SPA Software Sales Figures

You've probably read quite a few reports about Mac software sales declining recently. The only problem with this is that it just isn't true. This page discusses the major source of that information--the Software Publisher's Association. If anyone can fill in any of the blanks in the tables, I'd appreciate it.

 

FLASH: Since this page was written, SPA has dropped their survey. While they don't admit it, I believe that they have finally realized that their survey was producing inaccurate results. They state that they're looking for another way to measure software sales.

 

I will also add that I have received (second hand) a response from the SPA. Some of the information I posted earlier was apparently incorrect. I was happy to correct the information below based on the feedback from the SPA and others. These changes do not change the fundamental conclusions.

SPA recently added a section to their web page. In this section, they discuss frequently asked questions on their survey. My favorite part is where they state that you can see how the survey is "REALLY" done (clearly a dig at people who criticize their results). The problem is that they just don't get it. Their methodology is producing incorrect results and their own data proves it. They should be working on fixing the problem rather than defending erroneous results. Please contact the links at the bottom of this page asking them to correct their methods.

There are several major sources of error in SPA figures:

So, the figures are strongly biased. Let's look at the figures as they've been presented, along with a few simple calculations based on them. In all cases, SPA figures are in blue and my calculations are in red.

The interesting thing is that the SPA has reported that Mac software sales are "in decay" and that they've declined for at least seven quarters in a row. Clearly, their own final figures contradict that statement.

It's worthwhile commenting on how other analyses compare. While SPA says Mac software counts for 10-13% of total software sales, NPD's SoftTrends puts the Mac market share at 12-30% depending on the demographic segment.

Another study was done by a Dr. Robert Resnick. I don't have a URL, but it was reported in the SPA magazine called Upgrade. Dr. Resnick reported that for the 1996/1997 school year to date, educational MacOS software sales are up 20.5%. Windows sales are up 8.1% and DOS sales are down 5.4%.

1. Total Mac Software Sales

All numbers are in Millions of dollars and cover North American sales, only.

  SPA Initial Report of Mac Software Sales SPA Final Report of Mac Software Sales Percent Increase Based on SPA Press Releases (1) Percent Increase Based on Final Report
 1990   614    
 1991   824   +34.2 %
 1992   990   +20.1 %
 1993   1,052   +6.2 %
 1994 1,250 1,225 18.8 % (2) +16.4 %
 1995 1,054 1,530 -14 % +24.8 %
 1996 1,200   -23 %  

(1) SPA press release uses their initial estimate for the current quarter or year vs. the final estimate for the previous period. Note that the "SPA Press Release" figures are widely published. The "Percent Increase Based on Final Report" figures never appear any where.

(2) Note in this case that they overestimated Mac software sales. But they overestimated Windows sales by an even larger amount, so they still underestimated Mac market share.

 

2. Total Industry Software Sales

All numbers are in Millions of dollars and cover North American sales, only.

  SPA Initial Report of Software Sales SPA Final Report of Software Sales Percent Increase Based on SPA Press Releases (1) Percent Increase Based on Final Report
 1990 4,586  
 1991 5,710 5,036 24.5 % 9.8 %
 1992 5,745   14.1 %
 1993 6,810 6,329 18.5 % 10.2 %
 1994 7,380 6,718 16.6 % 6.1 %
 1995 7,525 9,787 12.0 % 45.7 %
 1996 10,600   8.3 %  

(1) SPA press release uses their initial estimate for the current quarter or year vs. the final estimate for the previous period.

So, what we have is that the initial estimates (which are widely publicized) consistently (except for 1995) overstate total market sales. Since the release of Windows 95 and the resultant media frenzy, they also underestimate Mac sales. By the time the final figures are released, they are old news and are never publicized. So the growing Mac market is constantly reported as dead while the modest Windows growth is exaggerated.

3. Mac Market Share

All numbers are in Millions of dollars and cover North American sales, only.

  SPA Initial Report of Mac Software Sales SPA Final Report of Mac Software Sales Market Share Based on Initial Report Market Share Based on Final Report
 1990   614   13.4 %
 1991   824   16.3 %
 1992   990   17.2 %
 1993   1,052   16.6 %
 1994 1,250 1,225 16.9 % 18.3 %
 1995 1,054 1,530 14.0 % 15.6 %
 1996 1,200   11.0 %  

Note that the "Market Share Based on Initial Report" figures are widely published. The "Market Share Based on Final Report" figures never appear anywhere.

Keep in mind, too, that 1995 and 1996 were exceptionally high years for Windows as the release of Win95 made so much software obsolete. Already, 1996 sales figures are showing single digit growth rates for total software sales.

Also, keep in mind the fact that hybrid CDs had to be allocated by the software vendor. It appears that the share of hybrids assigned to MacOS may be too small.

Still, even without including the major sources of error, in each year for which data was available, the initial (widely publicized) SPA data underestimated the actual Mac market share by a significant amount.

