Ask HN: Does targeted advertising work at all?
I'm too lazy to run ad blockers. So I see plenty of advertising on the internet. Here's the thing: all the ads seem to be for things that I already bought. After I placed an order with Select Blinds, I started getting relentless Select Blinds ads on YouTube. After I made some charitable donations to Mercy Ships and Heifer International, their ads started following me around everywhere. My wife recently got on a new immunosuppressant medication, and now I see ads for that all the time. Great job targeting! But I already spent the money? So what is the point of all this?
157 comments
[ 9.4 ms ] story [ 249 ms ] threadEspecially something like remarketing must be hard to do a-b testing on, as the control group is basically not tracked/targetet at all.
For example, maybe you do not browse a lot before spending money because you know what you want more or less?
I have the habit of clearing my browsing data often, all the tracking options in my Google account are off, and the ads I get seem pretty generic in general, but I sure get ads related to my Google searches if there is something concrete to purchase, like hardware parts.
I’ve posted this comment before, but:
1. It’s hard to tell when someone has bought a product. You buy it at a reseller and the brand will never know. Or you buy it on mobile, but have researched the brand on desktop. Or you purchased it at work, but were researching it at home. Advertisers suppress known buyers but they don’t perfectly match you to all of your devices.
2. Just like there are not so good developers and not so good code, there are not so good marketers who aren’t that great at their job. Or they have too much work and didn’t set it up correctly. DSPs are extremely extremely complicated, so it’s to make mistakes.
3. Retention and advertising to previous purchasers drives more additional purchases. Maybe you only purchased blinds for the main floor of your house but not the basement.
How about, "I've literally got it in my last-day purchase history on eBay and they keep showing me the same thing and a bunch of related stuff to the item that hasn't even arrived yet, but that they've taken my money for"?
... and then eBay follows me around showing the same stuff on other sites. The thing I've purchased. From eBay. At least, they did, until I got more aggressive blocking ads and started using the internet a ton less.
Also likely the app is a totally different team and getting them to change a link is difficult, but giving them a link where the web team can change the redirect is easy.
Genius.
(I'm only being half-facetious)
Interesting idea. Probably still would be cheaper than paying 30% tax to Apple.
For example, sometimes for suppression brands literally just export a csv of customers and upload it to google. If your marketing guy does it once a week..
I’d assume eBay’s doesn’t work like that, but it’s possible. Also possible they just don’t suppress recently bought items !
I've often bought stuff that didn't work for some reason. It was broken, or it had something that was bothering me. Maybe the power adapter was annoyingly loud (looking at you, Kyocera).
So I sent it back. But before I managed to go to the postal office a few days later I was already searching for a replacement online.
If I buy a TV set, I have indicated that I need/want a TV set. The rate of returns is broadly known (and maybe even specifically for me the return rate is known).
So Amazon et. al. do well to recommend similar stuff or things from the exact same product category to me.
Even right after I have bought a graphics tablet I'm infinitely more likely to buy a graphics tablet tomorrow, compared to my parents who have never bought one in their lives, and never will.
Or maybe the new TV turns out to be great, and I realize it would be nice to have a second one in my sleeping room, and why not buy a third one for the bathroom? ;-)
While plausible, that sort of situation doesn't sound common enough that it should underpin the entire online advertising industry.
Even right after I have bought a graphics tablet I'm infinitely more likely to buy a graphics tablet tomorrow, compared to my parents who have never bought one in their lives, and never will.
But you're more likely still to buy another stylus, or art software, or an accessory, or physical art materials, etc. Why don't advertisers understand that?
I still maintain that people who have bought a TV yesterday are more likely to buy a TV today than people who haven't bought a TV yesterday. Very few people buy a TV today. Or in the coming days.
It seems like a straightforward application of the base rate fallacy. Sure, most people who bought the TV yesterday keep their TV and don't need a second one.
But unless you have specific information about other people's interest in buying one, they may still be your best shot at selling another one.
Also, both you and sibling commenter argue as if there were a scarcity of ads. Amazon showing me a TV ad doesn't preclude Amazon from showing me a kitchen utensil ad on the next page.
And it is completely at odds with how I buy things. Or how people I know buy things.
> I still maintain that people who have bought a TV yesterday are more likely to buy a TV today than people who haven't bought a TV yesterday.
I never ever seen anyone buy second TV shortly after buying first one. Never.
> Amazon showing me a TV ad doesn't preclude Amazon from showing me a kitchen utensil ad on the next page.
I get monothematic ads. There is always that one thing that follows me everywhere. I don't get mix of ads. It is as if their system thought I buy only shoes. Or only TV.
