Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Bots are getting good at mimicking engagement

joindatacops.com

Post reply on HN
Print view View original

261–270 of 312 posts

Re: Bots are getting good at mimicking engagement

#261

Post by dgfl »

getnormality wrote: post #245
dgfl wrote: post #216

It’s interesting that this matches _exactly_ the ChatGPT writing style. There’s clearly some human data and work behind, but the entire article has the usual “it’s not just X, it’s …”, bold emphasis on specific words, bullet points, and probably more that I’m just subconsciously picking up. Nothing wrong with using ChatGPT for help of course, I just found it interesting and kind of ironic given the contents of the ar…

The really ironic thing is that this article demonstrates the correct, purposeful human usage of these stock phrases and is getting skewered for it because it superficially sounds like AI. It's not exemplifying what it's criticizing, it's exemplifying the exact reverse of what it's criticizing. "It's not X, it's Y" is for when you know your reader will think X, and you want them to understand the difference between t…

I mostly disagree.

> I was not just counting clicks; I was watching behavior.

This is completely superfluous. The rest of the paragraph is fine, but the only real ChatGPTism is this one.

> They are not just hitting your site and leaving; they are programmed to mimic engagement

In this case it’s a valid sentence. What makes it sound overly AI are the semicolon and the timing, which is exactly the usual “it’s not just {1-5 words}; it’s {1-5 words}”.

Also, perhaps I’m not a typical audience member (I’m not), but I was not assuming any of that. I would try to explain how I parse this kind of language, but the gist is that a 10 sentence word with unambiguous meaning doesn’t really leave space for interpretation. Your attempt to “anticipate what the audience is thinking” is just a writing style, nothing deep about it.

My point mostly being that, as I mentioned in the parent, this is clearly a text written in the ChatGPT style, with simple yet over-expressive language. I personally heavily dislike the style, but I’m a non native speaker used to either scientific or fiction prose. I don’t encounter it too often, and to me it sounds like a parody of Silicon Valley vernacular. Scam alerts immediately start firing in my brain. There’s no “good way” of doing it nowadays.

Re: Bots are getting good at mimicking engagement

#262

Post by logsr »

KaiserPro wrote: post #156
simul007 wrote: post #2

Hi HN. I run a marketing agency and fell down this rabbit hole after a client's analytics made no sense (50k visitors, 47 sales). I ended up building a simple script to track user behavior and analyzed 200+ small e-commerce sites. The average was 73% bot traffic that standard analytics counts as real. The bots are getting creepily good at mimicking engagement. I wrote up my findings, including some of the bizarre pat…

But what I'm struggling to understand is where is the money coming from to run those bots? How does the advertising money get to them to make it worth their while to run bots at such a scale? I mean sure, I suspect that people in the advertisers are doing it, but surely thats massively risky. there must be a grey market for this kind of transaction?

for ad networks and social media platforms that provide monetization the click fraud is direct.

there is also a massive industry of fake accounts and fake engagement for social media and SEO (google). bots are designed to create plausibly real engagement, which is used to trick ranking algorithms into boosting content. these bots have to be real enough to bypass platform detection. clicking through on ads is a way of incentivizing platforms not to shut them down and possibly improving the ranking results, working with the theory that platforms give stronger weight to engagement signals from clients that generate more revenue.

Re: Bots are getting good at mimicking engagement

#263

Post by Gigachad »

LeifCarrotson wrote: post #224
Gigachad wrote: post #195

Earlier quoted context omitted.

YouTube didn’t stop counting views with ad blockers. It was one blocker extension that added the view counter api to the block list.

That's splitting hairs, but in a way that's important to the conversation. If YouTube served gigabytes of the video file for 40 munutes and human watched it for that time, but they didn't send a request to `youtube.com/api/stats/atr/` and periodically to `/stats/qoe?`, did the video actually get viewed? I think a reasonable person would say that the person viewed that video. Only a programmer would suggest that wasn'…

I think most websites would break if a 3rd party script started blocking things. Theres also the fact that view tracking is fairly complex since they need to filter out bots / ad fraud.

And just the fact that if users have a privacy extension blocking the view tracker, is it not just respecting their wishes to not be tracked?

Re: Bots are getting good at mimicking engagement

#265

Post by VBprogrammer »

ww520 wrote: post #223
seviu wrote: post #68

Earlier quoted context omitted.

I once worked for the yellow pages in Switzerland. Our paid clients had a dashboard which reported how many users visited their business entry. We at engineering decided to filter out bots. Figures fell dramatically by more than 50%. In less that a day business mandated us to remove the filter. Bots are real people after all

Way back when I work in an ad based company, click fraud handling was under my overseeing. We caught about 20 percent of clicks as fraudulent and filtered them out before billing the ad placing vendors. It was a constant battle with the sales team to relax the rules, as any clicks filtered out cut into the sales revenue. Sometimes we got the customers on our side as they ran their own analysis on the billed click rep…

Yeah, the incentives there are obviously misaligned. I wonder if there is a potential way of making advertising click through tracking following the "I cut the cake, you chose the slice" model.

Some countries have property taxes where you declare the value and the government retains the right to purchase the property for that value for example.

My first thought was to make the advertising cost driven by revenue on the site. But that just reverses the incentive.

Re: Bots are getting good at mimicking engagement

#266

Post by lurk2 »

criddell wrote: post #217
lurk2 wrote: post #175

Earlier quoted context omitted.

