So, first of all, you actually don't.

This is an aside to my post Research does not belong to Google.

In advocacy, there is a case for being nice to people who have wronged you. It's so that they will let you into the room. The Name Change Policy Working Group has been nice to a lot of recalcitrant bureaucrats and gotten to meet with them as a result.

Google, on the other hand, has never let trans authors into the room (aside from their own trans FTEs, who compromise like it's their job, because it is). They just close ranks and protect each other from consequences for the way they enforce transphobia in academia. They have a clearly defined position on the issue: "we're powerful and you don't matter".

They have done nothing to deserve niceness, or to even make it a good idea.

Even so, I'm not this critical of most Googlers! Just the full time employees.

If you're surrounded by people who call trans people by their deadnames, you're most likely in a hate group. But a possible alternate explanation is that you're in academia. And it's not because that many academics are openly transphobic -- they just don't know that the site they fully trust, Google Scholar, is telling them to do it.

Google Scholar was developed in 2004 and has changed very little since then. It supplanted a lot of hard-to-use library search indices by providing a Google-style interface with a single search box. Now it's the most name-recognized site for searching for almost any paper by almost anyone. One aspect of the design was, authors are just a kind of search term. An author is a cluster of different ways to abbreviate a name, like Firstname Lastname, Firstname M. Lastname, and F Lastname, and you might see different forms in different places, but the underlying name will never change.

This is because Google Scholar was built by, and for, cis men with unchanging Western-style names. The "almost anyone" who you can search for excludes trans people, among a lot of other people it represents poorly. And because Scholar will not change, it should perish.