The Incredible Click Rate of Nothing | Hacker News Reader

The Incredible Click Rate of Nothing

adage.com

91 pointsDanielRibeiro27 comments

I'm astonished at the low CTRs the "big" advertisers are satisfied with. Most of the direct-response people I know (including me, although I'm a comparative novice) tend to junk anything getting below 0.1% clickthrough rate out of hand, and aim for .2% - .5% or higher.

I remember reading in passing something about companies being able to do funny tax stuff with their advertising budget. Maybe that is the motivation?

It wouldn't surprise me.

I guess it doesn't count as "funny stuff", but all advertising expenses are tax deductible.

http://www.irs.gov/publications/p535/ch11.html [ctrl-F "advertising expenses"]

Agreed. Bigger brands aren't as focused on clicks as they have agencies that push "branding" Branding is sexy because the metrics you are held to are (often) fairly loosely defined and illusory.

A lot of the entire adtech world is based on the idea that these giant brands are spending more and more on online ads, and spending it rather indiscriminately. So if you can make a product that's JUUUUST good enough to get into use at some of the big ad agencies and to get onto the media planning budgets of massive brands, you're golden.

In the DR (direct response) world, also affiliate marketing and so forth, the advertisers are only buying ads in so far as they generate sales/signups/whatever. More and more brands are jumping on the "our ads have to actually produce results" bandwagon, but most of the brand advertising you see out on the web is not held accountable in ways you'd expect.

The "big" advertisers are usually more focused on what is called branding - metrics like brand awareness, ad awareness, brand favorability, purchase intent - as opposed to direct response metrics like click-through rate and conversion rate. In fact many of the brand advertisers don't even sell products through their websites.

I would want to see how this stacks up with things like deliberately un-interesting and generic ads that at a glance still look like a normal product, or even ads designed to look like part of the page's background.

For all we know, they got a bunch of extra clicks just for looking different. (Remember the story about the banner ad made in MS-paint?)

Half of them were just curious, according to the survey the ad linked to.

Which brings "mistaken" back down to the average CTR for an ad of that type, instead of double the average.

They seem to have missed the most important thing about accidental clicks: the position of the ad matters more than anything else.

If your ad is next to an important button you will get a higher CTR. If you advertise on a site that gets a lot of mobile visitors and your ad is next to an important button you will get a very high CTR.

Quote from article: "When I want to make quick money on clicks," he said, "I just buy late-night impressions on women's gaming sites. I guess the users are tired. They click like crazy. I make a lot of money."

I don't know that much about online ads. How do you make money buying "impressions" like this? How does this work?

Ad guy here:

So this guy is probably talking about affiliate marketing.

My best guess:

1. He has access to lots of CPA (cost per action) offers from various CPA networks. Example: "We'll pay you $3 for every person who fills out this info request form about our product."

As an affiliate marketer, he's always hunting for offers where the payout seems high or the offer seems attractive, basically an arbitrage mismatch.

2. He has found some gaming offers that he thinks pay out pretty well and convert nicely. Good landing pages, good product, etc.

3. He buys ads on these womens gaming sites late at night because they are click happy and the impressions cost less due to it being night time. This sends people to the gaming site where they convert at a profitable rate.

Affiliate stuff is all about arbitrage.

I'm curious: On which platforms can I buy links targeted to "late at night"?

Most DSPs offer so-called "time parting" functionality, where you set your ads to run only during specific hours (and/or days). For example our product http://www.direct-ads.com/ or our competitor http://rtb.sitescout.com/ offer this.

Most decent independent (as in, sell their own adverts) publishers will offer this. Personally the main, or even only, use has been customers with age-rated products that they don't want to be seen before a children can be assumed to be asleep.

That's neat. So, out of the people who took the survey, 1/2 were curious and 1/2 mis-clicked. How many even bothered to take the survey? What's neat is that 0.08%, the reported CTR, is higher than on some regular ads.

I think the curious ones should be discounted. You would be surprised at how tempting a vague ad might be to naturally curious humans. The 0.04% figure of mistakes is interesting to me as 'background noise' for ad clicks.

This would probably be an interesting area for more academic research. How do you measure accidental clicks? Do some non-ads receive higher accidental clicks than others? What if it was a blank black ad? What if you mimicked actual ads but with the 'why did you click' question? There are lots of ways this phenomenon could be further explored, and I agree with the author that for a multi-billion dollar industry this exploration is warranted.

So a while back I wanted to make some ads for my web comic. I drew some pictures of the female protagonists and made an animgif cycling between them, because, well, sexy girls get clicks.

Then I looked at it without the text, and there was… something there.

So I did a poll on my blog as to which version people felt they'd click on. Things were split very slightly in favor of the one with the text, but I wasn't convinced. So I dumped 'em both into the same Project Wonderful campaign, and told PW to do the thing it does where it displays them at random, and starts changing the percentages based on which one gets more click-through.

A couple days later, the one with the text was almost never being displayed.

Of course, if everyone starts doing this then eventually people will get sick of ads with "important" information missing. Maybe I'm glad I got in before the rush.

Ha - I used to work with Ted in my P&G digital analytics days. Nice to see him shaking things up in this industry.

0.02-0.04%? That seems higher than Facebook's usual CTR of 0.015%. Maybe Facebook should use blank ads, too, and see their ad revenue double overnight.

Where do you get that 0.015% is a usual CTR for FB? That's actually quite low, even for them.

From 2 articles that appeared on HN a while ago, that tested Facebook advertising campaigns.

Remember, average CTR's don't really indicate much - it depends entirely on what you are running.

For the direct response ads I've been running, each campaign has a totally different average. One product has a 0.02% CTR average over 3 years and ~200k spent, another has a 0.155% CTR over 1 year and ~45k spent.

Different messages garner different responses, so take the typical FB CTR with a grain of salt :)

This experiment probably means this - most of the AD designs are Awful, with a big letter "A". They are so bad that they actually decrease CTR of AD campaign even compared to the blank rectangle image.

Reply on news.ycombinator.com