“This site used to be better” is one of the Internet’s oldest recurring posts.
Sometimes it is true.
Sometimes the user was nineteen.
Those are not the same diagnosis.
Nostalgia is a terrible benchmark because both the service and the person changed. A platform may genuinely become more expensive or restrictive while also gaining useful features the old version never had.
The way out is measurement.
Dropbox gives us a change that nostalgia cannot explain away
Dropbox’s current documentation says free Basic accounts can be logged into on no more than three devices at a time.
The same help page contains the historical clue that matters: devices connected before March 2019 were grandfathered even when a Basic account already had more than three. Adding a new device requires disconnecting another or upgrading. See Dropbox’s device-limit documentation.
That is an objective product change.
A free user who synchronized five computers and phones before March 2019 could keep those existing links, but could not later replace or add devices without falling under the three-device ceiling.
You do not need to remember the old Dropbox homepage correctly to measure that.
Decline needs a defined dimension
A platform can improve in one way and decline in another.
Useful before-and-after measures include:
- subscription price,
- storage or device limits,
- number of ads per task,
- percentage of feed occupied by recommended content,
- API capabilities,
- export completeness,
- response time,
- outage frequency,
- moderation turnaround,
- number of steps required to perform a common action,
- features moved from free to paid access.
If one of those gets worse, say that got worse.
Do not automatically promote the finding into the entire platform is ruined.
Personal behavior can create a fake decline too
A social network may feel boring because your original friends stopped posting.
A game community may feel empty because you changed games.
A music service may appear repetitive because your listening narrowed.
A search engine may feel worse because the kinds of queries you make changed.
Those experiences are real, but they do not isolate a platform change.
A better investigation looks for old documentation, archived screenshots, price tables, terms, changelogs, support pages, independent measurements, or user-interface captures that establish a before-and-after difference.
Enshittification should be falsifiable
If every memory of a better Internet automatically counts as evidence, the theory becomes useless.
A serious claim should be able to survive a question like:
What exactly changed?
Dropbox Basic’s three-device limit is a good example because the company itself documents the March 2019 boundary.
Maybe a user still prefers modern Dropbox.
Maybe the service is faster, safer, or more capable in other respects.
None of that changes the narrower fact that one free-plan capability became more restricted.
That is how platform decline should be studied:
one measurable deterioration at a time, with nostalgia kept outside the lab.