[shoguning]

My interpretation of what's going on:

CEOs finally have leverage against their marketing departments to do black-out testing on social media ad spend. It's difficult to actually measure impact of ad spending, and marketing teams spin results to their advantage.

They will observe what happens and re-introduce spending where it actually works.

[Aperocky]

When a recession happen the first to go is marketing.

As a test, you would need to control other variables, the present economic uncertainty basically rendered that impossible.

It's just the CEOs finding a convenient excuse to cut costs.

[adrianb]

It's an even better deal: companies are replacing paid advertising with free exposure from being mentioned in news stories and bonus points for "standing up for what's right". In addition, plenty of people (and the news organisations themselves) share these news stories on the very same social media, with 0 costs for the companies involved.

Overall it means increased brand exposure for less money.

[will_pseudonym]

Good ol' public relations.

[jjohansson]

> When a recession happen the first to go is marketing.

As a marketer, I have found the opposite to be true for some industries. For example, during 2007/2008 my digital marketing agency saw a big surge in business from companies who saw marketing as a way to grow lagging demand.

[Aperocky]

I guess it also had to do with cost effectiveness. Digital marketing might have been small but cheaper than paper/cable medias. Right now given the size of Google and Facebook (and the fact that their whole revenue is basically ads), it seems the advertising spend on them has become excessive and not cost effective.

[sheeshkebab]

Everything “non-core” usually gets cut... it depends I guess whether someone views marketing “core” or not.

[efdb]

Zuckerberg: "We're not going to change our policies or approach on anything because of a threat to a small percent of our revenue."

Facebook Mission: "Give people the power to build community and bring the world closer together."

Please update the mission statement

[nostromo]

2020 is very atypical and not a good time to run a "marketing off" test.

Even if you did an A/B or multivariate test across different markets, you'd probably be measuring local responses to Covid more than you'd be measuring Facebook marketing effectiveness.

[politician]

On the other hand, with stay-at-home orders media consumption is at an all-time high. Ads should outperform in this environment.

[CaveTech]

Engagement rates are way up, but conversion rates are down. Overall a net benefit in my line of business, but it's affecting other segments different ways.

[mmm_grayons]

This assumes your conversion rate from media is the same. It's reasonable to hypothesize that most are significantly more reluctant to spend money in this environment.

[politician]

Agreed; it depends on the product.

[lalos]

With that same argument you can conclude that since 2020 is very atypical then it's not a good time to run the usual strategy of last year. Might as well switch it up, cut costs or react (in case they saw that ad clicks went up, but not the actual sales).

[garmaine]

It doesn't matter. The test is "should we be spending money on this in 2020", not testing a hypothetical situation where the pandemic doesn't exist.

[pphysch]

I doubt this very much. Doing black-out testing silently would be just as easy and allow a much greater degree of experimental control.

[wand3r]

Also, it is being done during one of the craziest markets and climate ever to occur so it would be difficult to really get good numbers.

[ramraj07]

Well if it fails, and revenue drops, that would be hard to explain to shareholders right? Especially if the test has to extend for months to be realistic. Now they can at least say it's for a social cause.

[pphysch]

Just blame it on the 'rona or geopolitics.

[shoguning]

Maybe?

Part of the leverage is going public and saying it's about some principle.

For experimental control, executives tend to think simple is better. The more complex, the more room for fudging.

[usrusr]

Yes, I think it's a fault line that has been waiting to crack for quite a while. Facebook has never been a good match for those big household-name brands because for them targeted ads add very little value over blanket ads. They begrudgingly pay a heavy premium on targeting only because some demographics have completely detached themselves from blanket ad media. That's not how a happy customer relationship looks like.

Marketing departments (or external intermediaries) don't mind the lower value proposition at all, because they get all those micromanagement opportunities and (perceived) necessities offered by self-serve targeting platforms. Huge opportunities to look important. I wonder how the management overhead per dollar spent on ads has developed in the last 20 years.

[emmp]

It's not exactly blackout testing if the blackout testing itself is highly visible and gets positive earned media.

[spootze]

Facebook is actually one of the few platforms (if not the only platform?) that allows running self-serve RCTs (conversion lift tests in their lingo): https://www.facebook.com/business/help/688346554927374

Conditional on them honestly reporting numbers, it's a pretty credible way to measure effectiveness in terms of incremental conversions you get by showing ads on FB/IG

[hellomyguys]

>It's difficult to actually measure impact of ad spending, and marketing teams spin results to their advantage.

Have you used any self-serve ad tool? It's quite easy actually.

[sokoloff]

It's quite easy to get a measurement on your screen. That doesn't necessarily translate to measuring the impact of ad spending.

If I have social media ads, paid search, and retargeting ads all live on multiple platforms, there is undoubtedly interaction among them and the quite-easy self-serve ad tool is likely to show me a simplistic "last click attribution" at best.

[shoguning]

Exactly. For really large companies with many channels, it's very difficult to disentangle this.

Here is a good discussion of the difficulty: https://thecorrespondent.com/100/the-new-dot-com-bubble-is-h...

[willcipriano]

Often on Google I see ads right above a organic link to the same site. I wonder if marketing teams count the ad click as a "win" when in fact I was looking for the company because I learned of it via word of mouth.

[sokoloff]

I’ve always seen that considered to be “paid search-branded”, which is tracked but is tracked and treated separately (for the reason you said).

[jjohansson]

Most analytics platforms will report on any straight direct line between spend and incremental revenue. The hard part is measuring the indirect part, as you mentioned, but it is possible.

[TeMPOraL]

It's easy to bullshit yourself into thinking your ads are much more effective than they really are. It's even easier to bullshit the people who pay you to do ads for them. Advertising and analytics platform happily oblige both use cases.

As 'sokoloff wrote, it's easy to get a measurement on the screen. But gaining real insight from it, or even knowing whether that measurement makes sense, requires a certain amount of analytical thinking and knowledge of statistics. Maybe among corporate marketing departments such skills are plenty, but in my experience, it's not usually the case with small marketing agencies serving smaller companies. Bullshitting has a better effort/reward ratio, particularly when your customers understand statistics even less than you.

[sanxiyn]

Let's See Your Treatment Effects!

https://www.youtube.com/watch?v=xd3gQAV1ncs

Run an experiment And let the truth fall out

[fullshark]

My interpretation: old media doesn't understand new media.

[JMTQp8lwXL]

The results would be tainted by the free advertising brands get for boasting about cutting ad budgets.

[neonate]

That is an ingenious interpretation.

Probably too ingenious though, especially across all the companies involved.

[fourstar]

Bad take. These companies are using FB as a scapegoat to reduce ad operations during this pandemic/company budget reductions.

[shoguning]

Seems synergistic with what I said, no? Reduce budget and do some testing at the same time.

[fourstar]

In what world do you think CEOs don't have power over their marketing departments?

[shoguning]

On paper, of course they can do whatever they want. This just gives them more cover to do it, shielding them (at least partly) from internal blowback.