Thursday, November 09, 2006
What a Difference a Point Makes
Last week, we reported about September printing industry shipments, and noted that the September to November period is the busiest of the year, starting that they were 26% of the year's annual shipments. We received some notes that were along the lines of “So what? If three months is 25% of the year, what difference does that little percentage point really make?”
That's a very good question. Facetiously, the answer is about $900 million dollars for a year. We didn't explain why it's important, or how these months, and others, have changed over time. This September-November represented 26.56% of industry shipments from 1992 to 2000, but this dropped to 26.17% in the period of 2001 to 2005. In 2005, it was 26.02%. The historic seasonal variations between months and quarters is lessening considerably to create a flatter shipments pattern.
Over time, even the calendar quarters have changed. The fourth calendar quarter (October-December) used to be 26.7% of the year's shipments. In 2005, it was 25.4%. The first calendar quarter (January-March) used to be the weakest, at 24.1%; last year, it was 24.9%. The weakest three months of the year were traditionally May through July, at 23.89%; last year they represented 24.28% of shipments.
January was typically the least active month of the year, at about 7.7% of the year's shipments. Last year, January was 8.1%, and July became the slowest month, with 7.8% of the shipments. The strongest month has always been October, but even that has slackened. From 1992-2000, it represented 9.15% of shipments. Last year, it was 8.88%. March is growing in importance.
Each month of the calendar is about 1/12 of a year, or 8.25%. The best way to show the difference between the months is to calculate how each month differs from 8.25%. The average variation was + or -0.27 percentage points in 2005, or from 7.98% to 8.52%. That's very small. From 1992-2000, the average variation was + or -0.41 percentage points, or from 7.84% to 8.66%. That's not even half of a percentage point, but the change shows that the months are becoming more alike.
What is the reason for this change in seasonality of printing shipments? It's quite simple, or at least the hypothesis is simple. Print's product mix is changing, and it's primarily the change in printing purchased by retailers that has affected the seasonal pattern. Just this past week or so, there have been many reports about the decline in newspaper circulation. These declines are even in Sunday circulation, which had been steadier circulation than the other days of the week. Declining Sunday circulation means that there are declines in the number of newspaper inserts needed. The rise in direct mail, though a boon to some, is a sign of the rise of e-commerce, resulting in many catalogers have reducing frequency and page counts of their editions, in favor of direct mail promotions. The result is less printing of one type, more of another, and it nets out to an overall loss in volume.
The real change is not a surge of volume in any month, it's a reduced volume of end-of-year printing related to retailing that has given other months a greater piece of the shrinking pie of printing shipments. Retailers have also be spreading many of their dollars throughout the year because consumer spending is changing. Note how Halloween has become the second biggest “holiday” for spending on decorations and the like. Also, as the population has gotten older, Christmas gifts are more likely than ever to be “experiences” and not hard goods. Gifts of vacations, travel, and event attendance are consumed year-round. Generally rising affluence, and smaller families, also means that Christmas is less of a focus for spending. The rise in self-employment and changes in corporate pay policies mean that Christmas and holiday bonuses have become less important to many households as part of their total income and plays less of a role in the timing of their spending. The rise of the credit card business also means that Santa can work his special magic at any time of the year.
What's the new print year look like? March is gradually becoming our new October. That is, as end-of-year printing recedes, other months are becoming arithmetically more significant. But it's just numbers. What they indicate is that the print business continues to change and that old rules of thumb and common wisdom about strong or weak parts of the year are out the window. That's what caught many people by surprise in 2005, as they kept saying “but the year always finishes strong,” and then, unfortunately, it didn't. Now the pattern is gradually becoming “one month or quarter is just like another.” Anyone who complained about how difficult it was to cope with seasonal swings in their volume should now be happy. Those seasonal swings helped define our industry. It's like having the hiccups... once they're gone, we miss them.
Regarding the recent increase in print volume and its possible attribution to political spending, we have this to add. A client brought this question up as they looked at charted data as saw some rises in even-numbered years. There aren't that many years to analyze, since the data series that we use start in 1992. What we can say is that up until 1998, there were no significant monthly changes in shipments compared to odd-numbered years that could be solely attributed to political printing. From 2000 to 2005, however, there appears to be upward bumps in the share that election months have in the year; but it's not that easy. In even numbered years, the entire period from July through December is stronger than in odd-numbered years, including a robust December. This is where the statisticians say “that's interesting,” and then go onto the next topic.
Right now it's just a statistical curiosity. The political season explanation is plausible, because the decline in retail printing may make political spending more pronounced and easier to detect, and therefore it may be an important factor to consider in prognostication, when in the past it was not. It's like the analogy of a river: when the water level decreases, you can see rocks and formations that were previously hidden from view.
To be certain of the validity of this new pattern requires more data than we have. That statistical disclaimer is, of course, absolutely no help to executives who have to make decisions today. This is definitely one of those “sounds good, let's run with it until we learn otherwise” situations that experienced executives, and good consultants, know quite well. The honest ones are always on the lookout for “otherwise” and those things that go against common wisdom. Among other factors that play into a short-term rise in print are the shortage of experienced new media personnel, the shortage of good Internet properties on which to advertise, and the price increases that many new media properties have demanded for ad placement for their best viewing. The lack of local new media properties that could address political needs may have sent more money to print in some extremely contentious races; national races may benefit most from online advertising. The dollars end up going to the payback of least resistance, and temporarily, that may be print. It could very well be the case that both trends are creating the recent print increase.
