In response to my Symantec article and other articles, some people mentioned numbers, and said that it didn't make sense to spend 50% of your development costs (for a Mac product), to gain 5-10% of the market. I would agree if that were the case -- the problem is that those numbers are not even close to fact -- but they are the fallacy.
Let's pretend the Mac hardware marketshare is as low as 5% of the total computers sold, it isn't, but I'll get to that later. That is still a silly measurement. It would be like comparing Boeing's market share to Piper's or to all vehicles. Remember, something like 75%+ of the game market (unit sales) is to things other than PC's (GameBoy's, Sega's, Nintendo's and so on) -- does that mean that the PC's are doomed because the market is too small percentage wise? Don't be silly.
Marketshare alone is a stupid way to measure potential market because people are not being wise about what they factor out. Why don't we compare PC sales to all programmable calculators, toasters with computer chips, cars with computer chips, game consoles, and so on? Because that measurement has no value -- people are using them for different things, they have different markets and so on. Same with Macs and most PC's. Apple doesn't go into many markets like Uganda or many markets that PC's are in -- comparing Macs against those markets is silly because few of those sales will effect Mac sales (or Mac software sales). Compare Macs only in countries that they participate in, or more accurately in YOUR country, since that is what you care about. In Canada I think Mac is closer to 25% and in the U.S. I think it is about 10%. But that still isn't an accurate measurement of anything.
If you wanted to get a more accurate measurement of System sales you need to factor out replacement parts -- many PC sales are reflective of all motherboard sales. PC's have a much shorter life span than the average Mac (about half), that means more of the sales are just replacements to begin with. Because motherboards are cheaper for PC's, and required for more upgrades, the numbers look larger than they are. So many PC's (most) are just replacements for other broken PC's, but they are counted as new sale! Many more PC's are being used for test-equipement, kiosk, some are used as servers (not running application software) than are Macs. Many PC are used for a thousand uses other than as a PC. All these things mean that a new PC sold is less likely to be used to run commercial software than a Mac (at least not normal productivity applications, games, educational software, web browsing, and so on). So those PC sales need to be factored out, but aren't. Then you have to remember things like over half the PC's out there still run either DOS on Win3.1 (which don't really count since new products don't support them). And then you have to fragment the remaining market between Win95 and WinNT -- not to mention Unix -- sometimes you can only support one. So the PC hardare market is not nearly as big or as relevant to software as people think.
Either the people making the measurements know this and are intentionally trying to make Apple/Mac look bad by not accurately representing things, or they are complete fools that can't be trusted with statistics. (I think that it is both).
Potential Sales Market
What should really matter to consumers is the sales market in their country, for the application software that they want to use. Like Graphics market (where Mac is probably 40% of the Market), or Education (60%+), publishing, general productivity, games and so on. Each market should be compared seperately, and more importantly does the computer have a reasonable marketshare in what you care about? I don't care that the Mac doesn't have a good elevator control and analysis package, as most people don't -- that should be factored out of sales as well. The real question is, are there enough machines in markets you care about to keep development thriving. Is there enough money in writing software for that platform (and the return on investment high enough) to keep software development going for your platform.
To get more realistic idea about sales, and what a Mac product would earn, we can either look at companies that sell Mac products, and see what their ratios are. This will give us a practical metric for measuring survivability for a platform. For many companies the Mac can approach 50% of their market (like Adobe and others). For some, usually the bad companies, it is as little as 20% (product to product). But conservatively lets say a Mac product will be 30-40% of total sales (1). Still, if you have to put in 50% more development costs to get that 35% return, then that is not a good ROI (Return On Investment) -- good thing it doesn't cost near that much.
(1) These numbers can be backed up quite easily by going the other way (look at the PC market). Remember that according to the SPA, PC software was like 4 or 5 times the size of the Mac -- and over $.60 of every software dollar in Windows market goes to Microsoft. If you are an ISV (independent software vendor) on the PC, your market is only 40% of what you think it is (the rest is going to Microsoft). Then figure there are 2 to 4 times as many competitors for every product. And you will be very lucky to create a PC software product that is much larger than the Mac version of that product.