I simply cannot believe that this superficially irrational behaviour is as irrational as always asserted on HN.
The big companies are spending a gazillion dollars on ad tech, and while I believe that there is lots of stupid money floating around, I don't believe it's so obviously irrational. Just look at behavioral biology or psychology how much of seemingly irrational or downright stupid behaviour makes a lot of sense when you find the right vantage point.
Unfortunately, there will probably nobody come forward and tell us how it really is, so we can only speculate and build our own pet theories.
Yes, but those tend to make you notice that "yep I have actually seen this". This one does not make me think that. Not when it comes to TVs, chairs, beds or other durable items. If we talked about beer or chocolate, sure. Someone who bought beer might want to buy exact same beer.
Someone who bought book might want to buy another book. But, not that exact same book again, that person might want to buy another similar book. With TV, no, nope.
Like, I’m the opposite of interested in buying it, I’m trying to get rid of it.
There are zero incentives for anyone to figure this out and fix it within corporate structures. I'd be thankful it makes advertising easy to spot, and less effective in real life. ;-)
Imagine if you're in the middle management of Google and you realize that target ads don't work, and the results to this point are just luck. You'd want to do something like "Improve your advertising spend" by making the ads "more efficient", and selling targeted ads to people that you know have already purchased, but the advertiser doesn't.
Eventually you could advertise that your ads are measured to be more than 80% correlated with a rise in sales. (Leaving out the reversed causality).
If this IS the case... shhhhhh... don't tell anyone. It would crash Google's stock price. Nah, who am I kidding, the market stays irrational for decades. ;-)
Never attribute to malice, that which can be explained by incompetence.
The book Subprime Attention Crisis by Tim Hwang does a good job covering this! (It is a solid bear case for fb,g,etc as well :p)
Nobody is confusing correlation and causation. These are rigorous studies completed by literal data scientists and statisticians. They are employed by neutral third parties who make money regardless of the outcome of the study, and showing positive outcomes where they do not occur would be hugely detrimental to their business.
And to your point, realizing advertising does not work happens all the fucking time. I see it literally all the time. It’s why you run studies. Then you take your budget, move it somewhere else, and run a test to see if it’s effective. Every major brand does it this way.
This is not completely accurate in my experience. Some brands do, particularly emerging brands.
However, I can assure you there are a few major established brands that I could drop on here (but cannot due to NDAs) that everybody on HN regardless of their location worldwide would recognize. If I was to tell you the inner workings their marketing operations you would be surprised at the lack of measurement and analysis they actually perform on campaigns. The words “indiscriminate carpet bombing and assume the enemies are dead” are probably the best analogy I can think of to describe their efforts.
As much as I am critical of the approach, I can’t really says it’s a bad strategy, as these are brands that had worldwide reputations long before I became involved and most of my involvement was pretty narrow to a specific product or product vertical.
No, which is why advertisers are theatre.
Many large businesses have entirely halted online advertising and found little effect, but much saved money.
It's fairly known in the industry (worked in a range of different ad-tech/marketing companies for around a decade), that more or less this is what makes Nike happy.
> They are employed by neutral third parties who make money regardless of the outcome of the study.
This is a bit naive. Same is true for home appraisers, funny how it always works out that the house appraised for just over what you offered (even in insane markets).
This issue is you might get paid this time for being honest, but you won't get paid again.
> completed by literal data scientists and statisticians.
I'm both of these. Saying that "adtech is bullshit" is absolutely too reductionist but you are wildly too naive about the reality of the industry.
Take A/B testing as an example. I've helped people run A/B tests for many years. I establish rigorous testing setups, make sure people understand confidence in the results, distribution of possible outcomes, etc. I'm honest in my work and so are the people running the tests.
But the truth is that A/B test on customer populations are not controlled experiments. One thing that literally no one in the industry does is go back and review the results of the past years A/B tests. The reason why? People don't really want to know. However, it's obvious these results can't be quite right because teams will run 10 A/B tests in a year, each claiming a 5% improvement in conversion, but clearly you don't see this cumulative impact of a >60% increase in conversion over time.
No one want to know this though, because knowing this helps nobody. Nobody is outright lying, but nobody wants to ask. The one time I worked outside of adtech but saw a problem where model results were almost entirely random noise I got in huge trouble for pointing this out.
Same is true across the board in the adtech world.
Someone else recommended "Subprime Attention Crisis by Tim Hwang" and I very strongly recommend that you read this. It's very well researched and was so inline with my experience that I found it so obvious as to almost be dull.
