
Is Medium a platform for attention seekers & troll gazing?
Medium bucks a UX design trend which has important implications for what content will thrive on it
Veteran blogger Anil Dash, admirably explaining his friend Evan William’s new service Medium, says it is blogging in form, but not in structure. It’s the Winer-esque unedited voice of a person, BUT sans the reverse chronological timeline of posts. Instead says Dash, Medium is organised around collections, and the identity of the writer is of significantly less importance, so much so, that that part of its functionality is outsourced to Twitter. Dash also mentions — almost in passing — that you can’t follow a person on Medium either.
However in eschewing what has become ubiquitous design affordances for mass media services, Medium has made life very hard for itself. Its design may encourage trolling and other undesirable effects.

Discoverability is a big deal
How do you discover good stuff on Medium? Dash concedes that the inability to follow people on Medium might be a problem, because following collections are ‘tough’. He does not explain why it’s tough, but it’s easy enough to figure out why. Collections are — especially the open ones — not that interesting. The quality of writing in a collection varies widely, that is if you bother to look inside them.But since you don’t know which writers are inside a collection and Medium’s essays and collections do not carry the imprimatur of a publication like the New York Times, it’s unlikely you bother in the first place.
The biggest lesson I learned on my first job in digital media back in 1999 was that discoverability is a big deal. We launched a content site — eCountries.com — staffed with high powered journalists sourced from the likes of The Economist and headed up by Michael Elliot, ex-editor of Newsweek. Even though they were heavyweight hacks none of our journalists had the star power and the personal audience some journalists have in the age of social media.
Our site was new and was far from being a recognised name itself. That was a problem because we had no above-the-line advertising budget to build a ‘brand’ either. What’s more, unlike our competition, The Economist and the The FT we had no visibility in the news stands. We were entirely dependent on people discovering us through search engines.
For it was only when you launch a web only publication that you realise magazines and newspapers’ physical form, their vast distribution networks and presence in shops and stands, are very much part of their visibility and how people discovered and remembered them.But physicality and distribution is very expensive. In contrast a website is dirt cheap, but a very lonely place. Nobody just casually walks by it.
Cater for me please
Recently, Evan Williams founder of both Medium, Twitter and Blogger in response to a question by the same Anil Dash let slip something fundamental:
The big thing we missed with Blogger is the network. We competed for years on features—trying to not lose users to the power tools (MT and later WordPress), and not taking advantage of the fact we had all these readers coming to our (pseudo-) centralized system. From that, we could have built distribution and feedback mechanisms that are much more important than any other feature. Imagine moving your Tumblr or tweets to another system. It doesn’t make sense, because it’s about the network.
Evan Williams is right, if you have a large centralised network you can do things you otherwise can not, distribution and discovery mechanisms being of the most significant. But although distribution and feedback mechanisms can be built on top of networks, they are not core to what networks are about. Nodes and ties are core to what networks are about.What Williams seemed to not have forgotten is that without nodes and ties distribution and feedback mechanisms are very undirected. Who are you feeding back to and from who? Nodes and ties are as absent from Medium as they are were from Blogger.
To better understand why networks matter, we can perhaps find an answer in sociology. Just over a year ago,social scientists Lee Rainey and Barry Wellman coined a phrase — Networked Individualism.The point they make with it is simple. The rise of several communication technologies from the phone, to email, the mobile, and now social networking, has changed the ways we can connect to others. As a result we are no longer organised into closed and location bound groups or communities, but part of much more fluid,geography independent and larger networks with the individual autonomous at the centre of these.
Importantly —and very different from our old conception of groups and communities — while my and your network nodes might overlap, they are unique to each of us.We are Networked Individuals. This way of relating to people is very well encapsulated in the design of services like Facebook.We have ties (friending and following) with nodes (people, organisations), and the footprint of these networks are very unique to us.
Facebook has no face
Facebook has something remarkable ‘missing’ when you think about it.It has no face: It has no uniform home page that everybody can see. That it does not have a single entry point is as significant departure in the history of mass media as there ever was.From the moment you have signed in you are presented with your personal newsfeed, a individually tailored offering of content, based on your connections.
The importance of your place in between nodes and ties is something that Mark Zuckerburg seems to instinctively get. When he said that a squirrel dying in front of your house may be more relevant to your interests than a child dying in Africa, an outcry followed. But the point he made shows a dispassionate understanding of how people experience the world. We all watch the world from our personal vantage point as a node in our social milieu, our network.
If you are a content service designer, tapping into the network, is — so far— the unbeatable way to build distribution, discovery and feedback mechanism — precisely because it can target needs in such an individualised way.
So proven is this model and the individually tailored filtering it enables that even an established service like YouTube has been changing it’s design affordances to be Twitter like. YouTube used to have one front page. Now it has a highly personalised feed, again based on your position in the network. This trend toward discoverability and filtering based on Networked Individualism is near universal. That is until Medium came along.
We have yet to see an example of a network-less platform like Medium to have an effective automatic feedback and distribution mechanism, without doing that feedback in a very explicit and transparent way. The best example is Reddit and how it exposes stories in response to up-votes. But with its lack of a network Reddit has to carry baggage: an old school universal front page for all. It can not show us a personalised feed. And that constrains how universally popular it can be. Reddit manages to serve 6% of US internet users, Facebook manages 67% (Pew).
Medium is not a network
The decision not to have friending or following on Medium is obviously not without consequences.While Medium is a platform, Medium is not a network.
The first consequence is Medium struggles to be a place for discovery. Some articles I have read on Medium have been emailed to me by friends, but most have come via people I follow on Twitter.Medium does not only outsource it’s identity affordances to Twitter,but also it’s discovery mechanisms.
The second consequence is that the rich will get richer.Since Medium does not allow good writers to build an audience on Medium, users with existing large audiences on other networks like Twitter will have an advantage when “seeding” their content. They can leverage their existing networks for discovery.
A third consequence is that a Medium post’s visibility will more closely approximate a power law type distribution than say on Facebook, with most articles getting very little traction and some getting massive amounts of attention. In other words, something approximating traditional media’s fondness for one to many content distribution.
Another and perhaps most troubling consequence might be that since distribution is significantly predicated on being shared, trolling and attention seeking might become the order of the day.We already have some spectacular examples.







