Content Marketing Doesn't Work (Except When it Does)

I love content marketing. Well, I love the idea of content marketing. Actually, it’s not so much that I love content marketing, it’s more that I dislike what I refer to as ‘traditional marketing’.

To me, as someone who knows very little about marketing, the idea of traditional marketing feels like screaming at strangers until one of them gives in an buys what you’re selling. I’m sure marketers will strongly disagree with me, and maybe I just need educating, but that’s how it feels.

This is why I love the idea of content marketing. You write about something interesting, and people read it. Maybe they’ll go on to check out your product or service, maybe they won’t, the point is there’s no hard sell. I like that I’m providing something (hopefully) of value, and not expecting anything in return.

So of course when I decided to go full time on BugMuncher at the beginning of September, I wanted to give this newfangled “Content Marketing” a try. I set a goal of writing at least one new blog post per week, and so far I’ve stuck to it. Initially I thought it might be hard to think of blog topics, but it’s actually been quite easy, even if I’ve had to resort to the dreaded ‘listicle’ a couple of times.

The best part was it seemed to be working. I’d write a post, submit it to places like Hacker News, Reddit and Bootstrappers.io, and BugMuncher’s traffic would subsequently spike.

Consider this graph of my happiness plotted over time:

It turns out this graph also shows how many users hit bugmuncher.com - and each of those spikes represents a new blog post being published.

Compare that to this graph showing free trial sign ups (as well as the quality of my sleep) over the same time period:

The sign ups graph follows roughly the same pattern - when traffic increases, sign ups increase too. Thus, I concluded that my content marketing efforts were working, and sat down to enjoy a celebratory cup of tea.

Last week I published the first in my series of blogs documenting my mission of taking BugMuncher from side project to profitable start up, which got a lot of visitors, and quite a few comments. One particular comment on Hacker New caught my attention:

If traffic's not coming to you for your product the vast majority of it won't care for your product, your best case scenario on product-irrelevant blog posts is awful traffic and most won't even deliver that.

An interesting viewpoint, and one that make me realise I’d never actually measured how many of BugMuncher’s sign ups initially landed on a blog post. This meant I couldn’t be sure of the effectiveness of my content marketing efforts.

So off I went to Google Analytics and defined a couple of new segments - ‘Landed on blog’, and ‘Landed on website’. I then looked at the conversions for each of those segments:

Oh dear. Yes ladies and gentlemen, you have not mis-read that, not one person who first landed on a blog post ever went on to sign up for BugMuncher. This was an unexpected turn or events. I double checked my segment definitions, and even ran a similar query in Heap Analytics, but there was no escaping it, my content marketing had been a huge flop.

So why do the first two graphs seem to show that when traffic spikes, so do sign ups?

I have a few theories:

1 - Not enough data

The eagle-eyed among you will have noticed that the sign ups graph peaks at just 4 new sign ups in any one day. Now I’m no statistician, but I suspect that such small numbers of sign ups mean the relationship between page views and sign ups is not statistically significant.

2 - Network effect

Like most things in life, this requires a Venn diagram:

Even though there’s no overlap between blog post readers and potential customers, some of the people reading the blog posts could still go on to tell their friends about BugMuncher. This would help explain the spike in sign ups when a new post is published.

3 - Computers are for creating, phones are for consuming

I think this is the most interesting theory - people could be reading the blog posts on their phones and tablets, but then using their computer to sign up. This definitely matches my own browsing habits, and Google Analytics data supports this theory insofar as blog posts have a much higher percentage of mobile views.

  Phone Tablet Desktop
Blog posts 25.36% 4.77% 69.87%
All other pages 10.85% 2.41% 86.73%
Sign Ups 1.47% 0% 98.53%

Over 30% of all blog post views are on a phone or tablet, compared to just over 13% of all other pages. Even more interesting, sign ups almost exclusively happen on desktops.

If someone’s first visit BugMuncher is a blog post on their phone, and they then go on to sign up using their computer, there’s a good chance that will be registered as two separate sessions, so the sign up won’t be attributed to the blog post.

So, Content Marketing - is it worth it?

Honestly? I have no idea, but it seems to have a positive effect on sign ups, and I enjoy writing, so I’ll definitely keep at it. Having said that, if I really want my content marketing to be effective, I need to consider more carefully the subjects of my blog posts, gotta try and introduce some overlap into that Venn diagram.

- Matt Bearman