Bob Hoffman's take on advertising e.g. Advertising for Skeptics, is well worth reading - he's an ex-adman who thinks the ad industry and particularly the personalised ad industry is full of shit
When you look at the online ad industry it's reasonably clear that the people who are making the money are the ad tech companies and not the advertisers
The VP of marketing of ebay realized that the keyword for which they spent the most money was "eBay". He thought this was nonsense and that people who looked for eBay should know how to get to their website, so he decided to save some marketing money and stopped putting money on they keyword. Result: no changes in the amount of visit to the website. However, the click-through rate, which is the metric by which marketing department was measured, decreased significantly. So the VP was fired and they even increased the money spent on the keyword.
We see this in the software industry all the time. Large organizations act inefficiently or against their own interests because large organizations don't actually have consciousness, instead their decisions emerge from the sum of a lot of little decisions and actions, each of which might be individually reasonable.
As an aside, I believe that's how large organizations can come to be evil even if the vast majority of members are not evil.
https://freakonomics.com/podcast/does-advertising-actually-w...
Markets are imperfect and irrational, but you really don't think that there is a trillion dollar misunderstanding about the efficacy of the practice? Idk jack about advertising but it seems unlikely to me.
Just the fact that people buy ad campaigns doesn't mean that they have any effect at all.
And in case of doubt ... they found that it is more advantageous to them to just keep bothering you.
I still get a lot of crappy ads and get bombarded with the same product I just purchased at times. But, it's by far the most customized/useful ad experience I've had.
Reminds me of the now-periodic posts on HN grousing about how bad Google has become. The complainers expect too much. Google is the modern Yellow Pages, or Computer Shopper. You use it for the ads (and SEO stuff). It’s good to look for the highest bidder - that indicates a powerful bidder who can afford it. In the old days this was picking the plumber with the big Yellow Pages ad. At least they’re more likely to be in business with someone answering the phone. You’ll have much worse average luck calling the business with no ad that just has a one-line phone number listing.
I have a list of authors I follow on amazon.com. It would be a better use of everyone's time if they just picked a random book from any of my favorite authors, made sure I hadn't already purchased it through amazon, and only then show it to me. It's like they're not even trying.
Maybe you do. I'm subscribed to Asimov's and SF&F electronic editions digitally via the Kindle store. Not even once have I noticed an ad for scifi/fantasy books on my Kindle. Most of the time it's fluffy bestsellers.
A remarketing ad can be set up using exclusion lists to stop display after a purchase ( checkout page cookie), also most should be set up with limits. These are basic,obvious and easy things to implement. Attribution isn't always perfect but what you observed is more of a reflection of the high frequency of low tier ad managers. Many don't know any better, many don't care, many have no access to actually edit a client website (or know how)
It's the wild west still.
The example you used is also of a brand that I am overmarketed to, whoever is spending their money has some belief or knowledge of very broad targeting that I don't understand. I have not nor ever will buy blinds and my browsing behavior does not align with the obvious interest groups one would pick.
It's not just the small agencies that blow money, someone is patting themselves on the back about the 100,000,000+ impressions they are generating daily for selectblinds, but that funnel is going to be ugly when it's time to jazz up their presentation to show off that success!
There are ads that want you to buy something, and there are ads that, when you are making a choice between products, you’ll go with the brand you’re most familiar with.
Ads build that familiarity.
Except Oatly.
Because when I lived in NYC, their subway advertising was so aggressive, in your face, and annoying, I swore I'd never buy the product. Because no brand that spends that much on advertising can possibly be price-efficient for their actual product.
I've actually tried it at a friend's house recently, and really enjoyed it. But I still refuse to support a brand that advertises so heavily.
I believe that might not necessarily be true, in fact the opposite might be true.
A brand that can afford to spend a large amount of money on advertising cannot afford to sell bad products, as otherwise people would buy it once and then never again.
Reminds me of this scene from Pirates of the Caribbean..
Norrington : You are without doubt the worst pirate I've ever heard of! Jack Sparrow : But you have heard of me.
Aggressive subway advertising can indeed be annoying, yet it has been quite efficient in this case: the milk of a brand advertised in the subway of a big city just got mentioned in a positive way on a tech website read worldwide.
That's what advertising addresses.
Sometimes it takes years…
I did both academic research (recommendation engine algorithms) and industry work (leading online advertising at a large e-commerce ). I can confidently say from my background that the targeted advertising works.
In fact, it works so well that it is almost awe-inspiring to see it in real action. Any e-commerce or advertising firms, including my former team, have internal data to support this claim. For our case, the difference between a highly optimized traditional adverting vs a crude targeted advertising algo is a still factor of N in favor of the algo (N >> 1). As the algo is optimized, N grows larger.