So the reality, no matter how you cut it, is that the PC software market is likely about 2 to 4 times the Macs -- not the 9 to 10 times that people tell you. But let's figure out the costs in developing Apps, and see which is the better ROI.
Cost of development
To develop a product has many variables including:
Whether you are developing one product, or ten, for one platform, or all of them, these costs are fairly common. Let's look at what it would cost to make a Mac version of a product, assuming that the product was developed well in the first place.
Developing a Mac version of a product is not going to change 95% of your business costs. All these costs are pretty much fixed costs, you need them to run your company anyway. Cross platform development may raise your costs by a small percentage (say 5%), but it will not be a big deal. Of all the business costs, I would be surprised if 10% of infrastructure is actually in engineering -- usually the majority of real costs are in other areas (like sales, marketing, management and so on). In other words, adding a Mac team is likely only to add a very small percentage to your fixed costs.
Advertising (Sales and Marketing)
Advertising can be focused around either brand (corporate) recognition (with just some mention for each product), or it can be focused around each product. Usually there is a split. But if you are already trying to achieve brand recognition and have an advertising presence, the costs to add "available for MacOS" at the bottom of some ads is nominal (nonexistent). Excellent ROI for a Mac Application, since your costs don't go up, but your sales do. Furthermore, you expand your market since there are really 3 markets -- Macs, PC's and cross platform companies. Cross platform companies are much more likely to choose your PC products if there is a Mac version.
Of course to run ads in special magazines like Mac Magazines, does increase your costs, but it also increases your corporate exposure for ALL products, as well as the Mac. Plus these ads often cost less than the PC counterparts, and have more impact, since the Mac market is more hungry for software. Again, a better ROI.
There is also a big cost in paying distributors to put products on the shelves for PC's, and there is fewer mail-order and direct sales (which have lower overhead than store front sales) -- which all go to higher distribution costs on the PC side. Again, you are getting far bigger value for the buck on the Mac side.
Because there are more PC Apps, to make an impact in the PC arena it costs more and has less effect. Advertising in the Mac arena has more effect (since there are fewer competitors), and you can also have an impact on cross platform users. So ad money spent on the Mac side is far more effective than on the PC side. But these are a small part of your operational costs, as compared to the rest of infrastructure costs.
There are many parts of developing a product. People don't understand the real costs.
The heaviest costs are in the design of the product, both architecturally and in interface. Contrary to what many realize, probably over 50% of total development costs go here. Now I consider this to include documentation, user evaluation, and lots of other nitty gritty stuff that is required to make a product. Guess what, 100% (or 95%) of this can be leveraged cross-platform. So the costs of an additional Mac version of a product may be as high as another 5% of the design costs, which is half of development costs.
But the reality is really better than that for supporting Macs. PC users often think only PC'esk and ugly proprietary Microsoft-Windows like ways. Mac designers are often better in that they think cross platform, and often understand UI issues better -- so have a Mac design team can actually reduce the total design costs (or the costs of a bad design). Thinking Mac is better than thinking PC. Thinking cross platform is better than thinking just PC (Windows). The end result is that a cross platform product is almost always better than a Windows only version -- in design, sales, interface, and so on.
Let's say development (implementation) is actually half of what is left of engineering (or 25% of the total development costs). On a well designed package you can easily get 50% shared code (across platforms), and closer to 75% common code if you know what you are doing. So 50% of your implementation effort (conservatively) comes to any other platforms for free. Then, on top of that, Mac programmers are usually 2 times more productive than PC programmers. It isn't that they are that much better, it is that the platform and environments are that much better. Usually there is at least a 2:1 ratio of PC programmers to Mac programmers to achieve feature parity in all the companies I've every worked at or dealt with. (That isn't counting the cross platform or "engine" teams). I'm not kidding, and I've talked with people all over the industry. So what are the additional development costs for a Mac version? About 1/4th (25%) of the development costs -- and of course the development (implementation) costs are really only about 1/4th of the total costs.
