As a rough datapoint, I run a consumer targeted e-commerce site. We ran a campaign before Christmas were we were selling a new product that was only marketed on Facebook, we are certain that (almost) all customers found it though that Facebook campaign. Facebook was only able to attribute about 50% of the sales to the ads, it should have been close to 100%. This then meant that Facebooks estimated CPA was effectively double what it actually was.
Important to note about 60% of our customers are on an iOS device, which is a little higher than the global average but matches the market segment we are in in the UK.
The situation improved after about 4 weeks, I believe Facebook now uses some "AI" to help with attribution on iOS, but it's somewhat difficult to be sure as by then we had other campaigns running.
So, this will definitely be effecting marketers decision making process of where to allocate spend. It certainly made us more courteous about spending on Facebook.
Agreed. For advertisers with larger budgets, marketing mix models are still the only way to understand the relative performance of FB, Google, TV etc. - each of which is a "walled garden" that doesn't exchange data with others.
FB marketing is effective, question is at what price. If those prices drop, ad dollars will flow back. It will take a few quarterly modeling cycles to reflect this though.
The contra-contrarian view is this: FB, Google have an unusual mix of large, medium and small advertisers all bidding for the same inventory. That's what makes FB and Google somewhat immune to large advertiser pricing pressures (and issue of the day spend bans). However, only the larger advertisers have budgets for complicated cross-publisher modeling. If organic FB tools show higher CPAs, it will drive the smaller marketers to other platforms causing some interesting feedback loops.
Exactly, as a small business we are completely dependent on the ad platforms internal attribution tools, if they don't work or can't be trusted we won't use them.
It's unfortunate that the incredably invasive tracking and profile building has become conflated with ad attribution. For us attribution is essential and we have little interest or use for invasive tracking. We just want to know from which ad a customer converted.
Personally we avoid the more invasive remarking tools as I hate it myself when you are chased across the web by a product you have looked at once.
I work in adtech and hear "We don't want to track people, we just want to know what events led to conversion" all the damn time. (Or conversely, from the sell side, "we don't want to track people, we just want to learn/verify our audience's composition.).
Sorry to break it to you, but that's what tracking is.
You're saying you don't want remarking, but - you want to know something that requires marking. "Remarking" is just persistent marking. What you don't want is "retargeting", which is when the user gets to learn someone is building a profile of them. But that's just whether the "marked" profile is used to also "target" - the profile gets built either way.
What do you mean by conflated? The tracking and profile building is how they correlate results with ad impressions. That this data is then also used for training the models and identifying the user coefficients is just a different use of the same correlation data (a website visit is just another type of ad conversion, after all).
We shouldn’t identify the user. We should identify the content. “I want to be shown next to car pictures, because maybe the guy is missing carpets for his car” instead of “We want to be shown to guys 24-30 with interest in cars.” First is tracking the content, second is tracking the user.
Can you not just point the ad at https://your.website.example/item012628?from=facebookad20210... or whatever and save those from parameters in a cookie or something so you can see at checkout what ads were clicked before purchase? How could ATT possible factor in to this?
FB attracts big ad dollars as they will continue to claim to have huge audience reach
>if a user does not have their Facebook and Instagram accounts linked in the company's Account Center, those accounts will be considered as two separate people for ad planning and measurement.
I cant upvote you enough. This single comment contains many of the contrarian view against HN. It is nice we have these real world stories on HN to balance the ideological fight against ads, where All Ads are evil.
It was about the fact that ads were forced, privacy-invasive (due to customization) and generally were terrible overall (think malware/crypto mining risks, terrible UX - think annoyingly flashy gifs, or those gigantic banners in the middle of a scenic drive), not to mention poorly regulated, leading to lots of "double your money/phallus in 3 days!" type of scams.
There are a few "fair" advertising companies (the name slips me) that I am perfectly happy with. A static, discreet ad need not be bad. Several ads are absolute works of art and passion. The vast majority are not.
What we have is (techy) folks wanting to not have a shitty experience, and the average privacy-conscious user not wanting tracking. Companies do not respect (or care enough) for these which is why you have an anti-ad point of view.
(I should probably write about this.)
Exactly. In the old world, ads were attached to matching content, which gave an incentive to produce insightful pertinent content (and develop a reputation in a certain content subject area). Those ads were not much of a problem.
