Platforms, distribution and audience

I've been writing a blog at ben-evans.com since 2010, more or less, first on self-hosted Wordpress, then Tumblr, and now Squarespace. In parallel, I've built up 90k or so Twitter followers and I have a newsletter with 40k subscribers, both of which serve in part to drive traffic back to the blog. This is what the monthly traffic looks like. 

In January 2013 I posted a bunch of interesting and rarely seen Facebook data (and launched my newsletter) and in January 2014 I joined Andreessen Horowitz. 

This is blogging the old-fashioned way. It's not the really old-fashioned way - I'm not writing raw HTML in notepad and uploading it over FTP (or indeed editing it direct over Telnet). You could do that, though - you could start out the way Justin Hall did 20 years ago, or use Wordpress, Squarespace, or one of the other new WYSIWYG publishing platforms. The challenge, and the point of the chart above, is that writing is not the same as being read. 20 years ago Netscape might have made you cool site of the day, but today there are hundreds of thousands or millions of 'blogs'. You can publish, yes, but you won't be seen. So, I spent two and a half years blogging regularly before my traffic picked up, and that was with a lot of work and a lot of time on Twitter as well. Blogging has never been easier but getting read has never been harder. 

Now, suppose you don't have 2-3 posts a week in you, but 1-2 every six months. That is, suppose you're actually busy doing something else. You could publish on the open web, like me or Justin, but frankly, you'd almost certainly be wasting your time no matter how good they are. You might have a topic that works really well with SEO, but that only applies to a subset of topics. You might have a following on Twitter (or be able to get people who do to link you) but many brilliant people don't - they don't have the time or they're just not allowed to talk about their work in that kind of unmediated way. (A similar point applies to Instagram for a different kind of content). But otherwise, it's very likely that no-one at all will read you. 

The problem isn't freedom or openness but distribution. 

There are a couple of places you can go to for distribution. If you work at Facebook or Google you might post it on Facebook or Google Plus. Your friends might see it in their feeds (though this is largely random) and they might share it and their friends might see it in their feeds. Maybe. You might wangle a Forbes guest post or blog, if it works with the brand. You might post it on LinkedIn and your network (if you have a big one) might see it in their news feeds (if they ever look at it), and LinkedIn might feature it. This one can drive huge traffic: they reposted my brief post on Microsoft's mobile capitulation and it got 400k views (and 900 high-value comments telling me I was an ignorant idiot and PCs will always be needed for real work).  But that's rather a lottery - you can't rely on being featured. 

Then there things that are both dedicated publishing platforms and distribution networks, and come with an audience - most obviously, Tumblr and Medium. You could apply Chris Dixon's post 'come for the tools, stay for the network' to this - come for the great blogging tools (and Medium and Tumblr are much simpler and easier than Wordpress) but stay for the distribution. But actually, the network is everything. Blogging tools are a commodity but the network is not. Tumblr and Medium have different takes here. Tumblr is (ironically) closer to the Twitter model, with a time-based feed based on a list of people you follow that takes a lot of time and effort to build up. Medium breaks free of that by deliberately recommending new content. It's trying to be a discovery and distribution network - a solution.

In that, Medium is something of a pair to Buzzfeed. Buzzfeed starts by trying to understand the fundamental dynamics of how journalism is found and read and works back from that to the journalism itself. This is exactly how newspapers and magazines got created in the first place - by working out how to be native to the new form that technology enabled, not by starting with what you already have and trying to make the new form fit into it (which is how most newspapers and magazines then reacted to the web). Medium aims to solve the same problem - 'how do you get read?' - for blogging.

The other parallel with Buzzfeed, of course, is revenue. Buzzfeed has been as innovative in the revenue model as in distribution, thinking about what advertising should be and how it should work rather than just throwing up banner ads and 10 meg of ad-tech JavaScript and hoping for the best. Twitter didn't do that either, and I doubt Medium would. This is partly an advantage of scale. You have to be very big indeed to be able to create new ad formats from scratch and persuade advertisers to pay real money for them. So, again, you want to be part of a network - a platform - of some kind, not just a publishing tool. A platform can create new revenue models where a CMS cannot. 

All this is really an argument for platforms over the raw URL, filling in things that that command line system doesn't have, things like distribution, identity or revenue. Hence, Google and Facebook gave one kind of distribution and Linkedin, Tumblr, Twitter and especially Medium provide another. Then, my boss Marc Andreessen sometimes points out that the web contains an unused placeholder for payment -  '402: Payment Required'. That's an unsolved problem so far, but one could equally point out that the web lacks identity. That's a problem for security and log-in, but identity is also what a lot of that 10 meg of Javascript is really doing - bodging "who is this person and what ads should we show them that might be relevant?" onto the top of the web. Facebook or Twitter don't need to do that, and neither does a Google search results page. They have the relevance information that they need and they can create the revenue product to match. Platforms can create revenue models where a CMS cannot. 

Finally, any such platform solves problems with the open, flat, flexible common standard by moving it into a richer, more powerful and sophisticated but also in some sense closed and proprietary place. In particular, there is always an embedded value judgement in what Google or Facebook (or Medium) shows you. A filter is a decision. Some people worry about this, though I tend to think competition (remember MySpace? Microsoft?) provides good protection. But really, this can only ever be about relocating the problem. You can rely on being on the web, but then you have to tell people your URL, or in Google's index and people have to search for you, or in Facebook and people have to share and to signal to Facebook's filters that they want to see you. If Google or Facebook have arbitrary and inscrutable algorithms, so do people's impulses and memories, and their decisions as to how to spend their time. That is, the open web has the same underlying problem as a closed propriety discovery platform - it's just expressed in a different way. More stuff is created every day than you could read in a lifetime - there's always going to be a filter.

 

(For another angle on some of these issues, see my post from June, Search, Discovery and Marketing)