Conversion optimization is often talked about solely in terms of getting more sales or generating more leads, but there is so much more that you can do with it, because in principal split testing can be used to make any web page better at doing whatever you want that page to do..
So, in this post we will take an example: Let’s say that you want to make your blog more successful and your content more viral – how can you do that?
Step 1 – Outline your goals
This may seem obvious, but there are in fact a few goals that your blog might legitimately have and you need to decide on your priorities…
Some businesses see their blog as a purely promotional tool (which is generally a bad idea, but that’s a different issue) in which case, they may think that the main goal of any blog post is to get the visitor to a sales or contact page.
Another goal you might have is to get the reader to visit more of your other blog posts, in order to increase the time on site. Which isn’t a bad goal to have either.
But in our example, if we want our content to be more viral then our best bet is probably to prioritize getting it shared on social media channels and to an extent getting people to link to it. The reality of course is that getting links is much harder to optimize for, so we’ll focus on social traffic.
The Message Here Is:
If you want to use your blog to generate more traffic and brand awareness you may have to prioritize user experience and share-ability over hard selling. This may (or may not) mean that your sales drop in the short term, but creating a more successful blog is probably worth the cost.
Step 2 – Choose Which Metrics To Test For
You could just dive in and play around with your social sharing buttons, but let’s take a more methodical approach – after all, you’d rather make an improvement on your first attempt, wouldn’t you?
So first of all, have a look at your traffic and where it’s currently coming from. In particular, are you already getting much social media traffic?
If a good portion of your traffic comes from Facebook for example, then it is likely that your audience is in love with Facebook, so that is also likely to be the easiest channel to get your users to share your content on.
Obviously, if your traffic from other channels is lacking then trying to diversify your traffic sources is a good idea, but right now it is worth focusing your tests on increasing Facebook shares, since this is likely to yield the best results.
The Message Here Is:
Focusing too much on a single goal can be dangerous, but you do need to know what your priorities are. In the example above, it is likely that optimizing for Facebook shares would yield the best results in terms of making the blog more viral, but do make sure that you don’t do it at the expense of other social channels.
Step 3 – Decide What To Test
Next you need to set up a split test, but first let’s figure out which aspect of your blog most need testing. This is where it is helpful to use Analytics and Session Tracking. By looking at click maps (Crazy Egg is a pretty good place to start) for your most popular blog posts you can get a good idea of what your users like and what they don’t like.
Typical things to look for are which aspects of a blog post hold their attention and which aspects lose it. Many blogs do better with more images, but there might be a point where too many images can hurt. Likewise, aside boxes and bullet points may be helpful or distracting.
The idea isn’t to figure out what works, it is to create some hypotheses that you can test…
Step 4 – Setting Up Your Test
Actually, setting up your test will require a little bit of technical knowledge which is outside of the scope of this post, but check out Optimizely, it’s fairly cheap and pretty easy to use, so it’s a good place to start. For a first test you can do one of two things:
- 1. Choose a single blog post to test (testing with a new post works well)
- 2. Testing all of your blog posts at once
If you choose the latter you will get more traffic to test, but you will obviously only be able to test for more general layout changes. This is fine of course and you can test for things like:
- Position, size and style of social sharing buttons
- Moving or removing other navigation items
- Changing the look and feel of your blog pages
- Pop ups, banners and such like
There are a lot of cool things you can do and testing elements on multiple pages is pretty easy. Optimizely offer a guide to doing this here, although of course how you do it depends on the platform you use.
Alternatively, you can test a single blog post, in which case you can test for changes within the post itself. Now your findings will have to be more generalized and they won’t always apply equally to other blog posts, so it is worth running a few tests.
Here are some things that you might want to test:
- Using more/fewer headings within a post
- Using more/fewer bullet points
- Short vs long vs really long blog posts
- How you position your images (big/centered vs small/floated)
- Use of anchor texts (long verses short)
- Tone of writing (funny vs formal vs sarcastic etc…)
- You could also test including social sharing calls to action within the post itself
In general, running tests on an individual post will give you a better idea of what appeals to your audience. You will probably not be testing for social shares directly, but more for what keeps the reader on the page and reading.
Step 5 – Closing Thoughts
The great thing about split testing is that it gives you really numbers so you can finally find out just how much difference different things make. To get the best results it is a good idea to go big and test big changes because big changes get big results.
If you test a big redesign you can always test fine tuning the smaller elements later, and if you are worried about confusing existing users, consider just running the test on new visitors until you know which version you will be keeping.
The great thing about testing to improve the sharability of your content is that doing so will bring in more social traffic and help with your SEO. Just remember to keep testing as your traffic grows and as the makeup of that traffic changes – for instance, if you start receiving more traffic from Twitter or Google.