4. Quarterly Results

The above figures are based on SPA annual figures. The same trend is just as clear from quarterly results. These figures are especially troublesome in light of SPA's statement that Mac sales have decreased for the last seven quarters and that the Mac software market is "in decay." Their own figures do not support these statements.

All numbers are in Millions of dollars and cover North American sales, only. All increase numbers are based on the same quarter in the previous year. Paul Durrant has conveniently listed links from this data to the original source on his web page.

  SPA Initial Report of Mac Software Sales SPA Final Report of Mac Software Sales Percent Increase Based on SPA Press Releases (1) Percent Increase Based on Final Report for Both Years Percent Increase Based on Initial Report for Both Years
 Q1 1993   218      
 Q2 1993          
 Q3 1993   270      
 Q4 1993   321      
 Q1 1994 259 279 +18.9 % +28.3 %  
 Q2 1994 276 244  
 Q3 1994 316 327 +17.0 % +21.3 %  
 Q4 1994 397 426 +23.6 % +32.7 %  
 Q1 1995 308 335 +10.0 % +19.9 % +18.9 %
 Q2 1995 226 376 -7.0 % +54.5 % +28.1 %
 Q3 1995 242 375 -26.0 % +14.7 % -23.4 %
 Q4 1995 361 477 -15.2 % +11.9 % -9.0 %
 Q1 1996 318 307 (2) -5.0 % -8.3 % +3.5 %
 Q2 1996 297 -21.1 % +31.0 %
 Q3 1996 240   -36.1 %   -0.8 %
 Q4 1996 334   -30.0 %   -7.4 %

(1) SPA press release uses their initial estimate for the current quarter or year vs. the final estimate for the previous period.

(2) Not a final figure, but a revised figure. Final figure is likely to change.

At the suggestion of John Huang, let's look at the above information graphically:

Looking at the above figures, does this look like a platform "in decay" as SPA put it?

Presented another way:

The SPA stated that MacOS sales declined each of the past seven quarters. Using their method (estimates vs. previous year actual), they are correct. Using a more rational method (current year actual vs. previous year actual OR current year estimate vs. previous year estimate), MacOS sales have continually increased. Interestingly, the major errors started appearing in the second quarter of 1995--about the time that Windows 95 appeared on the market.

5. Mac Market Share -- Quarterly Results

All numbers are in Millions of dollars and cover North American sales, only. All increase numbers are based on the same quarter in the previous year.

  Total Sales Initial Total Sales Final Mac Sales Initial Mac Sales Final Mac Market Share based on SPA (1) Mac Market Share based on Final Figures
 Q1 1995 2,030 2,140 307 335 15.1 % 15.7 %
 Q2 1995 1,570 2,070 226 376 14.4 % 18.2 %
 Q3 1995 2,200 2,900 242 375 11.0% 12.9 %
 Q4 1995 2,070 2,630 306 479 14.8 % 18.2 %
 Q1 1996 2,400 2,327 318 307 13.2 % 13.2 %
 Q2 1996 2,180 297   13.6 %
 Q3 1996 2,350   240   10.2 %
 Q4 1996 3,400 (2)   334   0.8 %

(1) SPA press release uses their initial estimate for the current quarter or year vs. the final estimate for the previous period.

(2) Includes extraordinarily high jump for Windows 32 bit sales--more than the rest of the year put together. Data is inconsistent with previous years' behavior. Look for a major revision.

The bottom line is that SPA figures are not very believable even if you don't consider things like bundles and hybrid CDs. When you consider those factors, they fall into the category of fantasy. It's also interesting that they never state error estimates. In fact, give that some of their figures are off by hundreds of millions of dollars (based on the revisions which occur), their publication of estimates to the nearest hundred thousand dollars is ludicrous.

I want to make it clear that I'm not accusing SPA of anything. SPA uses data provided by their vendors. However, I suggest that the SPA, if it is interested in being more accurate, should consider:

OK. Maybe the SPA is a victim here, too. But they are in a position to stop it. For them to re-establish any credibility in their ability to report sales figures, they need to take immediate action.

Want to do something about it? Contact:

Sources of information:

SPA posted a response in a public forum which I've reprinted (and replied to) here.

Special thanks to Arun Gupta, Matt Deatherage (GCSF, Inc.), Don Andrachuck, and Peter Benson

This page has been seen 16,651times since December 15, 1996.

Updated May 7, 1997.

Other Complete Macintosh Links:

Advocacy Information

* Advocacy for Mac Users * Macs vs. Windows * Performance *

Other Information

* Apple Computer * Mac Clones * Games *

* Switching from Windows to Macs *

* Magazines * Markets for Macs * News * Macs Around the World *

* Other Information * Problems * Power Mac and Performa Computers *

* Rhapsody * Software * SPA * What Apple Could do Better * Where to Buy *

 

* The Complete Macintosh Home Page *

* E-mail me *

This page has been seen 16,651 times since December 15, 1996.

Copyright Joseph M. Ragosta, 1997. All Rights Reserved.