Testing probably costs half of what was left (12.5% of development). Everyone in QA knows that the amount of problems on the Mac are far far less than the PC (unless you are on Novell networks). Just the number of platforms you have to test on alone is far smaller, not to mention the 10,000 cards and configurations, and OS's (Win3.1, WinNT, Win95, Win98 that all behave substantially differently). But lets be ultraconservative and say the Mac is only 25% as much work.
But wait, it should be less, since many bugs are cross platform (and so fixing one helps the other). In fact, many times the Macs memory model is a great way to uncover leaks and other problems on the PC version of the product -- this means a more reliable PC product by doing a Mac version, that means a better reputation in the industry, and so on. So the Mac development ADDS value, and costs very little.
Lets say that support is the remaining 12.5% of the development budget. Anyone who has worked a support line knows this one -- if the Macs are 1 out of 10 calls then you would have some serious Mac problems (i.e. that would be high). But we'll play it save and say the Mac is 30% of the support calls and cost. (On what Planet I'm not sure, but let's pretend).
Total Development costs for adding a Mac version
If we are looking at the real totals for a company to make a Mac product, the development costs would be roughly as follows:
The Mac development is 15% - 20% more than having a Windows only product. And I am being pretty generous to the PC (against the Mac). A company that knows what it is doing in cross platform can get these costs down further. The Mac version (or cross platform development) added a ton of extra value, in fewer bugs, better design and interface, and so more.
The Mac side has easily the potential of being 40 - 50% of a products revenues, for 20 - 25% more costs. That is a 2:1 better ROI than investing money in more PC products. That isn't including the advantages of a more reliable product, the expanded market (because of cross platform sales), or the product recognition crossover (added sales on the original platform because of gained cross-platform customers), and so on. So the issue isn't whether it makes sense to develop for the Mac because it makes twice as much sense as developing for the PC alone.
But wait, there's more
Remember that the Mac costs less to develop for, and that the software market is less saturated. That means that if you are a company, it makes far more sense to release your product FIRST on the Mac. It will cost you less to create and support, and you are likely to get more attention for the product. Then, once you've built a name for yourself and the product, you can bring that product over to the PC and stand more of a chance of gaining market share and attention. Where if you release PC first, it is far easier to get lost in the crowd.
If you doubt this strategy, then you only need look at the big PC software companies. Microsoft, Adobe, Symantec, Quark, Aldus, all got their starts on the Mac for a reason. Most PC companies that tried to come the other way, failed. That failure also showed the fundamental problems in those companies, and with their products. Novell couldn't hack it on the Mac. Lotus. Ashton-Tate. Where are those companies now? Symantec started going away from the Mac, and now they are struggling in general. So it seems that good support of the Mac is an indicator of the company at large. If companies can't hack doing cross platform development, then they usually just can't hack the industry at all, and they often fizzle out.
Every time some so called "analyst" prattles off about marketshare, I have to laugh. These simians are prattling off numbers that are irrelevant, and are either too stupid to realize it, or they are intentionally defrauding the public. Neither of which impresses me with them.
When companies start dropping support for the Mac, it is usually an indicator of far bigger problems than just Apple or the Mac platform -- usually I see a company starting to go into a tailspin and having other issues. I think of Novell, Electronic Arts, Lotus, Symatec, and all the companies that coincidentally dropped Mac support and had problems either right before, or right after that move. Often dropping support was just an internal political decision or a turf war -- but it is rarely a good business decision. Some companies can survive their stupidity anyway -- like Autodesk -- but I wouldn't call them thriving or innovative companies. Usually they are just struggling to retain their marketshare and their present, and are doing little for their future. Usually, those moves are the beginning of the end, or at least the beginning of stagnation.
When a writer, an analysts, or the public at large babbles that Mac development costs too much (and so the Mac is going to die), I have to laugh. I believe it was Lucas Arts that said that the PC versions of their products pay for the development -- but the Mac versions are all profit! Usually the Mac version is at least twice as effective a way to spend dollars than on the PC alone. Microsoft, Adobe, and many of the big boys know this. Many small developers are Mac only, because they know this as well. Only people that don't know my industry and the realities of cross platform development don't know this. But now you can bludgeon them with this article to try to beat a little common sense into them. Good luck.