Their methodology defines 'project' as github repo. That's probably quite strong as the vast majority of repo's carry no formalised initiative or organisation driving it, but are 'merely' places for tinkering of individual users.
If anything, I'm surprised about 10% are still active. If you compare it to the failure rate of formal (money-backed) initiatives of startups in general, or say restaurants, or hell, people starting a workout routine this time a year, it's not a particularly unique failure rate.
edit: they later differentiate between users/organisations. The latter has about double the activity rate (15%) after a year. But they define organisations using GitHub's definition, which can be had for as little as $300 a year. At that level the difference between user and organisation is probably limited to the difference between casual user and casual entrepreneur.
GitHub organizations are actually free (although, of course, there are paid plans for private repos and enterprise features): https://help.github.com/articles/creating-a-new-organization...
It's often a way that people group collections of related subprojects of an open source project, e.g. Babel is an org, with babel/babel being the primary repo: https://github.com/babel
People who make an org rather than just a project are probably more serious about it, which explains the higher activity rate, but it's still just "someone on the internet tried out an idea".
Our analysis found that only eight percent of projects are active, which we define as being updated at least once in the last six months.
By that measure it's impossible for a project to be finished, and must carry on being added to forever to not be dead.
I have a few repos I've not touched in a while that are apps/libraries/etc that I planned, coded, updated, and now exist as things I use quite regularly. They're not dead. They're done.
Normally this is a valid discussion to have, but I don't think it has any merit here. The projects analyzed were launched in 2016. You're suggesting that a statistically significant number of people completed a project centered around a concept as fast-moving and heavily researched as blockchain could be feature-complete and bug free in under a year and a half. I'd believe that maybe a dozen projects like that exist, but not nearly enough to affect these numbers.
The bigger issue here is how many of those 26k projects are dead-end forks for playing around.
"Playing around" in many cases will mean "learning the technology", and those developers would have then moved to other blockchain projects. That means the code is "dead" in the sense that it's finished, but it'd be unfair to suggest it's "abandoned".
As several have pointed out this is click bait, I carry forks on occasion and never carry out on doing anything with it.
Let's say that wasn't the case and people were more serious and less casual, a failure rate of around 90% is pretty much expected for a newer technology. New business start up failure rates are around that number.
I'm more amazed that 8% are -not- dead apparently. Either is true for most other projects, but you can make it newsworthy by adding the word "blockchain" in the title
This is just a faulty analysis to show-off "expertise" in the blockchain space by Deloitte. Measuring about of changes in a github repo has to be the worst kind of "analysis".
Gartner's Hype Cycle for 2017 predicted blockchain to reach its maturity within the next 5-10 years. That 92% of the projects based on blockchain are now dead means that practical applications of blockchain are being narrowed down to a few. But I'd say those few projects are going to get attention sooner rather than later.
If anything, I'm surprised about 10% are still active. If you compare it to the failure rate of formal (money-backed) initiatives of startups in general, or say restaurants, or hell, people starting a workout routine this time a year, it's not a particularly unique failure rate.
edit: they later differentiate between users/organisations. The latter has about double the activity rate (15%) after a year. But they define organisations using GitHub's definition, which can be had for as little as $300 a year. At that level the difference between user and organisation is probably limited to the difference between casual user and casual entrepreneur.