Hacker News Reader

Comment by phlhr | Hacker News Reader

Are people purposely ignoring the fact that Twitter are clearly lying about the % of non-human users? They say its less then 5% but in reality its between 5% and 25% based on multiple independent research studies.

If it really was < 5% why don't Twitter just release the data Musk is asking for?

View on HN

This would be a great point if Musk hadn't waved due diligence. He knew this was a risk before he made the offer and now he's just posturing.

I think he didn't wave due dillegence, but his due dilligence was based on public filings that Twitter made, which may prove to be wrong.

I'm curious about the nuances of this, I haven't seen this argument made elsewhere yet, but it does seem to make sense that you would make that kind of exception. Even when you buy a house as-is, exceptions are generally going to be made for existential threats (leans on the house, title issues, etc.) outside of the _condition_ of the house. But it still feels like posturing, especially since he knew this was an issue before he made the offer, why not explicitly call this out in the contract? It definitely feels like the weight of the economy collapsing is the substantive issue.

No, the SEC agreement directly addresses this. Section 5.11 ("Parent" and "Acquisition Sub" is Musk, "Company" is Twitter):

> Each of Parent and Acquisition Sub has conducted, to its satisfaction, its own independent investigation, review and analysis of the business, results of operations, prospects, condition (financial or otherwise) or assets of the Company and its Subsidiaries. In making its determination to proceed with the transactions contemplated by this Agreement, including the Merger, each of Parent and Acquisition Sub has relied solely on the results of its own independent review and analysis and the covenants, representations and warranties of the Company contained in this Agreement

So, in essence:

1) Musk has been afforded the opportunity to address any concerns he has with Twitter

2) Any concerns of Musks have been satisfactorily resolved by Twitter

3) In determining said satisfaction, Musk is relying on his own judgement and analysis, and is not relying on any analysis by Twitter.

There is absolutely 0 ground to his complaints today and claiming he "isnt being given the information hes been asking for" or that "I was going off your numbers but I want to run my own".

Not only that, his biggest argument for buying Twitter was literally that there were too many bots and that he would solve it. Now that he's waived due-diligence and the market is tanking, he's "concerned" about the number of bots? It's such transparent BS that it's embarrassing.

I don't have a horse in this race, but I don't follow your reasoning. If you thought there was a large bot problem, part of the appeal would be that there's an opportunity to fix that problem. That would increase the value of the company after you've fixed the problem. Assuming that's the case, then finding out there's not a big bot problem would mean there's not as much of an opportunity.

If that's the case, then there is missing information that you should have done due diligence for?

yeah, maybe he could have spent a little time looking into this stuff before committing. Oh well, I'm sure his lawyers are happy to take the billable hours!

Eh, should've done my homework before commenting. Looks like Twitter has been saying 5% since well before Musk made the deal.

It comes down to not being able to state that he was unaware of a bot problem (as an argument for walking out of the deal) if he has very publicly said that he's doing this because of and to fix the bot problem.

> Assuming that's the case, then finding out there's not a big bot problem would mean there's not as much of an opportunity.

But that's not what he's saying, he's saying "the bot problem is way worse than you're saying, therefore the $54.20 price is too high."

It is a little more nuanced. Twitter is claiming than fewer than 5% of the monthly active users they report are spam bot, not that fewer than 5% or those on the platform. If they “don’t count” them but also don’t ban them, that would leave room for bit theories to be “true”

There's also a deeper set of definitions at play. Twitter the corporation cares about monetizable active users, as stated. Or, in English, "what fraction of ads served are really going to robots?"

Twitter users don't give a fig about Twitter's ad impressions. They care about what fraction of comments or interactions are in fact with bots or astroturfers. From the perspective of Twitter as a communications medium rather than an advertising channel, this is the relevant statistic.

it’s a nonsense argument, most people (including musk) couldn’t even articulate robust criteria for measuring a “non-human” user — musk himself has undermined his position on this topic multiple times.

Twitter has never claimed that the number is accurate nor important in absolute terms (“we aren’t sure” features prominently in their filings!) rather it’s a number that is important for understanding growth and evolution of the platform. The number matters quarter-to-quarter, not in isolation.

Active human users matters in the context of advertising, it’s a pointless distraction that musk is employing to back out of the deal.

