If you mean by "gaming the system" that people put their own referrer tag in the link, I agree.

But I'd game the system by spamming the whole service and recommend as many products as possible.

To be honest, I don't believe that refer.ly will succeed. It adds a layer of complexity to the end user. If I recommend a product, I just write a simple email to the persom I'm recommending to. If I recommend a product on my website, I just write about it or use affiliate marketing (e. g. my own Amazon tag). What would I need refer.ly for?

A couple years ago, I've seen a startup called loved.by (google it) which was the exact same thing. It seems that the pivoted now to a more generic affiliate marketing platform.

Thanks for mentioning loved.by. I had never heard of them. AFAIK the Referly team started working on the idea about two years ago about when loved.by first debuted.

I think Refer.ly could work if they add enough value to the end user by aggregating every imaginable affiliate service into one and offering some sort of easy to understand and consistent metrics across all of them.

An aggregator could work, but you're at the mercy of the big affiliates from which you need buy-in - and from what it sounds like Amazon isn't playing with them.

All it takes is a cease and desist letter, a la craigslist and padmapper, and you've lost a large portion of your market.

The code to do this is on Github (at least for the UK): https://github.com/fubralimited/php-oara Affiliate network integration is not a technical challenge at all.

I'm also sceptical. It's hard enough to get consumers to understand and bother with the affiliate model when they could earn pounds (http://www.quidco.com/), let alone pennies (refer.ly). I suspect they'll pivot (either towards Shopcade or towards Skimlinks) within 12 months.

Referly strips out other referrer tags, and we use several approaches to handle "gaming". These are issues that every affiliate network faces, but they can be managed.

how are you able to track what people purchase then?

My evidence would suggest that they guess.

You'd be surprised at the level of data the affiliate networks make available to the people using them. Even if you are a small affiliate, purchase/basket level data is on offer.

how could they build a business off of guessing if the purchases are made. especially with the 24 hour length of the cookie, it would make it nearly impossible to track if multiple referly users have the same items. they must have some arrangement or something going on that is not clear.