I'd be curious whether the conversion rates on the marketing side back up my gut instinct on the consumer side. Any marketers here with numbers?

Numbers: low nine-figures spent over the last 7 years or so on RTB display and GDN/YT

Outcome:

* Google on display tends to be much higher quality than most major DSP platforms. Incremental ROAS on Google's non-search ads tends to be better, with YouTube being the best performer.

* When you use GDN and YouTube, you are bidding on a much more accurate set of behaviors and first part data in addition to the content matching, which is second to none for large scale.

* One of the better audience types Google gives you access to is search intent. You can target people that have exhibited search behavior that indicates their interest or likelihood to purchase what you have on offer

* Despite these advantages, like any platform Google can also have major quality issues if you do not manage your campaigns well and click fraud can happen easily (thankfully in my experience they're decent at detecting and refunding fraudulent spend)

Edit: I would caution people commenting here to not use their own behavior as examples of how an ad platform is/isn't effective. HN users tend to be significant outliers vs. the general population.

Self-reported behavior is also extremely unreliable, and I would take all their comments with a grain of salt.

1. Self-referencing in my last point refers to gauging your own response to ads as a barometer of their effectiveness.

2. My points above are based on spending at scale globally (over 70 countries) for B2B and B2C products with multiple avenues to monetization

3. The outcome of my experience is based on this data in aggregate but I do not assume that past behavior indicates future success. The market is changing rapidly.

In other words, exactly what the top comment is doing.

Sample size=1. Methods=unknown ("seem 2-4x better")

That said, we could apply the same principle to the parent comment from the marketer. He is using his own behaviour as an example of how the Google ad platform is/isn't effective.

There is a serious lack of transarency in the "open market" system Google is both overseeing and participating in. In the long run, that opacity could spell disaster, so says a former Google lawyer.

This opacity is a problem that seemingly inconsequential litigation like this from the Texas AG is (unintentionally) helping to address. Discovery helps us understand what Google wants to hide from the public.

Move away from the big-boy ad networks (The likes of Microsoft, Google, Facebook), and there's a long tail of networks that will happily run the worst ads. Unsurprisingly, a lot of legitimate companies aren't super-keen on having their ads run next to ones for fake download buttons and penis enlargement pills.

Some ads are legitimately quite useful, others just abusive, lots in between.

I worked in the newspaper industry on the web side for many years. Remnant ads are the worst. These are the ad networks that you use to fill up the last of your ad spots that you weren't able to sell directly. They can fill volumes of ad spots for pennies per 1000 ad impressions. These are pure trash. Ignoring the common problems with actual virus payloads that they might try to inject, these are the most obnoxious, awful ads. They often have big animations with flashing background and jarring colors. They are full of click-bait titles. They are the reason people install ad blockers. I might not be able to tell you what a quality ad looks like, but I can tell you the other end of the spectrum, and it is pure filth.

Making sure clicks are real is a massive challenge with many ads networks. There are a lot of tools that help in sure spambots aren’t stealing all the ad money but they also slow down serving of the ads and sometimes they mess up and block things that shouldn’t.

We’ve had campaigns where we had to allow more spambots simply because the client couldn’t spend enough money, they were willing to waste more to get better reach.

But most clients absolutely love the results when you have really good filters in place. Very low spend and good lead generation.

For an advertiser side, one possible, measurable metric is advertiser value per spending money. Value can be defined by advertisers pretty arbitrarily but you can think it as something like how much money you earn from ad conversions. Of course, this is very hard to measure accurately, massive ad network like Google has some advantages though.

But note that this is just one of possible metrics, there are hundreds of metrics ad networks want to improve and some of them are conflicting each other by definition. And some of them don't even have an objective way to measure so still depending on user study.

Compare to most small news sites, that have top/bottom banners, inline ads and the curtains.... I hate the curtains!

Add to this how little time I use on Facebook vs Google properties and how I used to love Google and despise Facebook it becomes even worse.

