gregdoesit

Sure, but as a newsletter author, portability means that if this happens, I'll just move to e.g. Ghost or Beehiiv, or another platform.

The portability is the reason I am on the platform, and not on another platform. Take away this portability and I am gone.

The big difference with every other platform like Spotify, Medium, YouTube, Twitter, Instagram, LinkedIn etc etc is how those platforms do not allow for this kind of portability by design. They never have done so, and never will. And their take rate is also much higher than Substack's 10%.

Given full portability exists today, why are we speculating about things that are not happening, and what would likely undermine Substack's business model of both attracting and retaining newsletter authors?

And yes: Substack will need to make more money to justify their valuation. Taking away portability to do so would backfire: at least from my view.

> "Substack is to newsletters what Spotify is to podcasts, Medium was to blogs, and what Google Reader was to RSS: an aggressive player that dominates an entire segment with artificial and unsustainable advantages in a risky bet."

As someone who is writing a reasonably popular Substack newsletter - and have used several newsletter platforms in the past - I have to disagree.

The biggest difference between Substack and Spotify/Medium/Google Reader is how, with Substack - and with Ghost, Beehiv and other newsletter platforms - the author owns the relationship with readers.

When you sign up for Substack, assuming you use a custom domain (which is a one-off $50 fee), you own:

- Your domain

- The email list of your subscribers

- Your content that you can easily export and migrate

- Your subscriptions: as you are not paid by Substack, but subscribers pay you via Stripe, and then you pay Substack

This makes your Substack content, "audience" and paid subscriptions fully portable, should you ever want to move to or from this platform. These days, migrating to and from platforms like Ghost is fairly trivial without losing anything (content, subscribers or SEO ratings).

This portability and owning of my content, and the connection with subscribers is why I am on Substack. It means that I have the freedom to leave to another platform.

It's also why there have been many publications that have left Substack - and then some came back. For example, the newsletter Expontential View that has about 80,000 subscribers and thousands of paid subscribers left Substack for Ghost in June 2021 [1] and then migrated back to Substack from Ghost on August 2022 [2]. In both cases, they moved all their subscribers and content. And any publication can leave (or move to) without issues, any time!

Sure: this portability is not the best business to Substack. But it's the opposite of threat to newsletters: it's the opportunity to choose your platform, and move when you want to.

Show me another platform outside of newsletters that has this option.

[1] https://www.exponentialview.co/p/important-migration-from-su...

[2] https://www.exponentialview.co/p/announcement-move-2022

Books are not VAT free. On a €15 book I ended up getting charged €9 for customs: €1 for the reduced VAT for books €6.50 for the customs handling fee €1.50 for VAT on the customs handling fee

So it’s the customs handling fee being added that’s contributed to the price increase. Seems like the Netherlands started to enforce it for even small value orders from the UK.

I do wonder if Brexit and customs to be paid for all of the EU has anything to do with this decision.

Book Depsitory is based in the UK. I have inadvertently ordered from them, to the Netherlands, as they had a “Bookdepository NL” store, which made it seem like they are an EU entity.

However, when the book arrived, I was hit by unexpected customs charges that were half the price of the book. Looking at recent reviews, there was a surge of these complaints from all customers in the Netherlands [1]

Perhaps the timing on these customs surges are accidental, of course. But I can’t imagine that Brexit helped with their EU sales.

[1] https://twitter.com/gergelyorosz/status/1638824647390003206

sigh I am the author. I have zero stock or any financial exposure in Uber, Lyft, or any other company I mention: and in the rare cases that I do have (because I am an investor) then I disclose it. Here's the list of all companies I've invested (and I no longer invest in any, exactly to avoid biases) [1]. I don't get paid to write about companies. All of what I write is my analysis and opinion.

Here is my ethics policy [2], and the article starts by mentioning the above, as well as linking to my ethics policy.

[1] https://blog.pragmaticengineer.com/investing/

[2] https://blog.pragmaticengineer.com/ethics-statement/

I wrote the tweet. Didn’t realize it’s would be posted here.

I cannot edit tweets after it has replies (apparently by design). I added follow-up tweets, but seeing those are not read: I deleted the tweet.

