It happens all the time – you’re on Pinterest and see something so cool that you immediately want to find out what it is, and where you can purchase it. Basically, an experience right out of a marketer’s dream.
If the Pinterest image you’ve landed on is the focus item in the Pin, then it’s quite easy to find a place to buy it, either right on Pinterest or elsewhere online. But what if you’re interested in a specific item in only a small part of a photo? Or what if you want to see other photos of the item that aren’t exact matches, but are similar?
This is where Pinterest’s new visual search shines. It gives you the ability to search Pinterest for pins that visually resemble the one you’re looking at, or search for a specific item within the photo.
With Pinterest being a must-have platform for marketers, it’s important to understand how this feature works and what marketers stand to gain from it. In the sections that follow, we’ll illustrate this feature’s capabilities and suggest marketing strategies to begin thinking about in regards to this new Pinterest feature.
How To Use Pinterest Visual Search
I picked this pin of a guy wearing a sweater, which is an excellent example, since I’ve never been known as a particularly fashionable person. Let’s say I want to level up my personal style, I see this sweater on Pinterest, and I want this look for my own. I know nothing about it other than what Pinterest tells me and since it was pinned from Flickr, there are no details whatsoever. Let’s click the icon in the top right corner to get started.
Pinterest will immediately populate image results and create a tag crowd above the results field. Lo and behold, what I thought was just an ordinary sweater could also be referred to as a cardigan. But what if I’m not satisfied with the results? What if I’m interested in nothing but the sweater (*ahem* cardigan) and I want my results to be more specific?
Drag the corners of the selection box to focus on a tighter area and you can narrow down results considerably. In this instance I couldn’t find the exact sweater I initially stumbled onto. This could be a missed opportunity for the retailer selling this garment as it’s exactly the kind of thing a user would impulse-buy. A little research using Google’s reverse image search reveals that the sweater in question was part of H&M’s Fall 2013 collection, so it’s not available anyway. Rats! With that being said, I did find a few dozen other similar sweaters both in style & color, which is perfectly fine since I wasn’t dead-set on eggshell white anyway (it IS past Labor Day, you know).
Let’s try another example. Here I’m looking at living room setups to come up with ideas for my apartment, and I see this one that catches my eye. More specifically, I like the white chair in the corner, but not knowing furniture nomenclature I have no idea what to call it.
Time to put Pinterest’s visual search to work! Zoom in on that particular chair and sure enough, there it is, right in the results. A retailer selling the chair pinned it to their page and a few short clicks later I’m on their website looking at it. Maybe I’ll buy it, or maybe not – maybe I take a second look and decide I’m more interested in the chairs with the wood accents on the legs (they would really tie the whole room together).
Even in the image with multiple items, Pinterest was able to location pictures of chairs that look identical to the one in the original image.
It remains to be seen whether this feature is widely used, although I suspect it will be. In terms of accuracy it works quite well – in both examples featured here it found identical or close matches to the areas I zoomed in on. The image recognition capability is excellent, and the real limitation is whether relevant images have actually been pinned on Pinterest.
For exact match image identification, Google Reverse Image Search is still the way to go, as it pulls from Google’s much larger index of photos and images across the entire web. In this regard, Pinterest’s tool may play more of a discovery role than an exact match research tool. However, many users currently use Pinterest for their shopping inspiration. If this tool gains in popularity, it will be crucial for brands to upload more product images onto their profiles to increase the chances of users finding a brand’s product through this visualization tool.
How can marketers capitalize on this feature? After all, the search results here are not based on keywords, so how can you optimize for them? For one, look at popular pages in your vertical. A lower cost clothing retailer, for example, might look at J. Crew’s Pinterest page to identify their most popular pins. That retailer could then prioritize posting similar products, that customers can buy at a lower price. If an ecommerce site’s product pages have Open Graph tags, Pinterest will include the price immediately within the pin.
As shown in the sweater example above, users can also search for visually similar images as a whole. Let’s use a bridal company as an example. For instance, if one type of image is very popular, such as a bride running in field, bridal companies could develop insights into what style of product photography is favorited by users. As a result, brands can produce similar photo styles to showcase their products through the visual search tool.
Want more juicy Pinterest tips? Check out Seer’s Pinterest Guide For Marketers. Check back in on this blog post in a few weeks for more updates on Pinterest’s new visual search tool implications for marketers!
How do you think Pinterest’s visual search will impact marketers Pinterest strategy? Share your thoughts in the comments below!