[mparr4]

FWIW I run a D2C brand and the iOS 14 change hurt badly.

We are able to attribute something like half of the sales we used to. The main consequence of this is the FB models can't learn who our customers are as well, resulting in an increase in our cost of customer acquisition.

It makes selling niche products much harder.

[samwillis]

Similar position here, although we recently launched a new product that so far is only advertised through Facebook. Based on that we believe Facebook can only attribute 40-50% of conversion events.

Facebook is going to have to do something drastic to improve this situation, I suspect they are going to push advertisers to fold their store/checkout into Facebook so a customer never leaves the app. It will be the only way to track all conversion events.

[Tagbert]

Then I expect you will lose more customers who object to that practice.

[mensetmanusman]

These changes make it harder for local businesses to advertise directly, so it strengths the ‘winner take all’ amazons of the world.

Interesting feedback loops here that might not have been predicted.

[lanza]

lol Apple knew damn well what would happen. They are a business. This was a good business move for them. Being able to advertise it as privacy just boosted it to a fantastic business move.

[phh]

Am I understanding correctly that your issue is merely about attribution, and not that advertisement is actually less useful?

[ruhdgjns]

It goes hand in hand. First, if you can’t measure conversions and tie them back to advertising, then how can you know if it’s effective? Second, if you can’t measure effectiveness how to A/B test your approach? Finally, without labeled data how you train ML models?

The whole house of cards stood upon attribution.

[phh]

> First, if you can’t measure conversions and tie them back to advertising, then how can you know if it’s effective?

I don't think Coca Cola needed attribution to know their ads were effective in 1900, maybe we should ask them?

> Second, if you can’t measure effectiveness how to A/B test your approach?

You can still use either time, geographical or UA-based AB testing without requiring attribution

> Finally, without labeled data how you train ML models?

Are you saying the final objective of tracking is actually to feed AGI as most data about our world as possible? That's an interesting point of view, I never thought about ML modeling as an end-goal.

[ruhdgjns]

I don’t think you’re engaging in good faith. You can answer your own retorts if you think about it from the view of a modern advertiser.

[phh]

I agree that I'm being snarky, sorry about that.

However, I do think your points are moot, and my real understanding of targetted advertisement is that attribution is required only to steal the attention of the people paying for the ads, because unique attributions means you can send new numbers every single day, or even make a new mail at every conversion.

So in my understanding, losing attribution only reduces ad buyer's attention, but it doesn't actually decrease effectiveness.

Edit: But I totally agree that I have a very poor understanding of advertisement, so I'm genuinely interested in hearing more

[AnIrishDuck]

In the advertising world, there are roughly speaking two major objectives:

- Creative: also known as "brand awareness". Keep the given brand top-of-mind for as many prospects as possible. The hope is that they will then choose products under the given brand at some future opportunity. This is the type of advertising that Coke ads are. They aren't attempting to get you to buy a Coke right now; they are trying to brand "buy coke, it makes you happy" on your brain.

- Directional. These ads are trying to get you to do something as soon as possible. One obvious example is the annoying "you just looked at this thing, still want to buy it?" ads that follow us all around online. Presumably these ads are somewhat effective, otherwise they wouldn't exist.

From this perspective, removing attribution barely effects creative advertising. On the other hand, it is obviously quite destructive for directional ads. Whether this is a good thing is arguable; I personally find the desired objective of "creative" advertising to be more disturbing.

[titusjohnson]

I think the argument is that _you_ are not engaging in good faith, as you are assuming that the _views of the modern advertiser_ somehow appear on anyones list of things to even consider.

The individual you are replying to is giving you good indications that all the things you are worrying about _are not necessary_ for you to do your job. They're a crutch. Go back to the fundamentals.

[mparr4]

I don't agree with you that ruhdgjns is not engaging in good faith. He's engaging with my comment. Given that I am a D2C business owner and therefore a modern advertiser, that's a pretty reasonable view for him to consider.

Re: "go back to fundamentals" and "not necessary", I wish it were so simple. In my comment I mention that it makes selling niche products much harder, which is absolutely the case. Some businesses just can't exist without good targeting. Maybe we as a society don't want those businesses to exist, and that's fine, but it's not as simple of a matter of these businesses needing to be better at advertising. With good targeting a business can exist that only has a small subset of customers in the world. A business like this likely can't afford to acquire customers via traditional advertising.

This makes many D2C businesses much harder and a subset of those businesses won't be able to succeed at all in this climate. Again, I'm not passing any judgement on whether that's a good thing or a bad thing for society, just that it is the case. Good targeting is absolutely necessary for some businesses.