Here are numbers in context, not looking at stock-based comp: https://blog.pragmaticengineer.com/

I wrote this tweet but didn’t realize it would get picked up. The numbers are not correct (or ant least misleading due to stock-based comp) and I can’t edit it.

I deleted the tweet and here are correct numbers: https://blog.pragmaticengineer.com/

For anyone wondering about the underlying reason for these events suddenly being cancelled: it is connected to Google laying off the team that organized these four events, as part of their January job cuts. I covered more details I could find a few weeks ago [1]

[1] https://blog.pragmaticengineer.com/google-coding-competition...

Yes - he is referring to me.

Hi! The G.O. here who you refer to.

To reply to some of the factually incorrect parts of this comment:

On why I don’t cover other companies’ RTO? But I do! You asked about Apple: wrote about them and RTO last month [1]. Other companies doing this - also [2]. You can find my articles on my blog [3] and I really try not to “obsess” about any one company.

I am a bit puzzled on the comment about writing about Uber in “such a generic way.” Here’s a deepdive that happens to be about Uber’s data center history, just from two weeks ago [4]. I am not aware of these pretty specific - and not that easy to track down - details shared before.

[1] https://blog.pragmaticengineer.com/the-scoop-apple-rto/

[2] https://blog.pragmaticengineer.com/the-scoop-rto/

[3] https://blog.pragmaticengineer.com/

[4] https://newsletter.pragmaticengineer.com/p/uber-move-to-clou...

Take Adyen, a B2B payment processor - valued at $45B, and a company that we can no longer ignore when talking about Stripe. Founded a few years before Stripe, IPO’d in 2018, processing a similar volume and growing faster in 2022 than Stripe.

In 2022: processed €767.5B. Net revenue €1.3B. EBITDA of €728.3M. EBIDTA margin of 55% - similar to that of eg Visa or MasterCard.

The difference? Cost basis.

Adyen employs about half the people as Stripe, mostly in Europe (their HQ is in Amsterdam). It’s not just fewer people, but, most likely, lower cost per employee (European HQ vs US HQ, and the wage difference between the two).

They offer lower rates, collect less net revenue, and have a far bigger profit than Stripe.

Source on Adyen’s numbers: their annual report https://www.adyen.com/press-and-media/adyen-publishes-h2-202...

In what is interesting, TikTok / ByteDance was on an absolute hiring spree for devs October to January for sure, while most other tech companies took the back seat or did cuts. Here’s data I got my hands on, comparing the number of “software engineer” positions at companies and ByteDance stands out unmissably:

https://twitter.com/gergelyorosz/status/1628831860611506178

Makes me wonder how they missed the memo that Meta got by October/November, Google and Amazon also by end of year: that the market is slowing.

I assumed Bytedance had just as good - if not better - data on the market than eg Meta, but Meta seems to have responded far far earlier to what looks like a stagnating market for as spend, than TikTok.

The plagiarised article in question was on the front page of Hacker News: https://news.ycombinator.com/item?id=34287747

Interesting enough, it’s author claimed on the HN that they used ChatGPT because they are not a native speaker but maintained that the ideas in the article were original. This article questions this claim.

Gayle confirmed this site is an impersonation of her and she did not author this. Seems like the post is made up.

https://twitter.com/gergelyorosz/status/1611748759553150985

So why delete my comment on the publication instead of addressing it? Then why delete my second comment? And why not make it clear the article is generated by ChatGPT, at least partially?

Also, this is a good example on why anonymous accounts writing content will be trusted less. You say you used ChatGPT for parts of it. Unclear on what “parts” mean and how much input you had, versus what the AI wrote.

It’s a reason for people to stop reading anonymous authors, or articles that don’t make it clear that it’s not an AI writing part of the article.

I’ll be honest: I feel duped reading a wall of text to realise it’s at least partially generated, and this whole article could have been the prompts you used to generate it.

The whole article feels to me like it’s generated by GPT-3 based on a few prompts. There’s a reference to Forbes article not linked, but otherwise zero things backing up this 1% claim.

The thing that makes this very suspicious is the continuous repetition of the same content, the anonymous writer and the fact that the only other article in this publication is about a tool that’s even better than GPT-3.

