My recollection of Prodigy was that it focused on a GUI when screen resolution and connection speeds weren't really up to it. It tried to be a more mass market commercial system but neither the technology infrastructure nor the user base were really there.

It's a tempting narrative to focus on Sears' failure because of their catalog. But, truth be told, the Sears catalog hadn't been a thing for the vast majority of Americans for decades. In the mid-nineties Sears was mostly an appliance and hardware store (with a bit of an identity crisis) at a time when the big box home improvement and discount stores were starting to get big.

It's somewhat fair to say that Sears could have been Amazon while taking advantage of their retail footprint. But I'm not sure that Sears was special in this regard compared to any other of the big retail chains (none of which were oriented toward the way Amazon entered the market with books and then digital media).

The company itself had a dumb bureaucratic culture. They didn’t take credit cards for years, etc. See: https://www.nytimes.com/1993/01/26/business/sears-eliminatin...

But they built a last mile infrastructure that was amazing for the time and didn’t exist until Amazon prime. They also had a small parts distribution capabilty for repairs.

To me it’s always been a interesting “what could have been”, but also an example of how much more efficient and effective a modern enterprise like Amazon is.

Their custom graphics were a good plan for making use of a 2400bps connection, but they screamed EGA. I suspect they had to develop a lot of GUI features from scratch.

The result was that the Prodigy package looked like running a DOS package in a window, while AOL looked and felt like a native package.

I also suspect it did them no benefits that you got an assigned and largely illegible email address by default-- you didn't see people clinging to their [email protected] addresses the way they kept their [email protected] ones.