> When a company puts up a billboard or an ad on the bus, they don't care if the ad is seen by dashcams and dogs. Billboards sell for flat rates, online advertising is sold on the basis of impressions.

Billboard rates are based on impressions too. A billboard along a busy road is more expensive than one out in the sticks.

The pricing of a billboard is based on projected impressions but is decoupled from the billboard’s actual impressions. You pay the flat rate based on projected impressions. It doesn’t subsequently matter if an army of robots passes through and views the billboard 10,000 times, because the original projected audience should still see the billboard. You’re getting what you paid for. It’s also possible the projections are erroneous, fraudulent, or that the actual impressions just don’t follow the projected impressions for some other reason. This is the advantage of advertising online—if you pay for 10,000 views, the ad runs until you get them.

Contrast the billboard with an online ad: If you pay for 10,000 impressions, and 9,000 of those impressions end up being automated traffic, then you’ve effectively lost 90% of your advertising budget. Like you mentioned, ad buyers can and do account for this, so it’s not clear that it’s as big of a problem as this thread is making it out to be. The theory I see thrown around is that Meta and Alphabet are tacitly allowing this bot traffic to continue (or even that they are responsible for it themselves) because it allows them to sell another round of ads, but if this is what is going on it isn’t clear why firms would keep buying ads that aren’t converting.

Re: Bots are getting good at mimicking engagement

#267

Post by mustaphah »

chrismorgan wrote: post #17

I’m puzzled by this: I thought it was well-understood, at least in the industry, that traffic numbers were at least mostly nonsense, and that ad click metrics especially were suuuper shady, typically more than half fraud; yet OP, in the business of “accurate ad spend analytics”, only just discovered this!? It just doesn’t ring true. That aspect of the story isn’t novel at all, and someone in that line of work should…

> Yet OP, in the business of "accurate ad spend analytics", only just discovered this!?

Looks like a very fresh business [1]; it only became worth writing about once they were ready to pitch their product.

[1] https://whois.domaintools.com/joindatacops.com

Re: Bots are getting good at mimicking engagement

#268

Post by joomla199 »

oconaros wrote: post #250
simul007 wrote: post #2

Hi HN. I run a marketing agency and fell down this rabbit hole after a client's analytics made no sense (50k visitors, 47 sales). I ended up building a simple script to track user behavior and analyzed 200+ small e-commerce sites. The average was 73% bot traffic that standard analytics counts as real. The bots are getting creepily good at mimicking engagement. I wrote up my findings, including some of the bizarre pat…

I do web analytics consulting. One of my first projects at a digital marketing agency in 2021 was investigating weird traffic patterns for a global logistics firm. The findings are summarized in this blog post[1]. Bot traffic has been an issue for years. It has given rise to a new cottage industry of ad fraud detection services, none of which I’ve found particularly valuable. It always comes down to “so what do we do…

Given that Wanamaker died in 1922 it’s safe to say his quote was in the context of a different problem entirely (which still exists, on top of the bot issue). Maybe it’s time to update to “three fourths”?

Re: Bots are getting good at mimicking engagement

#269

Post by arccy »

LeifCarrotson wrote: post #224
Gigachad wrote: post #195

Earlier quoted context omitted.

YouTube didn’t stop counting views with ad blockers. It was one blocker extension that added the view counter api to the block list.

That's splitting hairs, but in a way that's important to the conversation. If YouTube served gigabytes of the video file for 40 munutes and human watched it for that time, but they didn't send a request to `youtube.com/api/stats/atr/` and periodically to `/stats/qoe?`, did the video actually get viewed? I think a reasonable person would say that the person viewed that video. Only a programmer would suggest that wasn'…

if a bot (or whatever latest of download tool) was served gigabytes of video, does that mean a human eventually viewed it? no.

Re: Bots are getting good at mimicking engagement

#270

Post by jimkleiber »

oconaros wrote: post #250
simul007 wrote: post #2

Hi HN. I run a marketing agency and fell down this rabbit hole after a client's analytics made no sense (50k visitors, 47 sales). I ended up building a simple script to track user behavior and analyzed 200+ small e-commerce sites. The average was 73% bot traffic that standard analytics counts as real. The bots are getting creepily good at mimicking engagement. I wrote up my findings, including some of the bizarre pat…

I do web analytics consulting. One of my first projects at a digital marketing agency in 2021 was investigating weird traffic patterns for a global logistics firm. The findings are summarized in this blog post[1]. Bot traffic has been an issue for years. It has given rise to a new cottage industry of ad fraud detection services, none of which I’ve found particularly valuable. It always comes down to “so what do we do…

I'd suggest legislation. It's like when hotels or rental car companies tack on additional fees and consumers hate it but don't have much incentive to cooperate to change it. Consumers don't unionize much, but maybe we should?

But maybe not legislation, because it might upset a ton of people. Any time a social media site goes through and seems to get rid of bots, people complain about how many "followers" they've lost. The fake user/clicker situation is quite pervasive, and a lot of parties, not just the advertising networks, benefit from the inflated numbers.

But it can be so insidious..."wow, that video got 1M YouTube views? Must be very popular!" Orrrr a lot of those views were from bots? Who knows?

So maybe a better approach than legislation is to talk about how the ad/bot fraud can both help us and hurt us and not demonize one side but see how we all may be implicated in it somehow. Maybe that will help us to be more aware of the problem and not fight against people, but try to work together to solve it.

Post reply on HN