Today, the ad industry tracks the user, and ads are attached to matching user interest, which gives an incentive to produce arbitrary, but addictive content, with most of the benefit accruing to the ad oligopolies instead of the content producers.
also, while ads were certainly dispersed through magazines and the like, a huge chunk of them were really just in the back and not with the main stuff.
Also, it was unlikely you'd see the same set of five ads on every page of every article. Because such a thing would not just be pointless, but counterproductive.
Today's statistically-driven advertising does not relate to the context and thus has no bridge for engagement. We are all desensitized to them, so they just have to block content, flash more incessantly, and make the dismiss click box smaller (or just start putting fake dismiss boxes to trick people into more engagement).
I think the scammy/dark side of advertising is driven by the cost of impressions being so low. This is because of the pricing disparities allowed by the advertising broker tracking attributions rather than the business themselves. Legitimate products suffer as a result, similar to the effect of spam on email.
Like I wrote in another post, most on HN couldn't understand the difference between placement ad on Google Search Engine and Google Ads Network. And suggested we should ban all "tracking ads". And later all ads. I explicitly ask them and suggest what you just wrote. That there could be good ads, and they, by majority disagree. This isn't just on HN, it is pretty much across the whole tech industry. Benedict Evans wrote a lot about this [1] and on Twitter. We even went to ask people offline to make sure we are not in an online bubble. But so far the results suggest otherwise. Especially with Tracking [2].
I think it's just a reaction to the adtech community's total disinterest in user's objections to tracking. They keep invading our privacy and only really consider our wishes when it starts hurting their bottom line for real.
I'm sure that's what inspired my hard line against ads. I block ads and tracking where I can. And I've become so used to an ad-free web that I don't think I'll go back anymore.
If they'd only listened to us this before things got totally out of hand this could have been avoided.
> This single comment contains many of the contrarian view against HN. It is nice we have these real world stories on HN to balance the ideological fight against ads, where All Ads are evil.
I still don't understand your point. The only reason Facebook's ability to properly attribute went down drastically is specifically because of the iOS privacy changes that require user opt in to track across different sites.
Fine, I don't doubt there were other HN commenters who were arguing different meanings of "tracking", but that's all a moot point. In this instance, we know exactly why Facebook's ad effectiveness went down - it went down because they could no longer track all your interactions across the majority of the web. Tough shit.
Why does ability to attribute an ad depend on the ability to track individuals?
IIUC, you can do attribution with a very basic ad: <a href="website.com/ad-34" ping="fb.com/my-product/ad-34/ping"><img src="fb.com/my-product/ad-34/img.jpeg" /> Come buy my product, you'll love it!</a>
Then you sum up all the "/ad-34" hits, and figure out how many of those user sessions (which are now on your own site -- single domain, but perhaps leveraging a script supplied by the ad network) went and actually bought something. The ad network can correlate those sessions with the "ping" it receives to determine clicks vs. conversions.
Not all purchases come immediately after clicking through an ad. They may come days later, even. Knowing that "this user who purchased this product saw this version of the ad" and "this user saw the other version" is what attribution is.
Facebook (and other ad networks) want credit if someone clicks on your ad, spends time on your site, but then doesn't actually buy anything until a couple days when they enter the URL directly (i.e. not clicking on an ad). This is the problem generally referred to as attribution. Facebook (and others) do this by tracking everyone all around the web. They track the ad click and then they track you going back to the website the next day, and they know it's the same person.
Ad attribution directly depends on pervasive tracking.
Allow some facebook supplied javascript to set a first party cookie when the user hits website.com/ad-34, and then reference that 1st party cookie the next day when the user returns. Phone home when they buy something. Facebook can now correlate (1) the ping they received to begin with, (2) the user session cookie they initiated on the first visit, and (3) the user session they observed when a purchase was made.
I'm sure things get easier with 3rd party tracking, but fundamentally you can do it without cross-site tracking.
Out of curiosity, does attribution also work if the ad is not clicked?
For example, I bought a product the other day after seeing a Facebook ad, but I didn't click on the ad itself and rather went searching for it on DuckDuckGo, then after reading a bit about it online, bought the product.
I'm curious as to whether that got attributed to the ad impression or not.
So would I, but I’d also not have a job, or at least make a lot less income. Without marketing, advertising and sales, there is very little economic activity.
It’s like the old adage: this job would be great it weren’t for those pesky customers.
