Publons

» see publons.com as it lived, on the wayback machine

~ the obituary ~

I work in industry. Edge AI and computer vision for unmanned aircraft, in Ankara, which means the output of a research field arrives on my desk as papers to read rather than as papers to referee. Publons was built for the people standing on the other side of that door, and I only ever stood on this one. What made me stop and look it up was the problem it named, because the problem is real and it is not confined to academia.

Here it is. Every paper that gets published has been read carefully by two or three specialists who were not paid and were never named. Afterwards they could not prove they had done it. The reviewing is a load bearing part of how science works and it sits entirely outside every system that measures a career. A department head counts your publications. Nothing counts the forty papers you took apart for somebody else's journal.

Andrew Preston and Daniel Johnston started Publons in 2012 to put a number on it. The company was registered in New Zealand with an office in London. The name came from the word publon, a joking term for the smallest unit of publishable material, which tells you the founders came out of the culture rather than into it from outside.

The mechanism was simple enough. You told Publons what you had reviewed and the publisher confirmed it, without anybody breaking the anonymity of the review itself. What came out the other end was a verified count on a public profile, a line on a CV where there had never been one.

~ the case file ~

~ what it counted ~

By the end more than three million researchers had accounts and over a million reviews had been logged against roughly twenty five thousand journals. The business model was publishers rather than researchers, which is why the profiles stayed free. Journals paid for reviewer discovery and for the review data, since a company that knows who reviews what also knows who to ask next.

There was a training arm too, the Publons Academy, teaching early career researchers how to review properly. That part mattered more than it sounds. Peer review is a skill almost nobody is formally taught, and most people learn it by being handed a manuscript and guessing.

~ the acquisition ~

Clarivate Analytics bought Publons in 2017. That is the company that owns Web of Science and the Journal Impact Factor, which is to say the company that already owned the scoreboard. Buying the one independent system that measured a different kind of academic labor was a sensible move for them and a warning sign for everyone else.

For five years it was left mostly intact and even grew. ResearcherID, Clarivate's older author identifier, was folded into Publons in 2019. By mid 2022 Clarivate was describing the platform as serving 2.5 million researchers with close to five hundred publisher partners.

~ the day it died ~

On 18 August 2022 publons.com began redirecting to Web of Science. Profiles were not deleted. Every one of them became a Web of Science Researcher Profile and the review counts came across intact. In pure data terms nothing was lost.

What went was the separate address and the separate identity. A reviewer's public page had been a thing you could link to that belonged to a project about reviewing. Afterwards it was a tab inside a subscription product sold to universities, and the difference is not technical.

~ what it leaves behind ~

The idea won. Recognized peer review is now a normal thing to list on an academic CV, and ORCID carries review records for anyone who wants them there. Publons made that expectation exist and then handed it to the incumbent whose scoreboard it had set out to sit beside.

The reviewing is still unpaid. That was never the part Publons claimed it could fix.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »