1. It's the most scalable way for them to generate revenue without shoving ads down your throat.
2. Skimlinks doesn't "modify" your links, they simply put it through a redirect to see if it affiliates. The link you've submitted to Pinterest will likely stay as is.
3. If users submit affiliate links already, then services like Viglink/Skimlinks don't overwrite your cookie (i.e. Reaffiliate it) so those users will continue to see revenue.
It also brings up another interesting point. If this revenue model works for pinterest, it means it could also work for pinterest users. Someone can create a browser plugin that automatically converts user links to affiliate links when they get posted to pinterest, and kicks back the commission to the poster. Probably too hard to distribute the plugin, but still thought it was worth mentioning.
In what bizarro world is this a problem?
If the customer acquisition cost (%) becomes higher than the vendor would like to continue paying, they will lower the commission they offer.
Vendors that pay affiliate fees do so willingly. They do this because all traffic comes from somewhere, and it is seldom free. If Amazon only has to pay one-referral for a completed sale vs. 100 click throughs required to complete the same sale, they're happy. If not, they'll down-correct their referral amount offered.
What's Pinterest going to do if this option doesn't exist, remove the links? No.. they could change the display algorithm to display links with their affiliate code more often, but that may mess with quality of content and click-rates..
Pinterest has every right to monetize their website in the least intrusive way possible. These referal links are transparent and provide a neat way to get some income.
What suprises me is how they are giving away 25% of their profits by using a third-party service instead of making something themselves.
Some (many?) programs will need to adjust their terms for this practice, or they may face one of those Groupon-like situations where a family bakery needs to serve 100k muffins, at a loss, in a weekend.
They assume a proportion of the buyers will be repeat purchasers. Paying for a loss-leader product for 100k casual visitors with unknown demographics or purchase patters is riskier than many businesses are comfortable with (for good reason!).
Also, if you are Fred's custom napkin business, and you usually sell 10 napkins a week and suddenly you receive 100k orders you are going to have a problem scaling a physical business that quickly.
Say if those 100,000 people came through an affiliate link, they only make $900,000.
If you're Amazon's affiliate manager, indeed your first thought is, "Wow, Pinterest is great! Look at all these sales"
But your second thought is, "Wow, if I just drop Pinterest from my program, I'll make an extra $100,000." In its current state, Pinterest will not do anything to stop these sales from being generated, so the merchants are just throwing away money.
The real problem here is that Pinterest doesn't deserve the commission, Pinterest's users do.
Also, I am not sure anyone "deserves" the commission. It isn't a right.