GPT-3 certainly has the effect that I have a hard time trusting that anonymous articles that are repetitive are not AI generated any more…

Update: I posted the exact same comment on the article and the author deleted it within two minutes. instead of responding. So yes, it’s likely I was on the money. I re-posted the comment. If it’s not there, you know that this comment is uncomfortable for the author for some reason.

Update 2: my second comment was removed within minutes as well. There’s a commenter claiming they are the author saying they used ChatGPT to generate parts of the article. Does not explain why they keep deleting my comment and not disclosing that this article is AI-generated.

Update 3: posted a third and final comment asking the author to not delete this comment and answer if the article was verenigde by ChatGPT. Comment also deleted within minutes.

This all underscores how it’s becoming hard to trust anonymous authors even now, and how this will just get worse.

If a GPT3 article can generate so much discussion on Hacker News, without most people realizing we are arguing about the output of an AI, GPT3 is ready to go mainstream.

Some insider details as someone who knew some of the Revue team - based mostly in the Netherlands.

The majority of the team was cut in the first wave of Twitter layoffs on 4 November (wrote about this [1]) - when 50% of people were let go. By this time, most of them were working on Twitter's long form tweets, which feature was being tested with customers.

So this decision was indirectly made on the first week of the Twitter takeover. In the end, it was perhaps two people left from the old Revue team. Today, there's only the founder of Revue, who wrote this article, everyone else let go AFAIK.

What was very, very surprising to me is how, a month and a half later, Elon Musk replied to a question on how he's open to the idea of acquiring Substack [2]. But he just shut down Revue, which was exactly "Twitter's Substack" version with all the same features: newsletters, support for paid, plus first-class integration into Twitter that Substack does not have.

[1] https://blog.pragmaticengineer.com/cruel-changes-at-twitter

[2] https://twitter.com/elonmusk/status/1607936197602562050

@treme most of my former team no longer works at Uber so I don't have direct contacts. Both of us can keep using the app, sharing an account. I am also not a fan of using my network when it's the system that is broken. I might be able to sort this by annoying people working at Uber I don't even know, but it's not a good use of my time, and even then it might not work (when I got similar escalations inside Uber, about a third of them went anywhere).

If it makes you feel any better: the same thing happened to my wife's account. Ironic that I used to work at Uber for ~4 years. Still, there's no way for me to resolve this for me unless I would escalate to someone working at Uber, bypassing customer support - which I'm not doing as it's a lot of work, a lot of annoying people I don't know and it might still not work. Customer support won't fix it, and there's no other way to resolve this.

It all goes down to either an ML system making a decision, or a flag was added to your account by some other if-then-else rule. When I was inside Uber, I could escalate tickets, internally, but this is no longer an option.

I wrote a longer summary on why such banning happens [1] and why - sadly - as a customer, we cannot do anything about it, until Big Tech prioritizes sorting out these edge cases, or there's regulation passed for them to do so.

And if you think Lyft is better: The New York Times ran an article Help! I Was Banned From Lyft and No One Will Tell Me Why [2] where they heavily quoted everything I wrote about why Uber - and Big Tech customer support, in general - works like they do. Lyft is pretty much the same.

[1] https://blog.pragmaticengineer.com/scaling-customer-support/

[2] https://www.nytimes.com/2022/11/17/travel/tripped-up-banned-...

Interesting enough, it was in August that I received a reachout for them, where they offered to pay me to write for Future.com and also keep my rights and re-publish later. I declined off the bat [1], because as I checked the site, the a16z association was pretty clear and I personally wanted to avoid being associated with the biggest crypto VC investors.

It sounds like those were the last few months when the publication was still running. I wonder if they simply struggled to find people to write?

[1] https://twitter.com/GergelyOrosz/status/1598666595274051586

Ah, got it. I still find it just weird practice to not link to the source, making it harder for people reading articles to find them.

Update: duplicate, the original WSJ article discussed here - https://news.ycombinator.com/item?id=33497012

The original article is from The Washington Post [1]. In this article, it is not linked, but mentioned. This dw.com article describes what The Washington Post reported.

