My general impression is that a lot of companies are more than 4x overstaffed.

Early Google had small, elite teams, whom it treated very well. That works really well. Late Google kept elite team salaries with bloated, mediocre devs. That works less well.

His infrastructure hasn't fallen over.

But by his own admission the site has lost a huge majority of its value, and the main revenue source has dried up.

> His infrastructure hasn't fallen over.

...yet. An airplane can still glide a great distance if you turn off the engine that actually powers it. I feel like Twitter/X is still at the "Wile-E-Coyote ran off the edge of the cliff and is still running in midair" part. The technical bits are probably running on pure momentum.

I doubt that the decline in advertising revenue has much to do with the staff cuts. Advertisers were driven away by bad PR, changes in content moderation policies, and the extension of the Community Notes feature to ads. Mainstream advertisers are rightly paranoid about anything that would harm their brand images.

One of the big units that he cut was to the trust and safety team, including content moderation and hate speech.

A big part of why he lost so much advertising revenue is advertisers knowing that there isn't the staff necessary to ensure that their accounts are being handled properly, and the content they appear next to is moderated.

A fair conclusion.

SREs predicated Twitter would crumble when Twitter's SRE team was laid off. It had some hiccups, but seems stable now.

I suppose the next line of defence is that it takes a while for architecture to crumble.

Time will tell whether that's accurate, or a No True Scotsman.

My anecdotal impression is the pace at rolling out changes and features seems to have also increased.

It sounds like the SREs did a good job of making resilient services

Takeaway is building resilient services is bad for job security lol.

Disclaimer: I've worked in SRE.

My position is that early SRE solved the hard parts, before I was involved. Services and platforms became reliable. The field is mature. SREs still point to risks, but execs have called their bluff.

The takeaway is to make sure you are actually creating value. Maybe Twitter's SREs were not actually improving reliability much.

They also vastly scaled back the number of services operating though. This whole thing of saying its stable with x << 100 % of staff is kind of nonsense when it its users, revenue, and features are also << what it used to be.

I don't use xitter so forgive my ignorance. What services did they scale back?

>SREs predicated Twitter would crumble when Twitter's SRE team was laid off. It had some hiccups, but seems stable now.

I think SREs knew better than to claim such. I saw tons of non-SREs claim that Twitter would surely immediately collapse. Then the 2022 FIFA World Cup occurred immediately after the mass layoffs ... and Twitter kept on working.

An Ebay outage would be more costly than a Twitter outage.

Depends on the makeup of those roles. Often it's not core engineering or infrastructure that's cut, but things like Innovation Evangelists, dev rel, swag diversity procurement specialists, TPS Report Format Change Control specialists, etc.

The floodgates were already wide open by the time Elon showed up at Twitter (in Nov 2022). Mass layoffs in tech started sometime in the fall. Microsoft laid off thousands in Oct 2022. Stripe did a similar number in early November. Facebook had the largest layoff in company history (11,000 people) the week after Twitter. Salesforce was the same week. Amazon laid off 10,000 a week after that.

So it wasn't really Twitter setting the trend. The company was simply reacting to the same market conditions as everyone else.