Think visually

Big social networks like Facebook and Twitter might get all the headlines, but visual networks like Pinterest, Instagram and Tumblr are a growing force to be reckoned with, says Sharad Verma, CEO of Piqora.

His company began life as Pinfluencer, dedicated to tracking pin activity on Pinterest – a site where people go to post and share pictures of things they love, be it art, fashion, brands or products. This week, it expanded its analytics capabilities to include Instagram and Tumblr.

The three sites have a lot in common, says Verma: “They are built around interest and intent” – whereas Facebook and Twitter are built around people and the connections between them.

For brands, trying to penetrate those social circles can be tough. But Verma says branded content is already widely shared on the visual networks – and people are doing so organically.

“Brands are being talked about whether or not they have already joined these visual networks officially. This constitutes a vast opportunity for brand managers to get deep insights about their audience and content and promote hashtags and content to acquire audience and turn them into customers.”

Research: What’s the background to Piqora?
Sharad Verma: Piqora used to be called Pinfluencer, and it was started in February 2012. Until now, the company has been solely focused on Pinterest, providing analytics, marketing and promotions, and content management for customers including Etsy, ZGallerie and ModCloth – companies mostly concentrated in the fashion, home décor and food verticals.

Now we’re announcing our extension into Instagram and Tumblr analytics and marketing to become the marketing suite for visual interest-based networks. We’re the only company that offers a unified dashboard for brands to be able to track their images, photos, hashtags, keywords and mentions across Pinterest, Instagram and Tumblr.

R: Why focus on those three networks?
SV: Consumers are spending a lot of time on tablets and on their mobile phones, which are lean-back entertainment devices. The hardest thing to do on a tablet or mobile phone is to type; the easiest thing to do is to tap on an image. So while the first generation of social networks mapped out the personal social graph, I think a lot of consumer time is now going to be spent on these visual networks.

We thought a lot about what brands are looking for and what type of networks they should they have presence in. We saw the rise of a lot of these visual networks, and the common attribute is that they are built around interest and intent. People are participating in these networks in such a way that creates data for marketers to drive insight and really learn about their audience; and that influences their offline and display advertising strategy; helps them understand their potential customers better; and helps brands have a more active presence in places where consumers are really interacting with their content.

Also, if you compare these networks with Facebook, you’ll see that Facebook is a closed network, where people are hanging out with their friends and family. It’s not a very viral network. It has just become a very large online destination where the majority of re-targeting happens today and where 25% of display advertising inventory is concentrated.

But the visual networks are all public, anyone can follow anyone, and they are public to big data companies like Piqora, who can mine all the data and turn it into insights. And because they are public, and because consumers are primarily there for themselves, they are recirculating their interests: they are re-pinning and re-blogging and that results in a very high amount of virality and reach, which is what brands are interested in.

So, even though these networks are not as big as Facebook, the virality and the reach is very high. 80% of the pins on Pinterest are re-pins and 90% of blogs on Tumblr are re-blogs. Also, we noticed that these networks are interconnected: 10% of all Pinterest pins are actually from Tumblr, so we also see aggregated virality and reach.

Our technology is really helping brands track their images across these visual networks, so they can understand which content, which images, are performing well, so they can shape their own content strategy. Tracking hashtags and mentions in conversations helps brands understand which keywords and hashtags their content is being associated with, and how consumers are incorporating brands into their lifestyles, and what are the associated lifestyles that brands should be paying attention to, and who their influencers are.

R: How much branded activity is taking place on these networks today?
SV: Mostly it’s all about organic and earned media. These networks haven’t opened up advertising options yet. They are beginning to take steps in that direction. Tumblr is the only visual network that has introduced advertising, in their dashboard feed. Pinterest and Instagram don’t have any way for advertisers to sponsor a pin or a post yet, but there are indications that they are moving in this direction. I think it’s inevitable that these networks will open up advertising eventually. It’s the best way for them to monetise.

R: So for now the brand presence is user-led?
SV: Yes. For example, someone might order a pair of Nike shoes and post a picture of them. Nike then wants to know what the user is saying as part of that, and what are all the other contexts in which Nike as a brand is getting shared and mentioned.

R: And you use image recognition to track what pictures are being shared?
SV: We can identify which brand an image is from. So far, we haven’t done identification of brand levels within an image, or identification of brand products within an image – that is a harder problem. But what we can do is look at an image and identify whether it originates from a brand’s website.

We’re doing that for Pinterest, and we’re sort of doing that for Tumblr, because a lot of the images shared there are picked up from a brand’s website and then incorporated into blogs. Instagram is slightly challenging because most photos are uploaded by users in a variety of styles and angles. Image recognition is not there yet where you can take a random photo and identify a product in it. But as and when it does become possible, we’ll be on top of it.

R: We talked earlier about the differences between social networks and visual networks. But what about the tone of the conversations that take place there, and what implications does that have for brand owners?
SV: Facebook and Twitter are about post-purchase sharing and complaining. Facebook is where you go when you’ve bought a pair of Nike’s and you want to tell your friends about it. Twitter is where you go if you don’t like your new Nike’s or you’ve had a bad flight.

Pinterest, Instagram and Tumblr, on the other hand, are really top of the funnel. They are about entertainment and improving your own life. Pinterest, for example, is about organising and getting new ideas and inspiration. Instagram is a casual creativity platform that is democratising photography as a skill. Tumblr is a place where people are consuming comedy and fashion. And a lot of the interaction is positive.

R: Are you able to build taste profiles for individuals across the three networks?
SV: That’s the plan, but we haven’t built it yet. The plan is to coalesce social identities – either using facial recognition, or name and identity matching algorithms – and aggregate all the interactions that a particular user is having across networks for a particular brand.

I think that by the end of the year, or by early next year, we will have a feed setup… a unified list of users who are advocates and influencers for brands across networks, a unified list of images that are trending across networks and a unified list of keywords and mentions and comments.