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

Uber discovered they’d been defrauded out of 2/3 of their ad spend

twitter.com

Post reply on HN
Print view View original

531–540 of 933 posts

Re: Uber discovered they’d been defrauded out of 2/3 of their ad spend

#531

Post by 1f60c »

nutanc wrote: post #522

We did a similar kind of experiment some time back and found both FB ads and Google ads were doing jack shit for us. Google was better. FB ads were absolutely useless. So we stopped our ad spend. We observed no change in our customer acquisition. Another way these ad engines extract money out of you is by allowing competitors to bid for your brand name keyword. Try searching for Ozonetel on Google. There will be six…

> Try searching for Ozonetel on Google. There will be six competitor ads and then our website will show up :)

DHH (of RoR and Basecamp fame) has been beating the drum about this for a while now.

Re: Uber discovered they’d been defrauded out of 2/3 of their ad spend

#532

Post by accountofme »

coliveira wrote: post #89
dllthomas wrote: post #59

Earlier quoted context omitted.

> Why wouldn't someone jump at an opportunity to prove the thing/service they provide actually works, unless they were unsure about it themselves? Because everyone is already acting like they know it works, so the only way that experiment can change things is in a way that's bad for the person in question. In that situation, they should (from a local, selfish perspective) be resisting even if they're awfully sure it…

Not really, scientists for example also have their whole careers based on the truth of some theories that they use. However, they're willing to put them to proof in different ways. The reason they do so is that they have a high degree of confidence that these theories are true. This cannot be said of people doing advertisement.

Truth is relative. Sampling can skew results. Scientists are people too. We all have flaws.

Re: Uber discovered they’d been defrauded out of 2/3 of their ad spend

#533

Post by hertzrat »

pessimizer wrote: post #357
disgruntledphd2 wrote: post #84

Earlier quoted context omitted.

Facebook's ad research team have run a lot of experiments. The goal was to demonstrate an impact of in-store sales from internet advertising. They did the first studies in about 2008-10, and have continued running these studies ever since. They even built a tool so that advertisers can run these studies, and get a sense for the incremental impact of their ad. Google have a similar (less full-featured) system. Really,…

> The goal was to demonstrate an impact of in-store sales from internet advertising. So the goal was to advertise advertising. A more scientific goal would be to see if there was an increase of in-store sales from internet advertising. Instead they were looking to design experiments that would show a positive effect, with the goal of giving advertisers a dashboard so they could run those experiments themselves.

To let those advertisers do the experiments themselves, and to relate the campaign to in store sales, would require Facebook and google to share a ton lot of internet-user tracking data

Re: Uber discovered they’d been defrauded out of 2/3 of their ad spend

#534

Post by Reebz »

onlyrealcuzzo wrote: post #50

You can A/B test ads pretty easily, and this is quite common. With some degree of statistical certainty, you can tell how one ad performs to another. You don't have control over your SEO results as well - but you can also measure against SEO traffic with a high degree of certainty. All big companies do this. Sure, you're never going to know exactly how many people you advertised to would have organically, eventually…

Across this and the other[0] highly active digital ads post, there's 3 interesting forces at play:

a) technologists screaming "ads are literally the worst societal cost, like ever" who don't want to understand the industry

b) advertising folks taking up arms to defend their shamanistic, money-printing machines without statistics (Reason No. 7 will SHOCK you)

c) A tiny, tiny group of Mandalorian-like voices who have a necessary statistics AND industry understanding who are being drowned out

(this is in jest, but I double-dare you to say it isn't at least directionally accurate)

Other commenters have mentioned it throughout this whole comment thread as "incrementality tests". Amongst other approaches, this is the way.

Freakonomics severely lack the industry understanding. Listening to the podcast was like hearing how HTML is a programming language from the kid in week 2 at code camp. Then too the article, all the issues with Uber was just doing a bad job of managing their ad spend and they can fall into group B, noted above.

The baseline of this work is a control group who see no ads and then you build your tests from there that factor in channels, cohorting, and other components to get a statistically significant outcome. Yes, this will get more challenging with upcoming privacy changes (IDFA removal, et al). However, the last 10 years this wasn't a problem and I'm sure the corporations in the identity resolution business will hand-shake on a bunch of 2nd party data deals that just move the deals done in broad daylight around identity tracking and audience creation to the alley. Further, any advertising that is tied to an already known customer is able to be backed into at an audience level with login and cookie data. Even Pi-hole users may not be exempt here.

To finish with some constructive advice:

1) Advertising is not a synonym for marketing. We're only talking about advertising in both HN threads.

2) Every industry has high and low quality. Pareto's principle should be aggressively applied to where one spends their budget in the cesspool that is the Internet.

3) If you're ever seeking an agency to provide advertising services and they don't have a qualified data science or statistics leader (10+ years work experience, degree in Stats/Math/Econ, an MBA, or similar), run. Run from those shiny-shoes gurus. Channel your inner Usain Bolt and run.