I'm unsure why so many publications don't directly link the article that they effectively are re-wording in parts?

FWIW I also cover Meta every now and then in The Pragmatic Engineer. After reading The Washington Post article, I reached out to engineering leaders working at the company who basically confirmed what the Washington Post wrote. Layoffs are to be expected, likely this week. Also, some other signs pointed to this, like Meta having people return from business trips by the weekend (so by now everyone is in their primary work location), cutting them short if need be [2]

[1] https://www.wsj.com/articles/meta-is-preparing-to-notify-emp...

[2] https://twitter.com/GergelyOrosz/status/1589504531804876802

Keep in mind that inviting strangers to the site carries risks. After being an active member on the site for about a year - and liking it - I sent out my first ever invite for someone I didn’t know personally, but who seemed to be alright based on Twitter. That person was promptly banned for self-promotion [1], I apologized on behalf of this person I didn't know but invited [2] and was subsequently banned for having sent the invite [3]

Making it hard to get an invite, and not inviting strangers is by design, as I understand.

[1] https://lobste.rs/u/sebastienlorber

[2] https://twitter.com/GergelyOrosz/status/1295409271119765504

[3] https://lobste.rs/u/gregdoesit

I was having the same question.

Original situation: tickets are priced too low, and sell out almost immediately, with scrapers buying much of the tickets, then reselling them. Fans are furious! Plus: the band+seller is leaving money on the table, scrapers are making a killing.

Current situation: tickets are priced at market value thanks to dynamic pricing, so they no longer sell out instantly, and scrapers are discouraged from buying them. Fans are furious! Plus: the band+seller make more money than before, scrapers make close to none.

Taking a step back, it's worth understanding why there's so much clickbait. Turns out that if the business model of a publication is selling ads, they're interested in reach. Generating reach is easier by having large quantities of shorter, and easier to produce stuff. This is true for videos, articles.

So if you're looking for more thoughtful things - which take time to produce - your options are:

1. Pay for publications that produce these. Ad-supported publications are unlikely to be able to budget for in-depth content. Just look at how eg BuzzFeed shut down their investigative reporting (which was unusually good). It just made no business sense to produce those articles when a meme piece or two would generate more ad revenue, while being 100x cheaper to produce.

In the tech world, publications that fall into the “paid and in-depth” category can be likes of The Information, IEEE, MIT Technology Review, and many newsletters, tech publications etc. Look for ones where ads is not their main business model.

2. Another source are people who do this for free... because they have a main job, and it's not a business for them to share their thoughts on things. These will typically be blogs, YouTube channels and other places. Based on your interests, you should be able to find plenty. Also, see this Hacker News thread about interesting blogs [1]. The only real downside is you won’t get these on a schedule, as it’s not a job for these folks.

3. Books and podcasts. Books are straightforward enough: they're meant to be deep, and reviews help do some justice on them. Podcasts are usually based on ad-based models, but most ads are less intrusive, and the format lends itself for thoughtful commentary. It's more time-consuming to listen to them over reading, of course.

I collect RSS feeds of both my paid publications, and thoughtful blogs using a reader (I use Feedly) and find this works pretty well.

[1] https://news.ycombinator.com/item?id=27302195

The original Business Insider article - referenced by the article linked on this post - writes [1]

"Executives told directors across the company that they should select at least 15% of their teams to be labeled as 'needs support' in an internal review process, one of the people who spoke with Insider said."

So, as per Insider - and the article this post is about -, this information should be below the VP-level now, at director levels, and widespread in the organization as per Insider's reporting. I talked with people who supposedly should have been told about this 15% target. In engineering, I could not find any sources (including at director-level people).

We'll see soon enough if the Business Insider article is right, as the article claims:

"Several employees told Insider that as much as 15 per cent of the workforce could be cut within the next few weeks."

Let's check back mid-November to see if this did happen by then. If it would, there will be plenty of news on it.

I would not see the business sense. It does not mean that I am right. I offer my analysis, and I cannot predict what will happen, beyond assigning likelihoods to events. I assign a low likelihood of engineering layoffs happening at Meta the coming months. I see an even lower likelihood when considering that right now Business Insider reported about the performance review touchpoint happening - as previously planned - which tends to not be a trigger for letting people go, at least not in bulk.