Let’s imagine there is a real measure of “non-human users” and lets say it turns out that Twitter underestimated by 50%… so? Musk long said he wasn’t buying Twitter as a financial move, and that he has a plan to be wildly profitable off of a small proportion of twitter’s users so unless 95% of twitter’s users aren’t real, it doesn’t impact his (absurd and ridiculous) plan.

It matters because Twitter have misled advertisers as to the quality of their audience.

If Twitter did come out and say they underestimated non human activity by 50% or the like then yes Musk should be forced to continue the sale.

Twitter haven't though. They are still pretending that less than 5% of daily active users are non-monetizable.

That’s not how advertising works, advertisers don’t care about “quality of the audience” in the abstract because digital advertising is results driven. There’s a great deal of fraud in advertising across the industry, and it’s certainly not good, but it’s a cost of doing business for advertisers and is factored in: when an advertiser pays Twitter $100 to generate $150 in revenue, it doesn’t matter if $50 of that $100 was spent because it “non-human users”. Twitter could reveal tomorrow “oops our numbers are wrong, our non-human user count is actually twice as high” and it would have zero impact on advertisers. Not a single professional in the space would care, because advertisers are not measuring their value from Twitter based on twitters KPIs. Advertising is a means to an end, advertisers measure the end.

That's completely false. I've worked in both the media buying and inventory side of advertising and the large accounts definitely care about the quality of audience. Nike, Apple, Nestle etc have huge internal agencies that are tasked with finding brand safe platforms to advertise against.

The type of attribution based advertising you are talking about certainly also exists whereby the advertiser pays $x per (milli)impression, then a further $x for click-thru and then a final $x for a conversion. However the vast majority of twitter's revenue is in the first bucket (CPM) which is entirely valued based on the size and quality of the audience.

You’re drawing an arbitrary line. The majority of digital advertisers are using digital advertising because the value is quantifiable through attributable revenue.

There are certainly a minority of major brands that operate as advertisers with awareness campaigns — sponsored hashtags are a good example — but if you read the Twitter filings, it’s very clear that their focus is performance based advertising — and that’s where they see their future, too.

Maybe a decade ago you could have said that advertisers were just trusting platforms to deliver value, and that ad-fraud could make or break a platform if they got caught, but that’s not true anymore, it’s a much more sophisticated market. Advertisers aren’t (as) dumb (as they once were).

Brand safety is a whole other kettle of fish — that’s a concern across all types of advertising, and not relevant to the audience, rather the content of the platform.

>You’re drawing an arbitrary line. The majority of digital advertisers are using digital advertising because the value is quantifiable through attributable revenue.

If you are talking about the majority of ad impressions being programmatic/attribution based then yes you are correct, however if you talking about dollars spent that direct sponsorship with large internal agencies is still very much king.

To give some context, large corps would routinely drop $5 million on a direct deal with Twitter for a combination of promoted tweets, hashtags, trending etc. This was also almost pure margin as there was no middle DSP/SSP taking a cut. To get the same profit from the method you are purporting to be most common would take years.

>Maybe a decade ago you could have said that advertisers were just trusting platforms to deliver value, and that ad-fraud could make or break a platform if they got caught, but that’s not true anymore, it’s a much more sophisticated market. Advertisers aren’t (as) dumb (as they once were).

Maybe a decade ago? So the people who are now in senior positions at the agency and call all the shots are the ones making the major deals. Well then it would stand to reason that the biggest profit comes from deals that are structured like they were 10 years ago.

It would be interesting to know exactly what the definition is for a bot / fake account.

For instance, I have 5 accounts which automatically tweet out the latest posts from 5 of my websites. They are automated and probably technically bots, but that seems different than an account that is meant to give fake likes/retweets.

They are never in an active "logged in through browser/app" stage because they use the api.

The comment was 5% of active users never receive ads. Some real people turn javascript off as well and some user ad blockers.

Nowhere Twitter claims that. For years they don’t report active users, but their own metric of monetizable active users. Elon just pretends he doesn’t know that.

“Active users” does not mean “users who tweet”

Majority of Twitter users don’t tweet, they’re still active

Not really a lie. They said it was an estimate. I.e. they are not claiming 5% is true, just that it’s their best guess.

Links to those multiple independent research studies?

Those studies have no way of measuring users who only consume tweets.

Reply on news.ycombinator.com