Twitter limits unverified accounts to reading 600 posts/day | Hacker News Reader

Twitter limits unverified accounts to reading 600 posts/day

twitter.com

157 pointsmadars33 comments

Comments moved to https://news.ycombinator.com/item?id=36552324. Normally we wouldn't do that, since this has the original source and was posted earlier, but it seems to be paywalled without a workaround and that's where we draw the line (see https://news.ycombinator.com/newsfaq.html and https://news.ycombinator.com/item?id=10178989).

FWIW, both are exactly the same in terms of work-arounds: one had the tweet copied in the comments (just like archive.is links customary for newspapers), other in text/screenshot.

This seems like death for an ad supported model on Twitter. Why would you buy ads on a social network that deliberately restricts the number of posts (and thus ads) that a user can view?

My (cynical) guess would be that once you pass 600 posts, all you see is ads...

You don't have to guess. It's already happening. There's a rate limit error, and no tweets.

Even before 600 given "filter:follows" search used to provide a more accurate timeline than the Following tab, and is now half full of random tweets that have nothing to do with the accounts I follow.

Can't speak for the big advertisers, but as a small business owner, I'm more likely to spend ad money on platforms that take an active stance against bots. In general, advertisers (should) care about impressions by unique users, not total number of views.

I'm not sure about requiring an account to view contents though. That to me will just kill total usage across the board.

The problem is this will reduce impressions by unique users as well. Once a user hits the daily tweet limit (easy to do if you scroll and view a few viral threads), they can't view any more ads for the day. A few viral or high engagement threads could significantly reduce your audience for 24 hours.

Additionally, if Twitter CTR is low, then you need a _lot_ of unique users to actually generate conversions. Low CTR + Low impressions reduces the value of the ad.

> if Twitter CTR is low, then you need a _lot_ of unique users to actually generate conversions

You don't want to on a platform that has loads of unique "users" and impressions if most of those users are bots, though.

Perhaps quality is actually better than quantity?

Pre-Elon the bot count was something like 5% of active users, meaning a broad approach like this will affect significantly more real users than bots. There are probably more fine tuned ways to reduce bots that don’t degrade the experience for your biggest users.

> Pre-Elon the bot count was something like 5% of active users

Is there a reliable source for that figure? Also, are we talking about 5% of accounts (active accounts/total accounts?) or 5% of pageviews or some other metric?

I'm struggling to see any incentive for platforms to be honest about how many bots they get.

Twitter reported in their financial statements that less than 5% of their monetizable daily active users were spam or fake accounts. They did disclose some errors in methodology in 2022, but not enough to make a substantial difference to that estimate. They shared their methodology for estimating with Musk’s acquisition team and with the SEC.

One can claim that Twitter was lying about this number. But if he could prove they were lying, Musk would have gotten out of the deal.

> Twitter reported in their financial statements that less than 5% of their monetizable daily active users were spam or fake accounts

If a human signs up for an account, and uses it occasionally, but there's ongoing automated scraping or posting using that account for a good chunk of the time - or more specifically "for the majority of the requests" - would we call that account spam or fake?

That sounds like a huge speculation/hypothetical — is there evidence that was or is happening?

To answer your question, it would depend on what the account was doing. Scraping abuse could likely be handled in a more targeted manner.

> That sounds like a huge speculation/hypothetical — is there evidence that was or is happening?

I've signed up for websites before, manually, and switched to using automated tools to download all the content. Certainly more efficient than downloading stuff by hand. Isn't automation the kind of thing most developers would do in this scenario?

If you're doing it right, the website doesn't even know. At least not until it's too late ;)

Aside: a friend of mine wrote a script to perform the online check-in whenever he flew with his favourite airline. He reverse engineered the airline's official app, and duly wrote his own (Perl!) script to do the same. It spat out a pkpass file which he could load straight onto his phone.

I agree to an extent.

The only metric that matters, and unfortunately the hardest one to measure, is conversion rate. Everything else - impressions, clicks, watch time - are just really bad approximations of real value.

_If_ Musk can establish Twitter as a high-conversion platform, he will succeed; however, that's a big "if" because I think this latest decision is a huge miss. I was on twitter for 15-20 minutes today and got rate limited. I'm not a bot.

Are enough people going to give Twitter $8 because of this to offset those who stop being active on the platform and driving ad views? I highly doubt it. I know I'm just moving on to doing other things instead.

Yeah, but twitter has more bots then ever. If you want platform without bots, twitter is loosing that fight on pretty much all fronts.

But is that not exactly how Instagram works already?

Instagram requires a login to view most posts, but it doesn’t rate limit it’s most active users.

It absolutely does rate limit logged in users. try scrolling any view (tag, explore) - you will hit a wall with a cooldown in hours to days. Not a rate limit, a full-on "not available"

One reason I could think of is that real humans only read a certain number of tweets per day. Showing ads to users that "read" large numbers is showing ads to robots which, sadly, do not convert very well.

Current rate limit is about 20 minutes of use for me. Remember that it loads a couple dozen replies at a time if you click on a tweet, so 600 tweets is really not much.

Yeah, this basically restricts users to viewing a handful of popular threads per day.

600 posts is about 10 minutes of browsing for me. It seems to be counting replies in large threads to the limit. So even if you make a post with a premium account your content won't be read by the vast majority of your followers.

It's like how scams put in typos on purpose to save effort. If I was an advertiser I'd really want to target people who donate $8 a month to Twitter.

Because the value of accounts that browse extreme numbers of posts per day is near zero to advertisers.

Someone who has seen thousands of ads per day for years on end is basically immune to ads (or has blocked them, or is a bot)

One of the hard things to grok as an engineer or power user is that we are not the target audience for ad-supported services most of the time. How many people in this thread have aggressively blocked ads everywhere and are confused about why ad-supported services aren’t catering to people like themselves?

A few thoughts:

1. I'm not sure that an account that heavily browses twitter is zero value to advertisers. If anything, those are more engaged users with Twitter, _and_ Twitter should have more data and thus better targeting for that user.

2. The marginal efficiency of not showing an ad to a saturated user is probably offset by the reduced potential audience for your ad. Once you've seen 600 tweets, you can't use Twitter or view any more ads. That can shrink the daily audience significantly.

3. Users now need an account to view _any_ tweets. This further reduces the reach/audience for Twitter ads. If Twitter ads already have low CTR or referrals, and now their potential reach is cut, then they have no value.

I have blocked ads most places, but I use Twitter in ad-supported mode + receive ads through the app. I'm a good example of a user they'd be advertising to, and now I can't use it.

I assume they're hoping that people will start paying $8 for that stupid blue check.

Never gonna happen.

Probably a good thing. People should use less social media.

Elon paid $40 billion to cure Twitter addiction. It worked on me.

I tried to hit the 600 post limit by scrolling through my feed quickly, after about 2.5 minutes the tweets I was being shown were mostly the same ones but on a loop. I kept scrolling for another 3 minutes and yeah, it's like the last 100 odd tweets shown over and over and over.

Which honestly is an ironic metaphor for the platform as a whole.

https://docs.google.com/document/d/1l8dzqvpLyA-l0955IlBRLf5S...

Sounds like some folks were finding ways around the new API rules/pricing and Elon decided to put an end to that with a machete instead of a scalpel.

Reply on news.ycombinator.com