We relaunched the iDoneThis Blog on July 9th, and one of the major initiatives has been to blog on a daily basis on weekdays. Since July 9th, we’ve published 26 blog posts on productivity, management, and the future of work.
When you put the articles into 1,000-pageview buckets as above, what you see is that the vast majority of articles has gotten sub-1,000 pageviews. 38% of articles are sub-1,000, 19% are between 1,000 and 2,000, 19% are between 2,000 and 3,000, 7% are between 3,000 and 4,000, 4% are between 4,000 and 5,000, and 7% exceed 5,000.
However, just 2 articles (7%), account for a whopping 53% of all pageviews, and the top 20% of articles account for nearly 70% of all pageviews. What you get looks a whole like a power law distribution.
What is the Power Law?
The Power Law is a statistical concept “where one quantity varies as a power of another.”
In a power-law graph, you see that the few to the left, highlighted in green, dominate the value being created. Then it’s followed by the yellow section, which is the long tail.
Often, what’s commonly known as the 80-20 rule is described by a power law distribution. The 80-20 rule is that 80% of the results come from 20% of the effort.
What’s problematic about the power law in the context of content marketing is that it feels like a hit-driven business in which you’re on a constant treadmill. However, if we understand why content often falls in a power law distribution, that can provide us emotional solace and a way forward.
The Why of the Power Law and What We Can Do about It
I suspect that this happens because we don’t own any of the truly massive distribution channels. Our content lives or dies based on the whims of places like Facebook, LinkedIn, Reddit, and Hacker News. For instance, our biggest article, 95% of Managers Follow an Outdated Theory of Motivation, got a whopping 27,673 views because it was massively shared on LinkedIn. It’s unclear to us at the moment how to duplicate that kind of sharing on LinkedIn.
We have modest-sized email distribution lists which, when grown, will raise the pageview baseline and better expose each piece of content to the opportunity of blowing up. This is why email lists are absolutely vital for SaaS companies, and it’s something that startups like Buffer and Helpscout have done extremely well.
In addition, previously successful articles give us data points that we can use to understand what articles will blow up in the future. For example, the 13,147-pageview article, The Boring Trait Google Looks For in Its Leaders, was a follow up on the 22,505-pageview article, The Most Innovative Employees at Google Aren’t Stanford/MIT grads with Perfect SATs, that we published a year ago in May 2013.
* * *
The 20% of the articles that give us 80% of the results can show us the way forward in terms of both the topics we should be writing about, and how to frame them.
I don’t think that means we can get away with doing less work and getting the same results, though. Rather, I think that producing a high quantity of content is priceless in terms of gathering more data points on what will work and what won’t.
Ultimately, I expect that we’ll continue to see this kind of distribution, but with an ever-increasing floor on how the content performs. In fact, it’s reassuring to think that this pattern will continue to recur, just on a larger scale as we grow.