They didn't so much as "miss" it as they "swerved to not catch it." Ken Olsen was truly perplexed as to why anyone would actually want a computer if they weren't a scientist or a large corporation, and DEC had the biggest market share of those markets. He affirmatively shot down ideas suggesting otherwise. It wasn't until the IBM PC proved him totally and absolutely wrong (one could argue the Apple II did but the thing that really got his attention was one of his competitors entering the market) that DEC came out with some variations on their machines for personal use. But because he didn't understand why people wanted them in the first place, it was really challenging for folks to put through really good projects.
They tried. Remember the DECMate and Rainbow and all those? Our school taught us MACRO-11, pre-standard C, RSX-11, and Pascal on them. They aimed at the small office and word processor crowd but were poorly marketed. They included such dirty tricks as putting glue in the peripheral card slots of cheaper models, to differentiate the expensive models.
They had a lot of software licensing issues. They believed software cost hundreds of thousands of dollars to develop and should be self funded (edited rather than a gateway drug to sell lots of new hardware). Also they believed in industrial era depreciation, so the cost of, say, OS-8, should be more or less constant for 20 years.
So although someone might be able to obtain obsolete / obsolescent hardware, if you wanted legal DEC (non-transferable) licenses it was always like $5000 for OS-8 and everything no matter how small had a price. Not a big deal for a corporation buying a $500K CNC milling machine, but it did quite adequately kill the small resale and reuse market.
They didn't understand software as it relates to the small market.
If my Dad could have legally run an old -8 or an old -11 at home or at work for clandestine-ish jobs, then in his official day job inevitably the office would fill with VAXes. But no, they wanted to nickel and dime everyone, so we ended up with with the inferior junk we are still recovering from. A world of VAXes and VMS would be a wonder. Home computers capable of running BSD on a -11 in the early 80s instead of rom msbasic. Could have been a different world.
About 20 years after DECs fatal decision to nickel and dime software, around '93 I was installing linux on commodity PC hardware using SLS floppy disk sets I downloaded for free, legally, from BBSes. So much for that bright financial idea.
As I recall, DEC didn't like the idea of people formatting their own floppy disks, either, and didn't supply a FORMAT utility, even with the version of MS-DOS shipped with the Rainbow. The idea was that you were supposed to buy preformatted disks from DEC instead. A major exception was that the Heathkit H-11 (an LSI-11/23 in Heathkit form) actually could format floppies under HT-11, its variant of RT-11.
So much potential, and they blew it all on petty BS like that.
It's easy enough to see the mistakes of yesteryear in hindsight but there is plenty of similar behaviour today.
Everybody that sells objects really yearns for a subscription model and everybody that sells objects through a subscription model would like to have a subscription based service model, and failing that they'd like some advertising money to go with their other offerings. And some companies want all of those at the same time.
Yes, I remember those. I (and a number of my friends) were waiting for the DEC PC, believing it would blow everyone else out of the water. We thought it would be an 11. I owned an H-11, which was an awesome machine.
When the Rainbow was unveiled, you could hear the jaws dropping on the floor. That moment was the end of DEC. We knew that x86 was the future at that point, and never looked back.