Maybe Google just failed spectacularly for a decade and everyone else had fantastic experience but I doubt it.

It's amazing their incredibly terrible advertising actually funds a huge breadth of services. Marketing departments really are that dumb I guess

Facebook has the best direct response ROI (e.g. a donation, a volunteer, a tracked conversion to voting).

I used to be able to get > 100% ROI immediately but with iOS changes it's becoming hard to impossible. I think FB will become more of a brand ad and loose a lot of app install or direct to consumer business.

Those are what most people consider relevant, for instance i get a ton of athletic shorts ads after I hover on one (usually because they use sexy images lol...).

Google has great DR ROI for search, though if your website is already #1 is that valuable? Big brands have to pay so their competitor doesn't get the slot though.

Google video, e.g. YouTube is very high quality and low CPMs compared to say Hulu and FB video 'view optimization' sucks like they count a 3 second scroll pass and the total complete watch % are absolutely tiny.

If you spend big (like 10k a day i think) you can access google's surveys which show a good amount of recall.

I've done my own surveys I built mobile survey interstitials and retarget people who watched video ads. Generally see a few point % bump in head to head voting (e.g. vote candidate X over Y) and actually a really high amount of recall candidate names with very few picking the fake spoiler. Again though with iOS this becomes hard/impossible.

Stories ads are valuable to me I love a digital native ad that fits the format.

Open market RTB video is mostly shit. Unless you do a kind of hybrid called a private marketplace deal (and a lot of other acronym buzz stuff) which gets you the high quality inventory but it's just an easier way instead of reaching out to each of the sites individually. Most websites are either auto spam content generation or layer so many auto play videos with sound off you're just wasting money. And they can lie in their bid saying it's a large size or audio on. Have to then spend $ and time on fraud prevention.

Open market desktop display ads (banners) are scraping shit off the floor usually IMHO. And then open market stuff like Taboola or when a website puts a link farm at bottom of page are like rectum licking in the scale of quality, to extend this gross metaphor.

Snap/twitter are interesting but haven't provided a ton of audience/value for me in my campaign audiences. Twitter is cool like targeting specific handles - havent done it in a while maybe not able to anymore with iOS idk. Snap is cool to try and get younger people IMHO.

Mobile interstitials can be valuable. I've built fun 'playable' interactive ads but I don't know of any other campaign or political vendor who has done something that is truly mobile native. They mostly just have static banners.

examples for game install ads they show previews that are playable. I like to think of interactivity that works on mobile to pull a voter into the ad experience. Twitter is selling mopub maybe? That's a higher quality network for this inventory.

Connected TV is great (a bunch of acronyms and definitions/sources, but basically anything steaming to a tv). Though it's insanely expensive during the election because all the TV buyers are fighting to buck viewership trends and selling campaigns equivalency - which is fairly valuable.

all the other corporations in the world are spending more on digital as broadcast TV dies but the incumbents in the political space have too much power and money at stake to follow the eyeballs. Most only spend like 15% on digital which is INSANE to me. A 'good' campaign i've never seen spend more than like 35%... IDK maybe bloomberg got to over 50% which is where it should be if you includ CTV.

TV people usually win because historically the TV vendor is the main consultant of a campaign setting message and direction.

Instead of a digital firm buying/owning as a whole package which I think makes more sense.

How do you measure this with reasonable certainty?

and then they black box tie them to impressions/ad events.

It only works with early voting, where you get a list of voters who have cast or sent in mail ballot each day.

though arguably you're already targeting a tight list in FB of likely early voters (who are likely to vote themselves even without seeing an ad) so who knows if the ad actually got any extra votes in.

