Originally Posted by wankawonk

That's exactly what I'm doing, and also A/B testing landers (thumbnails, order of categories, site designs), bids, countries, and more. It's just not profitable. My initial ROI hasn't gone much above -70% on anything I've tried, and I'm paying over .20 per engaged session (where the user makes several clicks on my site) AFTER optimizing...engaged users are worth about .001-.002 per session, so these engaged users would have to return at least 100 times, possibly 200, to be profitable.



Right, you build some repeat users with skimmed traffic and Google will like your site better (probably because you generate brand searches on google for your domain). I've seen this happen in practice on some of my sites.

But a lot of the sites buying a shit-ton of feeder traffic have no SE traffic and have been buying feeder hits for years.