0. https://news.ycombinator.com/item?id=25620707

Re: Uber discovered they’d been defrauded out of 2/3 of their ad spend

#535

Post by hertzrat »

accountofme wrote: post #532
coliveira wrote: post #89

Earlier quoted context omitted.

Not really, scientists for example also have their whole careers based on the truth of some theories that they use. However, they're willing to put them to proof in different ways. The reason they do so is that they have a high degree of confidence that these theories are true. This cannot be said of people doing advertisement.

Truth is relative. Sampling can skew results. Scientists are people too. We all have flaws.

If you really believe truth is relative, walk to the grocery store tomorrow without wearing clothes and convince yourself that it will be fine. Or run a lighter under your hand - maybe it won’t burn today? Of course you wouldn’t to those things though because, at the end of the day, we rely on truth for everything we do. We just don’t always know what it is. Relativism is not the same as “it’s hard to figure out because it’s so complex”

Re: Uber discovered they’d been defrauded out of 2/3 of their ad spend

#536

Post by nutanc »

1f60c wrote: post #531
nutanc wrote: post #522

We did a similar kind of experiment some time back and found both FB ads and Google ads were doing jack shit for us. Google was better. FB ads were absolutely useless. So we stopped our ad spend. We observed no change in our customer acquisition. Another way these ad engines extract money out of you is by allowing competitors to bid for your brand name keyword. Try searching for Ozonetel on Google. There will be six…

> Try searching for Ozonetel on Google. There will be six competitor ads and then our website will show up :) DHH (of RoR and Basecamp fame) has been beating the drum about this for a while now.

Yes, we have also been beating the drum since a long time. But unfortunately our drum is not as big as DHH :)

Re: Uber discovered they’d been defrauded out of 2/3 of their ad spend

#537

Post by pb7 »

nutanc wrote: post #536
1f60c wrote: post #531

Earlier quoted context omitted.

> Try searching for Ozonetel on Google. There will be six competitor ads and then our website will show up :) DHH (of RoR and Basecamp fame) has been beating the drum about this for a while now.

Yes, we have also been beating the drum since a long time. But unfortunately our drum is not as big as DHH :)

Curious: why do you feel your competitors shouldn't be able to do this? If your customers intend to go directly to you, why do they not go directly to your site but instead search for it on a general search engine? More generally: if I search for "apple", what should the first result be? Wikipedia or Apple.com?

Re: Uber discovered they’d been defrauded out of 2/3 of their ad spend

#538

Post by lotsofpulp »

friendlybus wrote: post #385
tqi wrote: post #122

Earlier quoted context omitted.

If I search "Kleenex", am I specifically looking for the Kimberly Clark product or should Procter & Gamble be allowed to bid on ads for that search?

If I search "Kleenex" on Google, I'm asking Google to find me Kleenex, not some ads. If Google is going to start listing things in highest-to-lowest bid, then it stops being a search engine and starts becoming an auction house of links.

It already is an auction house of links. I think you have to go down at least 6 or 7 links nowadays for actual search results.

Re: Uber discovered they’d been defrauded out of 2/3 of their ad spend

#539

Post by base3 »

adwww wrote: post #168
Nextgrid wrote: post #30

Earlier quoted context omitted.

> No one is willing to run the experiments necessary The people that would have the power to run this experiment have their entire careers depending on things staying as-is. Running the experiment carries a significant risk of exposing that the advertising operations they're responsible for provide much less ROI than they pretend it does. The unwillingness of anyone to run such an experiment is already an answer. Why…

I used to work in this industry. A small tech team investigated fraud on our platform and developed a system that was pretty robust at detecting and potentially shutting it down. But literally nobody was interested - even the people advertising don't want to know. The people spending money are typically networks, media buyers, ad agencies, etc, far removed from the actual brand. There are so many parties who want a s…

I worked in adtech. I built a simple ML system that detected signals of fraud on our network. When I reported my findings execs told me the group that was facilitating the fraud was our only profitable division. Soon after, our company went bankrupt. My final paycheck was months late.

Re: Uber discovered they’d been defrauded out of 2/3 of their ad spend

#540

Post by gundmc »

mschuster91 wrote: post #392
TheOtherHobbes wrote: post #215

Earlier quoted context omitted.

Coca Cola sells water with flavouring and quite astonishing amounts of sugar. Apple sells nice devices made by what might as well be slave labour. Both are propped up - actually maintained - by huge and vastly expensive brand management strategies. The criticisms of individual campaigns here are missing the point. It's not about micro ad spend, but the perception of value and manipulation of behaviour created by the…

> Both are propped up - actually maintained - by huge and vastly expensive brand management strategies. In the case of Coca Cola, I agree - in the case of Apple, not at all . "Keyboardgate" aside, their products are vastly superior to the competition in build quality and life time. A typical Windows laptop is unusable after two, three years (almost all are made from plastic which breaks or looks extremely ugly rather…

I've been using my windows laptop daily for 5.5 years without issue.
Post reply on HN