Hacker News Reader

Comment by lumost | Hacker News Reader

There are a few factors at play here when it comes to advertising.

1) Digital is taking over all ad-channels

2) Digital ads can be tracked and attributed ways that were never possible with traditional media

3) It turns out most non-targeted ads aren't actually that effective, and aggregate ad spend is falling.

FB ads are effective and high value because of tracking. If tracking goes away, FB ad revenue will fall to that of banner advertising at roughly 1/100th to 1/10000th the price per impression that FB currently receives.

Such a change would effectively force a rethink of the entire consumer technology business, and see FB/GOOG re-structure to some form of subscription revenue or shutdown large feature sets and product offerings which would no longer be tenable.

View on HN

> FB ads are effective and high value

Not really.

Digital advertising, frankly, sucks from the advertiser's point of view. It's just that people are online now and advertisers have no choice: the inventory providers are monopolistic and unregulated.

A more fair comparison might be that all advertising sucks from an advertisers perspective... but we just didn't have a good way to measure it.

In the world of Nielson and print the only consistently measurable ads were brand ads - and brand awareness had a loose correlation across businesses with revenue. There was no way of rigorously testing whether individual customers propensity to purchase a good from a given business was actually tied to any of the brand awareness ads that were being purchased, or customer's brands they had heard of for trust reasons after deciding on a purchase, or if customer's simply didn't care about the branding. Bear in mind, any negative datapoint that indicates customer's didn't care could always be taken as a sign that the "brand" had to be improved. Not that the ads were worthless.

Now on digital any brand exercise, direct pay per conversion, or other form of ad gets attributed to downstream revenue, and we're quickly discovering which half of the marketing budget isn't working.

> There was no way of rigorously testing whether individual customers propensity to purchase a good from a given business was actually tied to any of the brand awareness ads that were being purchased

This isn't right. Of course there was - statistics, sampling and A/B testing was invented way before computers or the Internet existed.

Advertising is just applied sociology, and sociology is just statistics when done right.

> Digital advertising, frankly, sucks from the advertiser's point of view.

What advertising does not suck from the advertiser's point of view? How are metrics gathered that demonstrate these forms of advertising work?

> 3) It turns out most non-targeted ads aren't actually that effective, and aggregate ad spend is falling.

While this is a commonly held believe in many corners it doesn't appear to be well backed by data & research. Freakonomics also covered this topic in some detail recently.

Some examples where it's proven untrue:

Nytimes drops targeted advertising in Europe: https://digiday.com/media/gumgumtest-new-york-times-gdpr-cut...

>“The fact that we are no longer offering behavioral targeting options in Europe does not seem to be in the way of what advertisers want to do with us,” he said. “The desirability of a brand may be stronger than the targeting capabilities. We have not been impacted from a revenue standpoint, and, on the contrary, our digital advertising business continues to grow nicely.”

Danish broadcaster grew revenue after dropping targeted ads: https://techcrunch.com/2020/07/24/data-from-dutch-public-bro...

>The data shows the NPO grew ad revenue after ditching trackers to target ads in the first half of this year

Typically, platforms with smaller reach ( read < 10 million DAU) will only have a few big advertisers who are willing to purchase most ad positions. Behavioral targeting on such platforms with large advertisers will probably only limit spend by cutting out bad impressions, and the in-house or external behavioral targeting isn't anything close to what FB offers.

FB/Goog dominate ad-spend as they are able to effectively match small advertisers with the appropriate audience, and these advertisers typically have specific customer goals in mind. If the goal of the ad buyers on the NYT is simply to associate their brand with the trustworthiness of the NYT, then there won't be much benefit from behavioral targets.

> It turns out most non-targeted ads aren't actually that effective

Non-personalised television advertising is still doing great.

No, it's not. Eyeballs are down, thus prices and market share are down compared to digital. Different industry sources have different figures, but everyone agrees TV spend was down >5% in 2020 and expected to drop faster in 2021 as digital continues to grow at >30%.

There's now a mad rush to standardize tech for personalized tv advertising, as providers plan to use the email/phone number you used for your cable/streaming service as a personal identifier that can be joined against 3rd party targeting data.

I'm not trying justifying this - just describing where the industry (and money) is going.

Is it not right that 2019 was the highest spend ever for TV advertising?

And a 5% drop would be less than the drop in broader TV production, via coronavirus?

Major events that are associated with large advertising budgets haven't happened: the Olympics didn't happen. The Eurovision Song Contest didn't happen.

How much does Facebook earn from one user, $1 a year? No one would've died if they switched to a $1/year subscription model. People tend to respect companies that are honest with them. Except Facebook won't ever be able to regain its reputation, but still.

https://mondaynote.com/the-arpus-of-the-big-four-dwarf-every....

Depends on the metric you use.

> Facebook’s revenue per user is roaring. For Q4 2018 vs. Q4 2017, it’s global ARPU increased by 19 percent to $7.37, and the US-Canada ARPU by 30 percent to $35. Between 2011 and 2018, the social network global ARPU rose nearly 6x, while the US-Canada grew 11x. With a domestic revenue per user of $112, Facebook is the equivalent of a publisher charging $9 a month. It does that with a free service.

TIL. I thought these kinds of numbers weren't at all possible with an ad-supported business model.

Assuming that the only users who would be willing to pay are those who use one of FB's services at least once per month; then FB would need to charge $26/yr to match their current revenue. Given that they already have 2.7 billion MAUs it's unlikely that they could grow their way out of such a change, and serving ads is ultimately inexpensive.

In Q4 alone, they made $53/user in US/Canada, and $10/user worldwide. This revenue is not evenly distributed, and the type of user who is willing to pay $50/quarter for FB is worth much more than that to FB advertisers.

source: Slide 4 (ARPU) https://s21.q4cdn.com/399680738/files/doc_financials/2020/q4...

Doing that would equate to $2.7 billion a year in revenue. They currently generate about $70 billion a year. So, yeah; they'd pretty much die if their revenue dropped that much.

Imagine how much money they could save if they stop redesigning every little thing all the time to squeeze every last cent. Imagine not needing hundreds of developers working on a single app.

Reply on news.ycombinator.com