Experienced people in this forums seem to act and think under the assumption that the number of potential customers for a game is infinite.

It looks like one can do experiments and the worse that can happen is that the sales will be lower for a while, the assumption is that once you undo the changes, the sitiuation will be the same as it was before - fresh customers will just keep coming. A popular variable is downloads per month, but nobody ever said that you can run out of customers (when everybody already has the game).

So... I'm sure this is is because my naive point of view about the market, but is this assumption valid? Are there really that much customers? I never thought about this, but at 10K downloads per month, 5 years mean 600K customers. What is the size of the market? It seems to be much, much, much, much bigger than that, but how much?

The question is not exactly valid.

The market size is different for each game. More importantly to you, it is different in each genre.

Smart marketing and good product positioning, a game can get a very very high quantity of downloads.

If Super Mario 3 could sell 17 million copies you can realize the POTENTIAL market size for any game downloadable on the internet is some number much higher than 17 million. (so, yes, much higher than 600,000)

A marketing test is most important by its ability to increase the 'long run' effect of the sales. Spend a few months finding the 'best' way to market it and stick with it, making small modifications so long as it reaches your goal in sales. When/if it falls below your goal level, begin making changes.

This all assumes you are producing a mass market product. If you have produced a game you would consider a niche product, your marketing attempts should be targeting that niche, and most changes within the game or the way the game is recieved wont make any difference in sales %s.

Our product is browser based, the true market size therefore is somewhere in the area of 1 billion potential customers. The actual market size is somewhere around 500,000.

Terin

I think that you must also consider that the market varies based on game-type.
For my soccer game I believe that the market is still quite big, but not as if I made a small cute tiny puzzle game (Dweep? ;)).
And also there's the "new" effect... when you release a game you get lot of downloads: but after the intial boost, your download count start to decrease.
My download now are about 5-10% less than in past 30 days (when I released the game)...