The Price of Software

When Zork I was originally released in the early 80s, it sold for around $30-40 in 1980s money (according to a few online sources I forgot to save a link to). That’d be $118-157 in today’s dollars according to internet calculators.

Since it was a boxed-retail model, Infocom probably only made at most 50% of that, so effectively they made maybe $60-75 in today’s money (assuming less than 50%) per copy.

If there had also been an App Store or payment processor skimming 20% in the middle, they’d have made maybe $50-60 per retail copy in today’s dollars.

38,000 copies were sold in 1982 according to Wikipedia.

Let’s say they made $50/copy (which might still be a high estimate - I have no actual idea).

In theory, for 1982, it could have brought in nearly $2mil in today’s dollars assuming around $50/copy (roughly 80% of 50% of the box’s sales price).

I feel like if Iconfactory brought in that much or sold that many copies of any of our software in one year we’d be throwing a freakin’ party.

Obviously not all of that is profit, though! There were a lot of other things they had to produce in order to sell a copy of Zork - things like disks and boxes and manuals (remember those?).

I have no idea how much of the actual cost of the product would have gone into producing all of those parts, so this isn’t entirely comparable to today’s world of no-boxes and no-manuals, but let’s say they could produce the physical parts of the product for 50% of their real (unknown-to-me-but-guessed-at) wholesale price. That means they could have had revenue of 80% of 50% of 50% of the sales price - so maybe $28 per copy? (Fuzzy math here since I keep changing my estimates and rounding but you get the idea.)

Zork went on to sell hundreds of thousands of copies over time - although to be fair, I imagine the price dropped over that time, too.

Today, people expect constant updates and software is pretty much never finished. In the days of Zork, there were different editions and bug-fixes for each episode over the years, but people weren’t entitled to free updates every couple of weeks - if they wanted new bug fixes, they often had to buy a whole new box which meant a chance for more profit for Infocom each time, too.

Right now, Iconfactory Tapestry goes for $1.99/month or $19.99/year. With the App Store taking 20%, that’s more like $1.59/month or $15.99/year depending on which subscription level we’re talking about.

Looking at just the yearly number, it’s close to half of my wild guesstimate of Zork’s $28/copy.

The modern expectation for many apps is that they get updated continuously forever for that price, too, whereas in Zork’s day it was largely one-and-done. (And that’s ignoring the astronomical increase in software complexity since then.)

If we don’t update an app for a month or two, people seem to think it’s dead and they cancel subscriptions, don’t share it with friends, walk away and never come back - despite the app probably still working fine!

If they were subscribed monthly, we might not even get a full year worth of revenue from them. In the worst case, if they only paid for a month, we get a whopping $1.59 which isn’t even enough to buy a medium drink at McDonald’s!

None of this situation is exclusive to Tapestry - it affects all of our software and probably the whole industry (but maybe especially the indie software industry that typically doesn’t rely on selling services).

The drop in software prices over the years was fueled in part by cutting out the physical retail stuff - not filling boxes and shipping things really does lower the cost a lot - but the prices kept falling because you could “make it up in volume.”

For a short period of time that actually kind of worked - the lower prices and “at your fingertips” nature of online software meant selling more copies for less money to more people.

But, I think, not anymore.