05/05/2013

Instagram Is Visual Twitter

Joe McCann

If you haven't heard, Instagram recently announced the ability to tag people in photos. They have cleverly marketed it as "Photos of You" which puts the focus on tagging people, but this also presents the opportunity for brands to be tagged as well.

You see, Instagram is an incredible medium for brands to communicate and engage with current and prospective customers in a visually appealing way. We saw the same sort of attraction by brands (and retailers) to Pinterest when user growth reached a critical mass, for brands were able to tell a story, without actually saying a word (a picture is worth...).

Yet Pinterest is a different medium than Instagram in that it is not conducive to what I call "content snacking™" or the short bursts of attention that users divert to media such as Twitter, Facebook and Instagram when they are standing in line at Starbucks or even waiting for an elevator.

Instagram, on the other hand, is ideal for content snacking and more importantly, it is more effective than Twitter.

Twitter, for the most part, requires one to actually read their Twitter feed. Sure, we can skim for the most part, but with Instagram, the story is clear immediately.

Instagram is actually visual Twitter.

Visual Twitter

Twitter is parseable, readable, consumable by computers (software) and is relatively easy for brands, retailers, celebrities, politicians, really anyone or any entity to find out who is "talking" about them...it's text afterall! For example, "Prosecco" in the following tweet is very easy to find with a simple search.

Prosecco and crab cakes. Happy Cinco De Mayo.

— Joe McCann (@joemccann) May 5, 2013

But Instagram is different.

Prosecco in my Instagram (the photo) cannot be picked out with a simple text search -- it's not text! It's an image!

Besides the incredibly overused and abused hashtag on Instagram, photos are not easy to parse, analyze and consume by software, particularly the text or even logos, because it simply is not text! It requires some specialization, such as Optical Character Recognition, or OCR for short.

OCR is not necessarily a new technology, but it is not widely known and certainly is not commonly applied to a use case such as parsing Instagram feeds for words or even logos.

However, this is precisely the type of information brands would be interested in. Who is Instagraming? When? Where and with whom? How many people are "liking" it? What is the broader reach and impact of someone's Instagram post when that brand is indirectly or directly involved?

To answer these questions, Instagram has provided a stop-gap solution with tagging photos, but it has it's own set of issues.

Applied Tagging

Allowing people to tag brands, err, people in an Instagram post, before they post the photo, is a decent start and sure the value behind tagging actual human beings is great; the network effect for Insta-book grows stronger. But as brands start to be involved more and more in the tagging, the value, in my opinion, shifts from Instagram/Facebook's network effect, and more in the earned social media that brands will obtain from the tagging.

In the image above, your's truly has tagged myself (Instagram is a bit narcissistic, innit?) and the Theory brand, button down shirt I'm wearing. Theory, now knows that I have attributed (tagged) them to something about my post. Prior to this, Theory had very little knowledge of whom was wearing their clothes, yet I've explicitly included them in my post.

In my example, there is no Theory logo here to "parse" or "scan", but imagine the following Instagram post:

Out of the absurd number of hashtags included in this post, not one of them is actually "#Nike". Yet, this photo has 65 likes and is in fact a Nike product. Is this not of value in any way, shape or form to Nike? What if the post had 6,500 likes? Would Nike's opinion of the value of the Instagram post change then? If so, how would they know about it?

Current social listening platforms have no way of discerning where Nike is truly being represented on Instagram.

Relying on acutal human beings to include your brand's logo in their Instagram post is also foolish, as most folks won't participate, unless they are incentivized in some way.

Software As A Service

Instead, this is a service that needs to be handled by software in (near) realtime and not rely on crowd sourcing (tagging, hashtagging).

A service that parses Instagram in (near) realtime, looking for characters using OCR (I wrote my own here) or logo recognition using something like IQ Engines is precisely the type of product I'm looking to build while in my new role at Mother NY.

The ability for brands, retailers or anything with a logo to get even closer to those whom are "talking" about them or better understand them is extremely valuable to the brands and provides a massive opportunity to the person or company who builds it first. Instagram is the first truly visual social network and the tooling and analytics will need to be rethought give its inherently visual nature.