Ads synthesize desire and cause people to be unhappy when the ads work. They encourage expending money that could have been saved, or taking on debt. They also create a funding situation where media producers are beholden to corporations.
The implementation of ads is also wildly invasive, creepy, and propagandistic. However, focusing on implementation allows ad salesmen to lessen the sharpness of the criticism by supposing there is some "nice" way this could be done.
Ads aren't inherently trying to make people unhappy (at least the good ones). Ads in the best case, are just informing a person about the existence of a thing. There have been many ads that I've found useful, not because of some evil mind games, but rather because i didn't know of a company providing such a service.
I don't think there is anything inherently wrong with expending money for things that are valuable. That kinda seems like the whole point of money.
The idea of an ad is that it says you don't have this thing. Your life will be so much better if only you had this thing. In some cases, that makes sense, but mostly it serves to reset people's expectations and become dissatisfied with what they have.
I think in many cases we should be aiming to do that for social fundamentals that would make all of us live richer and happier lives like healthcare and education, but it has a lot of negative consequences in consumer products which are more individualistic, competitive, and status oriented.
While I hate ads, it's impossible to live in a world without it.
Because even at the most extreme, where there are no ads, even the most minimal difference like location or appearance has value. Consider the simple example of a grocery store - what products do they place at eye level? What products do they place closer to the entrance? What color does P&G make their detergent bottle? What do they call it?
It's all "advertising", and there's no practical way to prevent it.
The alternative is that everything is packaged in the same sized gray box and described in perfectly utilitarian terms.
So advertising will always exist. We just need to know where to draw the line.
Ads are a corruption of organic discovery though. Yes it could help you but say person A hocks product 1 because they are paid, even though product 2 is actually superior for you. Ads corrupt their incentives. Even if product 1 is good for you, product 2 is better. Product 1 ends up succeeding because they decide to play the game of corrupted promotion.
First, it's not "all ads are evil". It's the spying and tracking of ads that most people oppose.
Second, how is this a "balanced ideological fight" counterargumnet. Let's say I oppose Facebook ads as invasion of privacy (and I do). A person on the internet used Facebook ads to make money. Those are contrary why? It's not like I'm sitting here and was like "oh, now that someone made money on the ads, I'm totally in favor of them."
> If you have an income it’s because of economic activity, most likely derived from advertising.
That reasoning can be used to justify any shitty practice.
Concerned about global warming? If you have an income it’s because of economic activity, most likely derived from emitting carbon.
Concerned about human rights in China? If you have an income it’s because of economic activity, most likely derived from trade with China.
Concerned about American workers getting fired for unionizing? If you have an income it’s because of economic activity, most likely derived from interacting with sales behemoths like Walmart and Amazon.
Therefore, I suggest you come up with a better rational for bad things. Or frankly stop trying to justify them. That also works.
> I pay for YouTube Premium to avoid ads because I hate ads. But I also like robust economic activity. So, I’m happy to pay a fee instead of seeing ads.
Yep, and I drastically limited watching YouTube on-site because their advertising became untenably invasive. The last straw was when they tried to get me to pay money to back-track how much they made their product suck, rather than actually providing me value over the out-of-box experience a few years prior.
Likewise - I am not entirely opposed to paying for Google products. I pay for G-Suite, for example (for now). However there is not a hope in hell I will ever pay for YouTube - they go out of their way to disable native OS features like Picture in Picture video, and the experience of ads is so shitty I’ve just added it to Pi-hole and moved on.
There’s some good content on there, but it’s not good enough to subject myself to that, or to reward such behavior with money.
Ads are evil. To work, they play upon human insecurity. In many cases, they create desires or perceived needs that weren't there before, making them wasteful in addition to scammy/conniving. Because they're successfully evil - the tricks work - they often leave a residue on people: beyond just never being able to forget a jingle from a cereal ad you heard when you were 8, a lifetime of very frequent exposure to advertising trains us to suspend our criticality, or hinders us from developing it in the first place. So it leaves people dumber, too, more pliable and dependent. Hilarious in a country like the US, whose national mythos is so obsessive about personal liberty and rugged self-reliance.
We have a family friend, retired now, who had a successful career in marketing and strategy in multi-billion dollar transnational companies. I once asked (probably naively) why [maker of extremely popular product at the time] ran so little advertising. The friend told me "they have a good product. They don't need to spend money haranguing people into buying it or spreading the word about because it's actually good. Word of mouth is free".
