be a problem then whoever was doing the addressing would just do something like go with an abbreviation instead of spelling things out in full.  The mythical "Frostbite Falls, Minnesota" worked just fine, for instance.

Then in 1963 a switch was made.  The postal zones within cities were dropped.  And two more digits were added to the sectional center facility number.  These two extra digits allowed a specific post office within a section to be identified.    And the system was publicized and given a catchy name, the "ZIP Code".  People and businesses were instructed to add the ZIP Code to the city-state line.  But adding the ZIP code could be a problem if the city-state line was already quite long.

So, at the same time a list of two letter state codes were promulgated and people were encouraged to use them instead of all the earlier (and usually longer) variants.  So my state became "WA" and that shortening allowed the ZIP Code to fit right in.

In order to encourage the widespread adoption of ZIP codes starting in '67 bulk mailers were required to use them if they wanted to mail at a discount.  So every Post Office in the country had its own unique ZIP Code.  This got rid of a lot of problems caused by near-duplicate addresses.  Any town could have a "123 Elm St.".  And it was possible that two close together towns with similar names could have the same address.  When the ZIP Code boundaries were set up the USPS made sure that each instance of "123 Elm St." (or any other specific address) ended up in a different ZIP Code.

And the 5 digit ZIP code was a big help.  But there were still a lot of addresses within a single ZIP Code.  So in 1983 the USPS went to a ZIP+4 system.  They added a "+" and another 4 digits to the end of the ZIP Code.  They also started rolling out a barcode system.  All of a sudden, a bunch of ticks started appearing below the address.  These ticks were designed to be read easily with a dumbed down version of the UPC Code scanners used in supermarkets.

The original barcode contained only the 5 digit ZIP code.  But improvements and enhancements were added over the years.  First it was expanded to include the ZIP+4 information.  Finally a two digit "Delivery Point" number was added.  A 5 digit ZIP Code gets us to the correct Post Office.  A ZIP+4 Code gets us to the specific delivery route (or part of a route) operating out of that Post Office.  The addition of the two digit "Delivery Point" number gets us to a single specific house, apartment unit, etc.

This 11 digit number (a 5 digit ZIP, plus a 4 digit +4, plus a 2 digit delivery point) gets us to a single number that uniquely identifies any specific location capable of receiving mail anywhere in the US.  And in 2012 the latest version of the USPS barcode was introduced.

It is called the "Intelligent Mail" (IM) barcode and it became mandatory in 2013.  It is much more elaborate than even the 11 digit barcode it replaced.  The old barcode used only short ticks and tall ticks.  So that each tick encoded one bit of information.  The IM barcode used four tick types - short-centered, short-centered+up, short-centered+down, and long (a combination of short-centered and the "up" part and the "down" part of the other two ticks).  So each tick encodes two bits of information.  The IM barcode consisted of 65 ticks encoding 130 bits of information.

Ignoring some "overhead" ticks (you need to know if you are reading the barcode upside down or right side up, for instance) the "payload" translates to a 31 digit number.  11 of those digits are used to encode the ZIP+4+2 11 digit number described above that identifies the unique delivery point of the letter.  Most of the rest of the digits are used to create a tracking number so letters or packages can be tracked like they can at UPS or FedEx..

The tracking number is broken into two sub-parts.  The first part is the sender identification number and the second part is the item number.  The USPS assigns a sender identification number to each participating company or organization.  A unique item number is then assigned by the sender to each item that sender mails.  Two formats are used.

Some of the "sender" numbers are short.  They are assigned to (relatively) high volume mailers.  There are fewer numbers but there are also fewer high volume mailers than there are (relatively) low volume mailers.  By keeping the number short for some mailers there is room for a larger pool of item numbers for these mailers to use.

Other "sender" numbers are longer.  This allows the system to accommodate a larger number of low volume mailers.  But the trade-off is that these mailers have a smaller pool of item numbers to work with.  But they shouldn't need as many as the high volume mailers so it is a good trade-off.

In either case a unique item number is assigned to each piece of mail.  But the number only needs to be unique within a 90 day window.  After that it can be reused.  The combination of the sender identification number and the item number is a fixed size.  So the combination always takes up the same amount of space within the 31 digit number.

So it is cool that the USPS can now track an item.  But that's not the main point of adopting barcodes.  The main point was to permit machine sorting.  A mail sorting machine can sort mail by reading the barcode.  It doesn't have to have the smarts to handle all the variation in regular printed addresses.  And it is still legal to write an address by hand.  No machine can read handwriting reliably.  And they definitely can't do it at high speed.

There is a nice feature that this 11 digit unique delivery point number enables.  I was sending an Amazon package to a cousin a few days ago.  I filled in the address and Amazon came back with "is this the address you mean".  It had made a small change to what I had entered.  There are now lists available to Amazon and others of all the valid addresses in the US.  Amazon chunked through the address I had provided and found no match.  But it did find a very similar address, which it suggested to me.

A similar thing happened to me several months ago when I was at the USPS mailing a package.  The staffer entered the address I gave and it came up "tilt".  It turned out that the person who had given me the address had included a ZIP code that was completely wrong.  With a little poking around the staffer was able to suggest to me a completely different ZIP code that turned out to be the correct one.  That's definitely a good thing.

So is all this automation helping?  I'm sure it is.  But the cost of a First Class stamp has been rising quickly over the last several years.  (And I'm old enough to remember when the cost of a First Class stamp went from two cents to three cents.)  What's going on is that mail volume is dropping.  But it costs a lot (and costs keep rising) to keep all those Post Offices running all over the place.  City people like me heavily subsidize country people.

There is a high enough volume of mail to cover the overhead in cities.  But that is not true in rural areas.  A rural Post Office may cost less to run.  But it costs a lot more on a per-letter basis to run than an urban Post Office does.  The very people who benefit from this cross subsidization are the ones who scream loudest about how much it costs to keep the USPS going.  But they also don't want any of those rural Post Offices, the ones that lose so much money, closed.

I know what my ZIP+4 ZIP Code is.  But I don't know what my Delivery Point number is.  And if there is a way to find out what it is I haven't been able to find it..  So, I only know 9 digits of the 11 digit number that represents my delivery address.  Could the USPS switch over to the 11 digit number at some point?  Sure.  But it doesn't look like they are going to do it any time soon.

But there are lots of businesses that know what my 11 digit address is.  You see many bills and pieces of junk mail come with that IM barcode printed on it.  And I can tell it was printed at the same time my address was printed.  It wasn't added later by the USPS.  Instead, it's right there underneath the address that was printed by the people who are sending the mail to me.  And you can't create an IM barcode without knowing what my particular and unique 11 digit address is.

And on an apparently unrelated but actually related subject, the decennial US Census is coming up in 2020.  The Census people need to know where to look to find everyone.  Well, guess what?  The magic database that is used to validate addresses and to list what each address's Delivery Point is tells you almost every place to look for people.

It's not going to catch people who are living in their cars or otherwise have no fixed address.  But if you regularly get junk mail then your address should be in the Delivery Point database.  The Census people were able to work with bulk mailers to develop a "where to look" list when they were doing the 2010 Census and that was helpful.  But this Delivery Point database business means that things have gotten much better organized since.  And the Census people should be able to take advantage of that.

Giving people numbers is not just for secret agents any more.