My data at the time led me to believe that there is a small-ish part of the population on which targeted advertising is effective, and all the effort (information-theoretic wise) is locating this group, when your only tool (in real time bidding, anyway) is bidding on random ad impressions.
I’m sure the world is different now, and this inight might not be relevant to today’s world. But it does explain why re-targeting is way more effective than most people assume.
I do peruse ads in my muscle car magazines, because that's why I buy those magazines.
I enjoy the ad placements in those car hotrodding TV shows, because I am interested in how to do the projects they're doing, and the stuff needed to do them.
Amazon pushes ads to my kindle, but I couldn't tell you what they were for because they don't register on my brain.
Still use ad blockers a lot of the time but that’s cause of the insidious tracking not cause I’m against advertising as a concept.
To be clear, I suck at this. I've attempted to buy instagram / facebook ads and had little luck. For FaceBook ads you basically have to spend around $6k before the algorithm even starts to care about what you're doing. This is essentially because it has to first serve enough ads to figure out if people are actually clicking before actually targeting the demographic you set.
That said, as much as I hate it, I have no clue how to do social media marketing (TikTok or otherwise) and it's astonishing how much energy and time this takes. I used to joke that it'd be a better use of my time to find a college girl who likes TikTok and has a stimulant prescription to run social media for my bad side-project. (this is a joke)
However, targeted ads definitely work. I still get ads on instagram for the rapid-test clinic that just opened up across the street from my apartment. I can literally see the clinic from my window.
How is this working? I mean, I didn't think the claim was that targeted ads don't understand geography. The question was about if they drive behavior. And since you've seen the clinic every day, how does the advertising help?
Last year I began using THC to sleep better. And last week I experienced developing a deep craving for a specific snack food I saw an ad for. And everything clicked for me: advertising is probably very very effective on more easily manipulable people. People who are thrall to their lower brain impulses. I felt like I was six again. I just had to have a Skor bar so badly.
a) Someone recently bought blinds
b) Someone bought blinds 5 years ago
c) Someone hasn't bought blinds ever (maybe they're renting, or just tape newspaper to their windows)
Turns out, on average, the answer is a. Either because you're returning blinds you don't like and are in the market for replacements, or because you're doing a home remodel and once you fix up one window you'll need blinds for the next one. Even if most people who bought blinds won't be buying new ones in the next week, those are still better odds than advertising to someone who will never buy blinds in their entire life.
What do you think if you included the option d) Someone recently made a search that included the word "blinds" ?
The fact is, online ads are super cheap.
The thing is, they're all wrong. Advertising is a trillion-dollar business because it works. The human mind is susceptible to suggestion, and it's been demonstrated over and over again for basically as long as we've had civilization.
What is ALSO true, however, is that ad dollars are exceeeeeedingly inefficiently deployed. Which is why you experience things like weird retargeting on things you've already boight. I experience this too -- there are several brands that for which I'm a happy paying customer and I still see their ads 1-2 times a day.
* where I can. I still struggle with "sponsored" content, or fake review sites. It's becoming harder and harder to figure out what isn't an ad, and I hate it. If I see a good review, I don't know if it isn't an ad. If I see a bad review, I don't know it's not a competitor's doing.
I don't doubt that advertising works on me at some level. This is all the more reason to limit my own exposure to it. I'm one of those people who reacts with surprise and revulsion when we see what the Web looks like with ads.
>>> Which is why you experience things like weird retargeting on things you've already boight.
My favorites so far are the e-mails I get from online vendors, that say things like:
"People who bought shoes have also bought these items..."
"Here are some things to go with your batteries..."
All you've said is "A big industry does this. Therefore it works on everyone".
Which doesn't follow at all. The industry doesn't have to work on everyone to be profitable, and could operate effectively even knowing that. Whether they'd be able to take peoples' money as effectively if they said "our ads work on 73% of people, making them 0.45% more likely to buy the advertised product"... doesn't matter.
The existence of whales on the online gaming market suggests a counter-strategy might be more effective: some people are very, very susceptible to ads. Targetting them is likely to be both cheaper and more effective than trying to get the attention of people who are minimally affected by ads.
It's hardly shocking news that the advertising industry might lie, though. To the clients as well as their targets.
>Advertising is a trillion-dollar business because it works
How many months in the corporate world did you spend before figuring out that money = results and substance takes precedence over appearance?
