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

291–300 of 312 posts

Re: Bots are getting good at mimicking engagement

#291

Post by falcor84 »

frodo8sam wrote: post #79
malfist wrote: post #73

Earlier quoted context omitted.

Google ads are often like paying someone to stand in front of your business and hand out flyers to everyone going inside. Sure, a lot of your visitors were handed a flyer by google before they showed up in your store, but were any of them new customers?

It's worse, you're paying them so your competitor isn't standing in front of your store convincing them not to be your costumer. You can either buy the top spot on queries that are looking for you or google will sell it to your competitor.

It's even worse, because it's Google who built a high platform in front of your store in the first place, with comfy chairs for your competitors to sit on.

Re: Bots are getting good at mimicking engagement

#292

Post by dredmorbius »

VBprogrammer wrote: post #265
ww520 wrote: post #223

Earlier quoted context omitted.

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 th…

I cut / you choose works in some situations.

In others you'd want, say, auditing or independent third-party verification.

In this case, perhaps an audit involving the deliberate injection of a mix of legitimate and bot traffic to see how much of the bot traffic was accurately detected by the ad platform. Rates on total traffic could be adjusted accordingly.

This of course leads to more complications, including detection of trial interactions, see e.g., the 2010 VW diesel emissions scandal: https://en.wikipedia.org/wiki/Volkswagen_emissions_scandal>, or current AIs which can successfully identify when they're being tested.

Re: Bots are getting good at mimicking engagement

#293

Post by mtrovo »

xanderlewis wrote: post #285
dgfl wrote: post #261

Earlier quoted context omitted.

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…

> What makes it sound overly AI are the semicolon and the timing It's very disappointing that it now seems to be commonly believed that semicolons and em dashes are a sign of AI. Semicolons have been used as a useful tool for a long time; a comma here would simply be wrong.

I think the problem is the overuse compared to the times before ChatGPT. Same for furthermore, thus, delve and other words. Heck em dashes are so obscure that I doubt 1% of the population knows how to type it on their own keyboard.

Those heavily exposed to media before ChatGPT feel there's a shift on the way sentences and narratives are presented. We usually don't use that many semicolons because we structure our sentences not to use them, same for dashes or other weird constructs used by ChatGPT right now. It's hard to explain with words but there are a lot of tell signs when you start looking for them. And not even talking about the explicit ones like bullet points with "bold title:", one that really pulls the curtain open for me is the phrases that end with "one, two, and three of something". When you see it once it's hard to unsee it used everywhere, even on sites that are supposed to be user content only like here or reddit.

Re: Bots are getting good at mimicking engagement

#294

Post by V__ »

shiftingleft wrote: post #47
V__ wrote: post #38

Earlier quoted context omitted.

I could imagine that blocking bot traffic, would improve their retargeting and make sure that the retargeting budget is spent on real people leading to an increase in conversion.

What's the API here for Google Ads? How does their site report to Google Ads whether that was a good/bad user? Is this done through conversion tracking? If so, why would you track anything but a completed purchase in the first place?

I think Google calls it remarketing and it goes through Google Tag Manager. You can "tag" visitors how you want (duration, action, page scroll, etc.). It's just a javascript call to the API which you can trigger however you like.

You wouldn't necessarily want to track conversions for retargeting, since depending on your product or service, a second buy might be unlikely. But someone who checks out multiple product pages or articles on your site might be interested and buy in the near future. That of course are also actions bots could easily do.

Re: Bots are getting good at mimicking engagement

#295

Post by boringg »

buellerbueller wrote: post #242
boringg wrote: post #70

Earlier quoted context omitted.

This is well understood for well over a decade at this point -- the blogpost is a marketing effort by datacops.

Then it is securities fraud. Advertisers' shareholder money is being wasted in a process that is "well understood" to be lies, funneling money to tech industry.

That may be true but good luck ever bringing a case and validating any of it. I am sure those companies that are suspicious are also aware that they are looking at a money pit of fees without any chance of a reliable outcome.

I'm pretty sure that whole market is a game of horse shoes and hand grenades.

Re: Bots are getting good at mimicking engagement

#296

Post by maroon_unperson »

shiftingleft wrote: post #36

After we implemented advanced bot traffic detection and filtering, their reported traffic plummeted by 71%. [...] But then the sales report came in. Their actual sales went up by 34%. Their real conversion rate optimization (CRO) efforts had been working all along, but the results were buried under an avalanche of fake clicks. They were not bad at marketing; they were just spending thousands of dollars advertising to…

Sites send conversion events back to Google so they can target highly converting traffic.

If a bot network hits all the conversing events then Google will tailor the traffic to look more like the bot network.

If you filter the bot traffic out then Google can tailor the traffic to look like real converting users instead.

Re: Bots are getting good at mimicking engagement

#297

Post by bluGill »

RamtinJ95 wrote: post #288
bluGill wrote: post #173

Ads have been around for a long time. If your great-...-grandpa bought a newspaper ad in 1820 he would have been looking to see if the ad worked (back then 90+% of people were farmers who didn't buy ads, but lets assume you had an ancestor then who was a shopkeeper and thus had potential value in ads). He probably didn't have have modern statistics, but he would pay attention to sales and try to figure out how much v…

Do you have any links to sources around these other ways to get the data? I recently started working in this domain and trying to get up to speed on state of the art regardings these types of things specifically since im in the data team.

No. I once knew someone who did this for a big company back before the internet took off, but we lost touch. It is basic applied statistics, but likely there is a lot more than I don't know anything about. Likely you can ask someone who has done more study in that area to get details.

Re: Bots are getting good at mimicking engagement

#299

Post by dredmorbius »

dredmorbius wrote: post #292
VBprogrammer wrote: post #265

Earlier quoted context omitted.

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 th…

I cut / you choose works in some situations. In others you'd want, say, auditing or independent third-party verification. In this case, perhaps an audit involving the deliberate injection of a mix of legitimate and bot traffic to see how much of the bot traffic was accurately detected by the ad platform. Rates on total traffic could be adjusted accordingly. This of course leads to more complications, including detect…

On further reflection: I'd thought of raising the question of which situations cut/choose does work. Generally it seems to be where an allocation-division decision is being made, and the allocation is largely simultaneous with the division, with both parties having equal information as to value. Or so it seems to me, though I think the question's worth thinking about more thoroughly.

That's a subset of multi-party decisionmaking situations, though it's a useful one to keep in mind.

Re: Bots are getting good at mimicking engagement

#300

Post by VBprogrammer »

dredmorbius wrote: post #299
dredmorbius wrote: post #292

Earlier quoted context omitted.

I cut / you choose works in some situations. In others you'd want, say, auditing or independent third-party verification. In this case, perhaps an audit involving the deliberate injection of a mix of legitimate and bot traffic to see how much of the bot traffic was accurately detected by the ad platform. Rates on total traffic could be adjusted accordingly. This of course leads to more complications, including detect…

On further reflection: I'd thought of raising the question of which situations cut/choose does work. Generally it seems to be where an allocation-division decision is being made, and the allocation is largely simultaneous with the division, with both parties having equal information as to value. Or so it seems to me, though I think the question's worth thinking about more thoroughly. That's a subset of multi-party de…

I vaguely remember someone winning a noble prize for economics for coming up with ways to apply cut/choose in financial transactions but I couldn't find it in a quick Google. It may have been nearly 20 years ago though.
Post reply on HN