You could do an AB test, randomly show ads to only half the audience and measure lift but there are only a couple weeks of early voting usually and relatively small audiences unless you're statewide. I imagine it would work better for Presidential i've never worked on one, kind of a bucket list dream but also would suckkkk for a few years

You can also upload a final voting list after the fact as offline conversions, but doesn't make much difference after the votes are counted though I guess cool stats

this sounds like a good thing for facebook users

Though I do see the point that some apps relying on addicted 'whales' making micro purchases. FB is (or was before iOS started crippling) really great at targeting and optimizing to acquire those best revenue generators no matter if you're getting donations to a non-profit or running an app casino.

But this has nothing to do with search ads. It's about display advertising on 3rd party sites.

Google is the big bad competitor who you're trying to out-compete.

You can't out-compete google without going below it in price. Everyone is desperately trying by charging FAR less.

I just want to note, it doesn't necessarily mean google is overcharging, but that nobody can enter a well established market by charging more and providing no other real value.

I thought these advertisers paid per-click, so in some sense Google and any competitor are fungible. Do people really care where their ads are shown so long as people click on them? Are the competitors placing ads in those crappy web sites that encourage accidental clicks?

Of course they care. They pay per click, so they want the ads well-targeted so they have a high conversion rate, otherwise it's wasted money. Some ad networks literally scam businesses by placing ads everywhere so they get clicked on accidentally but of course this does nothing for conversion.

Perhaps temporarily, but search has been stagnant for 5-10 years.

Alternative smartphone app stores charges way less than 25-50% of Google or Apple. You need to compare electronic markets with electronic markets like this, and the big players tend to take out extremely high fees in those areas.

Although, I agree that charging high fees in these areas is a problem and should be fixed, but it isn't like this sort of thing is uncommon.

Edit: As an example, Apple takes 30% cut. A competitor taking 7.5% cut would cost 25% of Apples app store.

I guess those various people need to prove that they are needed so they make changes for the sake of doing something.. and quality of the product declines.

You're the person seeing the ad, but you're not party to the transaction.

Websites often run an auction at request time. The ad networks are actually competing, even if you get targeted by Google exclusively, personally.

A bit part of US antitrust law is harm to customer. And higher prices is a harm to customers.

I assume the argument is something like “Google is monopolizing search and charging higher prices as a result”

situation: A has 0% take rate, B has 1% => "B charges infinitely more than A"

Perhaps better in that case would be just stating the facts "A takes 0%, B takes 1%" or "A leaves you 100% , B leaves 99%".

The point is, from a consumer pov the take rate actually doesn't matter as much as the opposite, the "keep rate". And just adding in multipliers is so easily exploitable.

Not defending google, just as written the implied math has multiple interpretations, some okay some bad.

If I understand this correctly, this is about auctions/exchanges (middlemen which have little to do with ad targeting and quality), not ROI, as some commenters have suggested.

Correct me if I’m wrong, my adtech knowledge is limited.

Assuming people who dont know how ads industry works all they see is an article about an evil monopoly over charging.

While it might be hard to break up any given monopoly that offers "free" products with existing laws, this is the avenue that is straightforward.

(I don’t use the products and I’m sure the names have changed, but dv360/dcm, whatever they call their exchange, and dfp)

I’ll leave the morality/legality/etc of as an exercise for the reader.

I think there are much more compelling arguments around regulating Google.

I'm not capable of running a mile in 4minutes, that's not unfair for me to loose to someone that trained hard to be able to run that mile in 4 minutes.

WSJ charges two to four times as much as my local paper.

I believe Microsoft still operates a large ad network.

Taboola is everywhere.

Amazon has a huge ad network as well, ~14B in revenue, though obviously limited in terms of what is advertised.

Verizon Media also apparently operates a sizable ad network.

https://newscorp.com/2017/12/05/news-corp-launches-news-iq/

(from putlake's comment above)

That's most of the bigger players, but there are hundreds. Many aren't worth the time, and are spun up when one rep takes some budget from a large brand and does "Direct" deals and tries to start their own shop with "Guaranteed CPMs". It's all bullshit

It's not a complex business to start, you know.