While there is a lot of fraud, if you actually track from ad click to sale, you'll see SOME amount of results.

In most cases they will be underwhelming, but if the numbers make sense, why not?

example: 2$ per click for a $500 dollar product (assume no other costs_

if you can turn 1 out of 100 clicks into a sale, then it might still be worth it, despite 99 fake, uninterested, botted, clickfarmed or otherwise undesirable clicks.

If you can completely obliterate your margin like that just to make a single sale, I suspect prices are too high and you'd end up doing better by just offering the product at $400.

This is absurd. While I'm sure there is a lot of fraudulent activity and rubbish ad networks (IE Facebook likes, as claimed by the article), traditional online ads are pretty trivial to track at this point for a business that can complete transactions online. If you run a web store, you can track exactly how much sales you get from a particular ad channel and weigh it against the advertising cost by simply creating a unique link for each channel, or using any of the many available ad analytics platforms out there. Entire multi-million/billion dollar businesses are built around arbitraging this opportunity positively, particularly on Google.

Things like Facebook Likes are more difficult to track and are still relatively new, and that's why many people haven't figured them out yet, but over time they will either reach an equilibrium, where value provided is equal to cost or they will disappear.

I'm an anti-ads person, even though I know it hurts some of the sites I like who depend on advertising.

I'm anti-ads because the kind of ads that I am being exposed to are invasive, especially re-targeting ads. I.e. those ads where you've been on a website and thought about buying something, then decided against it? Then you get the same product advertised to you on random other sites you visit.

I cannot stand it when those ads follow me around. In fact, assuming I didn't buy the product or service for a good reason, and I do want to buy the product, I will deliberately buy from a competitor, because fuck you and your invasion of my privacy. It is my own little way of silent protest.

I assume I'm in a minority though and most people don't care. Most of the time I use an ad-blocker. iPad doesn't have a solution unfortunately.

Yes, you are in the tiny minority :) So your actions are pretty inconsequential to most companies

Ha! I love re-targeting... it keeps reminding me of companies who's products I've decided not to purchase automatically! And even better the company is paying someone to remind me not to do business with them!

The only targeting Ads that work for me are amazon.com ads.

I always try to deactivate and fallback to classic ads. It was at least possible in FB some years ago.

> "All on-line ads are mostly worthless"

I would tend to agree, if we keep the "mostly" in mind. You really need to target you ad money on places where the users are already trying to buy something.

Other than that I don't see ads on random sites working any better than television or print ads. Maybe that's okay, but let's face it, I'm not going to buy 10kg of fire wood because I see an ad on a news site.

This is just not true. There is much, much, much waste in online advertising, but there are companies who make it work and then go on to go public or get acquired for large amounts.

Right now at Perfect Audience, we have many customers using our tools and seeing anything from $5-15 in sales from each $1 they spend. It takes time and work to get to the high end of that spectrum, but people get there and it grows their business swiftly.

I disagree somewhat. Buying ads on google's search engine makes a lot of sense (for us), as google closes the loop from impression to click to sale (ads served to places where sales occur, keyword bids adjusted automatically based on sales). For my clients, google adwords adverts typically result in 100x-400x sales vs ad spend. A good deal for all involved.

So it's more of a hypothesis than a theory.

I used "theory" in the common meaning, but you're right, it would be better if I used "hypothesis" instead. I'll try to do better next time.