This guest post comes from Marc Mezzacca, founder of NextGen Shopping, a company dedicated to creating innovative shopping mashups. Marc’s latest venture is a social-media based coupon code website called CouponFollow which utilizes the Twitter API.
How much of the content posted to Pinterest is from retail or e-commerce websites? How much money can they make off these links? These are hot questions right now. I’ll get to them, but first let’s back up. Recently journalists have had a field day in thinking they found some new, hidden secret within Pinterest because the pinterest.com website utilized SkimLinks service to monetize the user generated content (Pins). Well, Pinterest has been using SkimLinks since around the time they initially launched. I should also point out many other sites also use SkimLinks and don't formally disclose it, as it can cause confusion to those uneducated in the affiliate space.
By browsing the site, I had a hunch a majority of the Pins weren’t monetizeable. Following Jay’s post at Business Insider around how VCs are excited for the Pinterest API and Adam’s semi-announcement of the Pinterest API launch, I decided to dig deeper into it utilizing their own API to find answers.
Note that there has also been a lot of confusion about the API. I can confirm, as Badgy points out, that the private part of the API (requires OAuth) isn’t available yet to general developers as the end-point is disabled. But the public facing part seems to be working. Good enough.
So I used the public API to gather some basic insights that hopefully translate in a rough approximation of how much e-commerce related content is actually going onto Pinterest.
Objective: Determine how much e-commerce is going through Pinterest as posted by users.
I should disclose that I’m using technology from my company to detect in real-time if a source website is e-commerce or not, meaning is it a direct point of purchase. It works with a high-level of accuracy and we use it in production with our CouponFollow website.
Sampling and Results
I took a sample of data of Pins from the API and broke the Pins’ Sources into four parts: E-Commerce websites with affiliate programs, E-commerce websites without affiliate programs, Non E-Commerce websites, and other unclassified websites.
- E-Commerce (aff) – These are retailers or service providers where users can conduct a transaction. A user can click on the pin and be brought directly to a place where they can purchase the item. The site has an affiliate program, so Pinterest will theoretically generate commissions on purchases.
- E-Commerce (no aff) – Same as above, except the site doesn’t have an affiliate program, so Pinterest would likely NOT generate commissions on purchases.
- Non E-Commerce sites –These are blogs, search engines, static images, image curators, other curators, or some other type of website.
- Unknown / Other – These are other websites we aren’t completely sure of. It is important to note that these also have a very high probability of being Non-Ecommerce sites.
The initial sample I took through their API was around 13K pins over the course of 24 hours. Of the initial sample (only popular pins), only about 9% were definitely from e-commerce sites. And actually, I think it’s safe to say at least 85-90% of the Pins in the sample taken from the Popular category were not e-commerce. That means the source link where it came from is not a purchase point. Instead it was a blog, search engine, static image, etc -- as mentioned above. Now, of those e-commerce sites, not all have affiliate programs that Pinterest can tap into to monetize the traffic. I cross referenced the e-commerce sites with an affiliate database to further refine the information, as shown below:
And here is a breakdown of the actual Pin sources. Of the 13K pins, there were about 5.4K different Sources (websites).
Rethinking the Sample
At this point it appears that only a very small percentage of the content is monetizeable through affiliate links by Pinterest. That’s what I thought. But wait a minute… Just sampling the Popular section isn’t a true test. So I went back and took a smaller sample on different categories. This time I choose categories heavily geared towards purchasable products. The next sample I took was about 5.5K pins taken from categories as shown below:
This new sample data changed the results significantly. The amount of e-commerce related Pins jumped up to around 30% from just 9% initially, with about 12% being websites with affiliate programs.
And the number of e-commerce Sources also jumped from 9% to 25%.
This was a bit higher than I anticipated, even though these categories are heavily geared toward shopping.
What Are the Most Popular Sites to Pin From?
Finally, I ran some aggregate analysis on where the Pins were coming from using the entire sample of about 18.5K pins. Tumblr was by far the largest source for Pins, with BlogSpot being second, and together accounted for almost 1/4th of all the Pins. Etsy rocked out in the number 4 position, being the only true e-commerce destination on the list. Amazon came in farther down the list somewhere around 13th, and was the top website source with an affiliate program.
Even with this information, I won’t try to estimate their affiliate earnings. Others might try, but realistically there are too many variables. What I can say is that Pinterest is already generating more quality traffic to retailers than other social networks. While I’m not certain if Pinterest will become the defacto social shopping destination, it certainly is filling a void that currently exists. Regardless, monetizing their affiliate traffic from the start was certainly a smart idea that could have easily been a missed step by a less savvy startup. Now that they have serious funding, they are able to move away from this method and focus on the product.
I’m sure Pinterest can provide much more insightful data themselves than these small samples produce, and likely (eventually) will, but this is certainly interesting and fun to see… and a good chance to explore their much anticipated API. Note: Be sure also to read Pinterest Data Analysis: An Inside Look by RJ Analytics which provides much more generalized information, buy from a much larger sample size.