Online media and political advertising
http://www.clickz.com/showPage.html?page=3623858
That's a very good question. Facetiously, the answer is about $900 million dollars for a year. We didn't explain why it's important, or how these months, and others, have changed over time. This September-November represented 26.56% of industry shipments from 1992 to 2000, but this dropped to 26.17% in the period of 2001 to 2005. In 2005, it was 26.02%. The historic seasonal variations between months and quarters is lessening considerably to create a flatter shipments pattern.
Over time, even the calendar quarters have changed. The fourth calendar quarter (October-December) used to be 26.7% of the year's shipments. In 2005, it was 25.4%. The first calendar quarter (January-March) used to be the weakest, at 24.1%; last year, it was 24.9%. The weakest three months of the year were traditionally May through July, at 23.89%; last year they represented 24.28% of shipments.
January was typically the least active month of the year, at about 7.7% of the year's shipments. Last year, January was 8.1%, and July became the slowest month, with 7.8% of the shipments. The strongest month has always been October, but even that has slackened. From 1992-2000, it represented 9.15% of shipments. Last year, it was 8.88%. March is growing in importance.
Each month of the calendar is about 1/12 of a year, or 8.25%. The best way to show the difference between the months is to calculate how each month differs from 8.25%. The average variation was + or -0.27 percentage points in 2005, or from 7.98% to 8.52%. That's very small. From 1992-2000, the average variation was + or -0.41 percentage points, or from 7.84% to 8.66%. That's not even half of a percentage point, but the change shows that the months are becoming more alike.
What is the reason for this change in seasonality of printing shipments? It's quite simple, or at least the hypothesis is simple. Print's product mix is changing, and it's primarily the change in printing purchased by retailers that has affected the seasonal pattern. Just this past week or so, there have been many reports about the decline in newspaper circulation. These declines are even in Sunday circulation, which had been steadier circulation than the other days of the week. Declining Sunday circulation means that there are declines in the number of newspaper inserts needed. The rise in direct mail, though a boon to some, is a sign of the rise of e-commerce, resulting in many catalogers have reducing frequency and page counts of their editions, in favor of direct mail promotions. The result is less printing of one type, more of another, and it nets out to an overall loss in volume.
The real change is not a surge of volume in any month, it's a reduced volume of end-of-year printing related to retailing that has given other months a greater piece of the shrinking pie of printing shipments. Retailers have also be spreading many of their dollars throughout the year because consumer spending is changing. Note how Halloween has become the second biggest “holiday” for spending on decorations and the like. Also, as the population has gotten older, Christmas gifts are more likely than ever to be “experiences” and not hard goods. Gifts of vacations, travel, and event attendance are consumed year-round. Generally rising affluence, and smaller families, also means that Christmas is less of a focus for spending. The rise in self-employment and changes in corporate pay policies mean that Christmas and holiday bonuses have become less important to many households as part of their total income and plays less of a role in the timing of their spending. The rise of the credit card business also means that Santa can work his special magic at any time of the year.
What's the new print year look like? March is gradually becoming our new October. That is, as end-of-year printing recedes, other months are becoming arithmetically more significant. But it's just numbers. What they indicate is that the print business continues to change and that old rules of thumb and common wisdom about strong or weak parts of the year are out the window. That's what caught many people by surprise in 2005, as they kept saying “but the year always finishes strong,” and then, unfortunately, it didn't. Now the pattern is gradually becoming “one month or quarter is just like another.” Anyone who complained about how difficult it was to cope with seasonal swings in their volume should now be happy. Those seasonal swings helped define our industry. It's like having the hiccups... once they're gone, we miss them.
Regarding the recent increase in print volume and its possible attribution to political spending, we have this to add. A client brought this question up as they looked at charted data as saw some rises in even-numbered years. There aren't that many years to analyze, since the data series that we use start in 1992. What we can say is that up until 1998, there were no significant monthly changes in shipments compared to odd-numbered years that could be solely attributed to political printing. From 2000 to 2005, however, there appears to be upward bumps in the share that election months have in the year; but it's not that easy. In even numbered years, the entire period from July through December is stronger than in odd-numbered years, including a robust December. This is where the statisticians say “that's interesting,” and then go onto the next topic.
Right now it's just a statistical curiosity. The political season explanation is plausible, because the decline in retail printing may make political spending more pronounced and easier to detect, and therefore it may be an important factor to consider in prognostication, when in the past it was not. It's like the analogy of a river: when the water level decreases, you can see rocks and formations that were previously hidden from view.
To be certain of the validity of this new pattern requires more data than we have. That statistical disclaimer is, of course, absolutely no help to executives who have to make decisions today. This is definitely one of those “sounds good, let's run with it until we learn otherwise” situations that experienced executives, and good consultants, know quite well. The honest ones are always on the lookout for “otherwise” and those things that go against common wisdom. Among other factors that play into a short-term rise in print are the shortage of experienced new media personnel, the shortage of good Internet properties on which to advertise, and the price increases that many new media properties have demanded for ad placement for their best viewing. The lack of local new media properties that could address political needs may have sent more money to print in some extremely contentious races; national races may benefit most from online advertising. The dollars end up going to the payback of least resistance, and temporarily, that may be print. It could very well be the case that both trends are creating the recent print increase.
Online media and political advertising
http://www.clickz.com/showPage.html?page=3623858
# posted by DrJoeWebb @ 5:00 AM 