[1] https://www.businessinsider.com/facebook-quiet-layoffs-will-...

When I did a deepdive on Facebook ~5 months ago ("Inside the Facebook Engineering Culture"), I confirmed with people inside the company the company having ~75,000 global employees, ~32,000 working in tech.

While I can't answer the question "what does the company need them all for", consider that the company's products are used by ~3.6B people every month, meaning they have about one full-time employee for every ~45,000 users.

Compare this with Google (one employee for every ~29,000 users) or Twitter (one employee for every ~52,000 users). The ballpark is similar, and shows the trend of it being harder to have fewer employees per thousand users (Twitter is much smaller than Meta, and Google twice the size in employee count while reached ~20% more monthly users).

The company generated ~$117B in revenue in 2021, which is ~$1.46M per employee. Profits per employee (net income divided by employee count) was ~$480K/employee.

There will, naturally be a cost of revenue, cost of customer support (even if not all of it in-house).

Also worth considering that Meta is composed like several companies: Facebook, Whatsapp, Instagram, Messenger, Reality Labs and several smaller/larger bets operating pretty independently.

I'd flip your question around: what if Meta had fewer employees, what would it mean? It would mean:

1. Increasing their profitability even further

2. Likely less support for current customers, and less of an ability to invest in forward-looking initiatives

I'm sure shareholders of Meta would be happy with #1. For #2, short-term shareholders would be happy to see all of this go. Long-term shareholders would likely realize there would be a cost to cutting down on customer support and R&D and I would not expect them to support this, given the very high profits Meta already has.

All in all, I find it remarkable that the company can operate with such revenue and profit numbers, and I suspect they are already deliberately hiring employees more conservatively, to keep up this profitability rate.

As a side comment, I have my thoughts on how non-applicable the linked article is for software engineers at Meta, as discussed in this comment [1] on this thread.

[1] https://news.ycombinator.com/item?id=33150495

I cover what's happening inside Facebook for software engineers, and I have confirmed that the 15% needs improvement target was not set as guideline for engineering teams, based on talking with several engineering managers within the company. However, this article is making its round amongst software engineers as well, creating lots of fear and uncertainty.

Right now, for engineering, what is happening is the revamped performance review process running its course. As with every year, the target for "needs improvement" is 10%, which is not new. Those going on "needs improvement", in the past, did not all go on PIPs. Also, PIPs at Facebook have not been nearly as draconian as at Amazon. With all these details, I find it a very strong stretch to say that "15% of Facebook's workforce may lose jobs."

Originally I wrote in a lot more detail about all of this - and why the Business Insider article likely doesn't apply to engineering - behind a paywall. Seeing this trending on Hacker News, which article I find misleading - an article referencing a public Blind post! - I un-paywalled my reflection, and my own analysis on the situation at Facebook. I personally doubt there is a need for layoffs at Meta, and I'm surprised no other publication points this out. Analyzing the financials gives a very different picture on whether or not any form of layoffs would be needed.

My response / reflection: https://blog.pragmaticengineer.com/meta-layoffs/

Updated with how shutting down Zenly means firing the whole team. The Zenly engineers always worked on Zenly, and not on Snap. If the goal was to acquihire, they are letting their whole acquisition go.

It smells like a complete change in direction. I am just unsure why not sell an asset that seems pretty valuable, or spin out, keeping ownership stake.

Buried in the announcement is how Snap is shutting down Zenly and firing everyone who worked on it in Paris, France. Zenly is a location-based social application (think: you can see & connect with your friends on a map). Snap bought them in 2017 for a reported $250-350M when they had ~4M installs. Now Zenly is at 40M MAUs, growing strong as I heard from employees working there.

Zenly is (was?) popular in Japan, Southeast Asia, Eastern Europe and the Nordics. In Japan, it is head-on-head as the most used social app with Line.

I find it odd that Snap just shuts down an app that was so popular in regions that is hard to gain foothold for any social media team. They could have likely sold it or spun it out, but chose not do so.

It's also not like the app is a major cost to the company. Less than 100 people work on it across Snap, most based in Paris, France.