Why ads stopped working
alexblom.com
alexblom.com
In the age of A/B testing, this is just incredibly stupid logic. If Myspace has 17 ads on a page, it's most likely because multivariate testing showed that 17 was the optimal number in terms of their eCPM.
It's always humorous when some guy with a blog assumes he knows better than the company with a zillion page views per day and access to abundant performance metrics.
That means Myspace may earn more money but the ads are not as effective. As soon as a more efficient channel is found Myspace is dumped. We see this issue happen over and over.
Actually, they probably leave some money on the table: they'll get a higher net present value if they reduce the number of people who quit in disgust, and for some people, the disgust from ads is cumulative. Thus, 20 ads might get them more revenue per page, but fewer page views in the future.
My point is that there is a balance and ad rates are falling in over-populated mediums because we know that the better chance people have of seeing the ads the higher chance we have of catching their interest.
A startup I am watching which is trying this is influads.com
Ad rates are falling (in some areas, not in all) because supply is growing faster than demand. Also more ads are appearing on less intent-focused sites, like Facebook, and branding ads always pay peanuts compared to direct sales.
There probably have been a lot of models showing that asset prices will not go down more than x%, because they have not gone down more than x% anytime in the last n years for which we happen to have data.
Similarly, 17 ads on a page might maximize eCPM today. But it is also possible that showing 17 ads on a page today could lead to fewer users visiting the site tomorrow. In other words, it is difficult to design an A/B test that takes into account long term effects on your business.
The second problem is that one of the costs of excessive advertising is to your brand. And testing things that affect your brand is something that A/B testing is not good for.
Perhaps they do A/B testing, but I think it's a leap to assume that they do.
No. Just because something is being done doesn't make it either the right thing to do or indicative of the way it should be done.
Just look at the financial mess we're in. All of that was based on the same kind of thinking.
We also started our own affiliate program (every customer automatically becomes an affiliate) and trained our customers how to refer their friends and family to us. That has paid off better than buying ads space on blogs or news sites.
There may be some types of ads that do very well as banners on general news sites (I'm thinking of those weight loss and bad breath ads) but it hasn't worked out for us at all. It all depends on your market, the people who visit the sites you advertise on as well the information contained in your actual ad.
(1) Ad relevancy must be above a very high threshold, else users will filter out all advertisements, even if some are relevant. This threshold is demographic dependent. (For HN crowd, I contend it would be around the 90% mark...I, for one, still ignore gmail advertisements).
(2) Users must trust that the advertisement will give a good deal, else they will just go to Amazon, eBay etc to buy the product instead. I think trust is an often overlooked aspect of advertising, and is a difficult problem to solve.
If ever I get around to building an advertising framework around my app, I will attempt to err on the side of not running advertisements rather than risk losing the relevancy and trust battle.
Sometimes people simply hear you better when you're not yelling.
Content ads make money because people see something that catches their eye while reading the content they came to the site for. This requires having several ad spots rather than just a couple at the top of search results. Content is usually unique, so there's a high annoyance bar to reach before you start to drive people away.
Yes, fewer ads make the ads shown more effective. I remember Hulu ads more often than I remember TV ads. (Then again, I watch far more Hulu than TV these days.) The problem with this approach is that advertisers aren't willing to pay enough money to cover the losses from showing fewer ads.
What you and the author are suggesting is a demonstrably unsuccessful strategy for someone who is running a website until you can convince advertisers to pay more money for ads with less competition on the page. I don't think this is impossible.
However Google has ads in a lot of places other than search. And makes a lot of money on them. Yet somehow they also haven't turned gmail, youtube, picassa, and maps into banner-fests.
It's unfortunate that people in the best position to talk about ads also find them distasteful.
The scenario I am speaking of is more akin to playing 4 infomercials at once. Suddenly neither really catch our attention and have the chance to suck us in like playing one infomercial does.
If you look at some of Apple's iPhone ads, those are sort of compressed -- in the JG Ballard sense -- infomercials, quickly demonstrating several apps that solve various problems.
edit: took a look at the Deck's page and they talk about "cost per influence".
So according to themselves, the publishers running these ads are making less money than they would with any other program.
if you have enough traffic to make $4,000 with the lowest CPM in the business...then you have enough traffic to making $4,000+X with companies that have better CPM
The are buying into the "cost-per-influence" idea, which means they have abso-freakin-lutely not idea if it's working (in terms of ROI). I seriously doubt that anyone is getting $7,900 of profit per day from advertising on the deck.
There may, for all I know, be some value in just being in front of the right people's faces (which is really the premise of brand advertising), but that's not really measurable.
They are covering Startups, Design, web development and productivity apps I belive.
With top members like venturehacks and answers.onstartups.com, they seem to be doing a good work, even if being 2 month old.