Buyer’s remorse and post-purchase confirmation: You’re seeing ads for products you bought because the brand or marketing team decided to spend some more money to have you avoid buyer’s remorse and to reaffirm to yourself that the purchase was a good one.
This may or may not work, depending on the person and the product, but some of these things are probably not measured well enough to decide who should see the same ad and who doesn’t need to.
Generally they work very well, better than audience category targeting. To work the best you have to set them up so.
If they are following you post purchase it often (but not always) means the marketing personal have not built a negative audience on a purchase confirmation flag to exclude this behaviour.
It may or may not matter depending on the ad pricing. Some ads you only pay on a click, others on views shown. Obviously the latter wastes more money while if click cost its not such a problem however if volume is great enough to effect your ads likelihood of being clicked it might effect how the ad network displays your campaign.
Basically ads are only as good as the product & implementation. And without trying to sound elitist many of the people setting up and running these campaigns are less technical than they should be and a bit useless at their job. In the same way there are amazing coders vs guys that do the bare minimum pasting other people's scripts as all they can do.
Odds are these campaigns are still making money overall on following the people that didn't initially convert so the companies are happy to keep them running as is.
TL:dr Most likely the ad manger has done a poor setup in remarketing.
For example, I have many pairs of awesome Portuguese shoes, and I originally found this vendor via their Facebook feed ad. That’s over $1000 of revenue over several years from a targeted ad which (I assume) cost them a dollar when I clicked on it.
I’m genuinely happy I saw that ad and found a shoemaker I like without having to research it.
Ads within the FB app still remain much more relevant than the stuff Google shows me on third party sites, which is much more random and often an eyesore. (Yes, I know nobody on HN uses the FB app either. What can I say: I joined at the age of 38 and was surprised that I genuinely like browsing FB because it’s so chill, just old friends and relatives. Nobody is angry or picking a fight, unlike Twitter where that’s 50% of the traffic.)
Without a doubt I'm sure there are some ads that are for decent products, but it feels like a gamble to even consider buying something being advertised to me on these social platforms.
If you find something you actually enjoyed at random, I think that is the equivalent of winning the advertisement lottery.
Basically I don't see how a targeted advertisement ever helps me. I find myself in these scenarios:
- I need a product, say a new pair of shoes. Since I need these, for example in the situation where my current shoes are literally falling apart of inadequate for the weather, I will simply start doing research, asking people, etc. I don't need targeted ads here, and they don't target me based on product quality, so there's no point paying attention to them anwyay.
- I don't need a product: well, then I don't need an ad for it either.
- I'd like a product, say a new luxury pair of shoes I don't need. Well, why should I trust an ad? Upon seeing an ad I'll just go and research it anyway, since again, nobody will list the negatives of their own product in any advertisement.
- I need a product, but am unaware it exists: the only area where a targeted advertisement really makes sense, since it saves me time (I don't know I should be researching it because I'm unaware it exists). But, uh, I've literally never had this happen ever in my life, and I'd be curious if anyone has any examples.
I have no idea why I would trust that company to deliver a product on par with, let alone better, than my current shoe supplying options.
I also wonder if you're in the EU or not, as that changes the story. But I understand a lot of people wouldn't want to say on a forum.
I think I was living in London at the time (pre-Brexit). There’s obviously no shortage of nice shoes in England, but I was new there, so finding a physical store I like was a lot more work.
https://thecorrespondent.com/100/the-new-dot-com-bubble-is-h...
Discussion on HN:
https://news.ycombinator.com/item?id=21465873
this might be an example of the targeting knowing you need the medicine but not that you have it, I recently searched for something I would need a medicine for, I already have the medicine which I got during holiday but the online companies have no way of knowing that, thus I get advertisements for the medicine. Great job targeting indeed.
All I'll say is that they used pretty low down ways to get to people, one example that sticks out was the amount of targeting that went on late Friday, late Saturday night in APAC when folks would stumble home drunk and go online.
Everything and the kitchen sink was thrown at these people, and it often resulted in multi thousand dollar sales, including a few automobiles!
- if the buyer was intoxicated, isn't that sale invalid?
- how is that in any way ethical?
Under what law? Am I not allowed to drunkenly order food at a restaurant? What's the threshold for a purchase that should be invalid?
> what is the threshold...
Far past the threshold is glaringly obvious, far below as well. If it is close, you do what is always done about any threshold question.
However, it's massively, objectively, unequivocally unethical, and part of the reason I quit that place. And for the record, selling crap to drunk people in APAC was probably the least morally suspect tactic this company got up too.
What does the acronym APAC mean?
Asia-Pacific
Sadly, we don't hold capitalism to consent standards