Governments are evil. To gain power, they play upon human insecurity. You can say this about anything providing a service. Its a wild oversimplification of the psychology of human buying decisions.
While there are plenty of people annoyed by any ads at all or who believe any form of marketing is inherently evil, that is not at all a majority view. Yet almost everyone hates the way Facebook does it except sellers. The problem isn't the ads. It's the level of surveillance required to make them as effectively targeted as they are. People don't want everything they ever do to be recorded, catalogued, and studied to build a psychological profile of their global purchasing habits.
And that’s the key issue: constant surveillance coupled with machine learning used to identify our desires and appetites.
It’s creepy for one thing. It reminds me of why we sometimes have to let go of a certain kind of relationship: they know too much, and are too effective at getting you to buy into action that is not ultimately in your best interest.
Well no one's arguing that targeted ads are not effective. This is just an example of how well it works, so upvoting it may not be the best way to bolster your side of the fight.
>Well no one's arguing that targeted ads are not effective.
Oh they sure were. All ads are useless. Targeted ads are evil. And Ads should not be targeted were the HN's view, which later became there shouldn't even be any ads, it should all be subscription. That is 2018/19. By 20/21 HN were even targeting those who were working inside the online Ad industry. And this is seen across the tens of thousands of comments. ( Say 50 Thread of 200 comments )
It became such a problem that people working inside Ad industry were even afraid to post their views on HN. And they have to be upvoted to make sure there are still sanity inside the HN community.
Arguing that "ads are evil" is different from calling them ineffective. Most people dislike targeted ads them because of privacy/resource usage issues. Other people dislike ads in general. Some of those want ads to stop existing exactly because of how effective they are.
I don't think I've ever seen someone say this. People hate that they make the internet impossible to use. I have a family friend that runs a small business and all of the revenue comes from ads on the websites they run. I hate using their website because of the ads.
Pop-up that takes up the whole screen asking to subscribe, auto-play video ad, content moving as ads get loaded slowly. It's all just awful to use. The ads are absolutely effective though.
Ublock origin and pihole make the internet a more pleasant experience for me. It has been suggested that I should feel some nagging voice causing guilt that I do not support the sites and apps I use that depend on ad revenue. Perhaps there should be, but there is not. Apple's position on this is just responding to demand. When wifi access points start offering built in enabled pihole by default those access points will sell well. Throw in a vpn to pipe your phone through it from wherever that is easy to use/automatic and "ads are useless" may become truth.
Edit: it would likely just cause a shift to self hosted ads, which is a dramatic improvement imo.
Self-hosted ads COULD be a dramatic improvement, if the companies start vetting these ads in some way
My guess is it'll result in backend web modules that make it easy to automatically dynamically re-host the same virus- and tracker-infested malware ads we all hate
This is a side-effect of media selling advertising through brokers rather than direct, that these brokers then built tools to track impressions, clicks and sales attributions, and then created a disparity between each level of several orders of magnitude.
When advertisers were paying an order of magnitude more for ad impressions, we needed far fewer ads and those ads tended both to be of higher quality and more topically relevant.
At the current CPM, I routinely get ads which aren't spelling or grammar-checked.
I have no problem with targeted ads based on what I specifically told Facebook. It’s when I’m shopping on Amazon and see the same products advertised on FB that there is a problem.
For the longest time I was getting Amazon ads on Instagram that were for products I had recently viewed - even viewed less than an hour before in some cases. Moreover, the Amazon account we use is my wife's, which we pay for Prime on. I would look at something on my phone, which is logged into her account, and then see the ad show up in my Instagram feed that day or a few days later.
Now, I get 90% ads for car parts and car accessories. I don't have a car, I don't drive, I don't have a license, and I don't even like to be in a car, but for whatever reason Amazon (or Instagram?) ads are assuming that what I really want are car parts and car accessories that I don't recognize, can't use, or don't understand.
It feels very stupid to me that those are the "default" ads I'm getting, even though I never tap on them for obvious reasons, but it's reassuring to know that they went from knowing exactly who I am and specifically what I had been looking at to having no clue whatsoever anything about me, or who I am or what I like.
The real world story is that marketing in one channel is a little worse now... I don't feel like that justifies the existence of data-mining to show ads.
