Website Layout Ideas & Tips From Top Publishers & Blogs

In fact, Website Visitor Segmentation Offers Big Ad Earnings Value For Little Work

This case study had some pretty interesting things to say about treating visitors differently.

It showed that by identifying and tracking user behavior that publishers could increase ad revenue and objectively give visitors a more engaging experience by slightly changing ad combinations based on how similar visitors had responded to these elements in the past.

Did these publishers regularly include website slideshows?

Slideshows — or swipable content on mobile devices — usually have both proponents and detractors.
Sites that use these mechanisms obviously hope it will lead to users visiting more pages, and thus more ad impressions and pageviews.
However, this may not always be a good thing. Detractors would point to research on navigation bounces and the impact on page engagement time as reasons why these features should be avoided.
Ultimately, every site should be testing this stuff for themselves. But, what about our 20 popular publishers?
How many used slideshows or suitable content to lead visitors through their site articles?

If you’re interested in measuring how slideshows impact your navigation bounces or overall engagement, you can do it for free using Ezoic’s Big Data Analytics.
We’ve found a direct correlation to engagement and ad rates from advertisers.

How many images are these publishers using in their articles?

We know that word count can play a huge role in engagement time — and that it varies from site to site — but what about image count?

Many publishers will only use one image at the top of the page to kick off the article (Wired is a good example of this).
But, how many use more images than that?
Less than you might think.

Only 25% of the sites used 3 or more images. The most common website layout for images was to only use one image at the top of the article.
Only 1 site in the sample size had no images in its typical article.
Surprisingly, only having 2 images was pretty uncommon as well.

Where did they ask for visitors to subscribe to newsletters?

A major trend among these major publishers is to build and expand their readership. Many are trying to build subscriber lists or newsletter lists to maintain a connection to their audience.

As most publishers continue to reduce reliance on Google and Facebook for traffic, what can we learn from these publishers about where they try to engage visitors for subscriptions or newsletter sign-ups?
There was a lot of diversity here with no clear winner among these publishers.

The most common place to find these types of calls to action were in the publisher’s sidebar.
This is a stark contrast with — the much more aggressive — pop-up window that was the second-most-popular form of newsletter or subscription sign-up.

The Washington Post is a good example of someone who uses both a fixed menu item and a sidebar call to action for subscription advertisements.

What to look for when testing website layouts

As you would probably imagine, few of these publishers are blindly implementing these practices.
Likely all of these popular publishers are implementing some form of testing to determine which layouts help them deliver the right balance of visitor experience with total website revenue.
This is actually a more difficult equation to solve than most people think.
A/B testing is actually a suboptimal solution for website layout testing if you’re a publisher.

It’s pretty easy to see in this example why simplified testing generally results in untrustworthy results in the world of publishing.
Imagine you test 1 image vs. 2 on a typical article. 60% of the visitors prefer 1 and the other 40% prefer 2.
An A/B test would mean that you would eventually opt to show everyone 1 image instead of 2 (since that’s what the test results showed was best for the majority).
Unfortunately, the 40% of people that may have visited more pages or engaged better with 2 images on the page are now forced into a subpar experience.
The best action for this type of website layout testing is to implement multivariate testing.
Once you have the results of these experiments, publishers should consider visitor segmentation as a means of delivering each visitor a more preferable experience; based on your new testing data.

This is what some of the most sophisticated publishers today are doing. See the example above of a publisher that tests and delivers ad locations differently based on previous visitor tests.
Here’s an easy way to do this form of testing on ads that is easy and free to use. You can also do this more broadly with website layouts here.

What determines a good experience from a bad one?

This is something we covered in a bit more detail here.
There is a phenomenon known as Fake UX. It is essentially the manipulation of objective metrics like time on site to falsely simulate a good visitor experience; despite the experience actually being bad.
For example, time on site being high isn’t necessarily great if the person is spending lots of time waiting for a page to load.
The best way to determine good experiences from great ones is to look at engagement metrics.

• Engaged pageviews per visit
• Engagement time

Both of these metrics help sort through when visitors are scrolling through pages quickly, bouncing quickly from internal pages, or waiting for pages to load.
You can track this automatically for free using this analytics app.

What should we learn from this?

It is obviously very interesting to see what top publishers are implementing on their sites.
This gives us some insight into what these publishers may have tested and ultimately decided worked best for their goals.
What we can’t see are the results of their testing and business motivations behind many of these decisions (are they more focused on subscriptions than visitor revenue).
It is almost always best to do your own website layout testing and to sort through your own visitor segments to identify ways to improve experiences for your visitors and boost overall website revenue.
Questions, thoughts, opinions? Leave them below and I will answer them as they come my way.