> Did you use a Facebook-specific link in the ads?
I do sometimes find it odd that such a thing is insufficient for tracking sources of traffic in and of itself. No doubt there is a complexity that I have missed
That is almost exactly what UTM url parameters are [0]. We use these and via them our other tracking collaborates our theory on Facebooks tracking.
There is actually a real problem with tracking via cookies on Facebook ads when the destination is a website. The ad click will open in a Facebook "In App Browser", any cookie that you (or any analytical service) sets will be within that IAB. If the user then uses the "open in Safari/Chrome" option that tracking can be broken as there is no cookie. Ideally you want your visitor to either complete their transaction within the IAB or to use the "open in Safari" option immediately so that any tracking parameters are copied to the other browser allowing the cookie to be set.
In our case the majority of our customers will have a better experience outside of the IAB and so we have a popup that prompts them to use "open in Safari" before navigating away from the first page view. We actually implemented this after noticing a very high drop out rate for iOS Facebook IAB users during our checkout. What was happening is address/payment card autocomplete isn't available within the Facebook IAB and people were clicking "open in Safari" during the checkout in order to use it, they would then find themselves with an empty shopping cart, hence the drop out.
So not only is the Facebook/Instagram "try to lock users into our own browser to track them" thing user-hostile, but it actually harms companies that are using ads by having them drop in the middle of the funnel?
That's really interesting, I'm going to check our analytics to see if the same thing is happening.
Another way you could potentially get around it is to fingerprint the user and store the basket contents server side and present them to that fingerprint.
This is interesting. Are you saying because some user switch to safari during checkout and tracking lost here? Since cookie will be cleared. But how about these special query param as you mentioned? That still can be maintained for tracking?
UTM parameters can be identifying but not generally considered private - rather the opposite - so while you can link your cookie to UTM parameters and join with other UTM-keyed traffic data for useful results, you can't do the inverse and use knowledge of UTM parameters as sufficient proof to recover a session cookie.
What is the shop? I partly live in the UK, don't run iOS, don't use Facebook and block ads about a billion different ways at once. There's a chance I might be interested in your products.
But then again. If it was an interesting product, Facebook would be the last place on earth where it’s advertised. Don’t know what it is but I’m predicting it sucks.
> I believe Facebook now uses some "AI" to help with attribution on iOS
Is that even legal? With AI, you can never concretely prove anything, so you then have Facebook literally making up numbers it pinky swears are legit and billing you accordingly (semi-directly because while you're not paying per click, the landing page analytics are also used to weed out robo clicks and other fraud that you shouldn't be billed for).
I am not sure it's ethical, but I don't think any laws would be broken - if their customers don't like paying for imaginary services, well, they can go elsewhere... or can they?
If you're so certain, it means you already have reliable data for attribution. Why do you need Facebook's tracking to confirm it to you?
I understand it's nice to have fancy reports etc but it sounds like you already know where your customers are coming from. And tracking is very invasive, in this case it doesn't really seem to add much value.
You could also ask your customers "where did you hear about us" for example. Perhaps you already do and that's the source of that 100%. If not you might even discover a way you didn't know of. Eg word of mouth, some obscure forum where your product was mentioned. As well as that it's a method where you respect your users' privacy.
> If you're so certain, it means you already have reliable data for attribution. Why do you need Facebook's tracking to confirm it to you?
We were in unique situation with this one campaign where, quite right, we knew where our conversions were coming from as we were confident there was no other source. Within a few weeks we had other campaigns running and no longer could have the same level of confidence.
The point it outside of this unique situate of a new product only marketed on Facebook the conversion attribution on their platform is broken. This is affecting marketers decision making process and reducing the spend on that platform.
The interesting thing is we were able to see the affect of Apples change quite clearly. Most people are not able to see that.
You attribute 0% of purchases to repeat loyal customers who just check in on what you're selling every now and again without needing to be reminded by a marketing campaign?
Important to note about 60% of our customers are on an iOS device, which is a little higher than the global average but matches the market segment we are in in the UK.
The situation improved after about 4 weeks, I believe Facebook now uses some "AI" to help with attribution on iOS, but it's somewhat difficult to be sure as by then we had other campaigns running.
So, this will definitely be effecting marketers decision making process of where to allocate spend. It certainly made us more courteous about spending on Facebook.