Much is written about conversions from various types of page real estate, but few share that data.

Working with the team at Live Casino we have spent the past 12 months looking at how different call to actions affect click through and uncovered some interesting findings around changing behaviours in how we react.

Is click-through banner advertising on the decline? Is a button better than a text link? All of this and more is answered below.

A conversion can come in many forms for a website owner, from a simple click on a link all the way to a large product purchase – whatever form it’s in, it’s a visitor action that makes you money and drives your website forward.

Most online business-focused blogs talk about how to gain more traffic one way or the other. From Google and affiliate referrals all the way to social networking sites like Facebook.

However, not enough talk about converting the visitors the site already has. Improving aspects of your site to aid conversion could be the difference between acquiring 10 customers and 20 customers a day, doubling your income without doubling your traffic.


For years, as far back as display advertising and affiliate websites have existed, banners have been an essential advertising medium. However, during recent years it has become increasingly reported that banner advertising is a dying method of gaining traffic, something we wanted to put to the test.

Below you’ll see the results of a test we carried out in conjunction with the team at Live Casino into how visitors react, in real time, to key conversion opportunities, including clicking banners, buttons and text links to attempt to put some data behind the suppositions. It was an exercise that we ran for 12 months in an attempt to get the clearest data possible.


Of the sites the business owns we decided to use as it features not just its own conversion buttons and text links but critical banner advertising also. We tested the theory with a sample of approximately 10,000 visitors over a 12 month period.

To extract the data we used a number of tools, including eye tracking software and used that to correlate data with what we were seeing in Analytics and other tracking software. Together they allowed us to see exactly which links are clicked the most, which banners are effective and what will help produce the best conversion rates.

What is clicked more?

The only way to statistically define which promotional medium is more successful is by tracking the outbound clicks from specific webpages and whether they were made through a banner, text link or button. Below you can see a summary table of our findings along with more detailed information beneath.

  • Banner: outbound click percentage:17%.
  • Button: 43%.
  • Text Link: 34%.

Text links/buttons vs banners

At the beginning of the research piece we were unsure as to what we would discover. What we found is that banners certainly do seem to be losing clicks as time passes.

That’s probably not surprising but what was more interesting is that the rate of decline seems to be steeper than we predicted.

 eye tracking

Across you’ll notice that the full banner at the top is placed most prominently, along with the medium rectangle banner placed in the sidebar to the right. Both banners are placed in positions which are typically more obvious than any of the text links or buttons.

Despite the more prominent positioning, we found that across the website 83% of outbound clicks were made using buttons and text links, whilst just 17% of outbound clicks were from banners.

The huge difference between buttons, text links and banners was not too surprising to us; however what was interesting was that the number of outbound clicks on buttons and text links were up 4% on the same 12 month period only six month previously.

This suggests that the already lessening power of banners could actually be diminishing further.

live casino

Banner removal

Even though the customer’s preference swayed towards text links and buttons, when removing all banners for the website (below image), outbound clicks dropped by 6%. This shows us banners are still playing their part as a clickable medium, perhaps to more inexperienced internet users, even if it isn’t the most effective strategy.

Buttons vs text links

Breaking this down further we found that the green buttons on the website, such as the ones under the advertiser logo and at the bottom of the page, resulted in 43 % of the clicks.

Whilst the other 34% went to the in-content text links, which shows us a large percentage of users read our content and use it as a guide before deciding further action to take.

Does design and context have an influence?

The natural progession from these findings was to properly test which type of banner/button/text link convert better and while a subject such as this has endless variants to take into consideration, from the specific page the user is on or even the brightness of the user screen could make one banner more appealing than another we still felt it worthy of monitoring to see what the difference could look like.

To ensure that this was digestible we are displaying two examples of each type below, one which proved to convert the best and one the worst.

Banner types clicked more/less

Worst = Guruplay, 4% click through

Despite being a casino the business earns decent revenue through, Guruplay’s 468×90 banner was the worst converting in this position – this was not surprising when you broke down the banner elements down.

 guru banner

Key characteristics:

  • No stand out colours.
  • No call to action.
  • No obvious message.
  • 50% of banner on branding with no real stand out.

Best = William Hill, 32% click through

William Hill, one of the largest online casinos on the web, had the best banners in regards to click through rates.

william hill banner

Key characteristics:

  • Stand out colours used, blue, green, gold.
  • Call to action, using a clear and different green colour.
  • Clear message, they run ‘THE WORLD’S FINEST LIVE CASINO’ with encouraging words emphasised.
  • 20% of the banner used for branding however very clear and different from the rest of the banner. Brand here clearly made a large difference.

Button types clicked more/less

The next stage was to test buttons. To do this the Live Casino team decided to narrow the trial on our recommendation, changing the colour of the buttons across the sites to see how it affected the click through rate.

We trialled grey, black, blue, red, yellow, purple and compared the results to the original green banner. Here is what we found:

Worst = red banners, 29% click through.

red buttons

Best = green banners 43% click through.


As you can see from the above, the green buttons were clicked significantly more than the red. We feel this is because of the mental association with the colours. Red is regarded as the colour for stop universally, whether that is at traffic lights, a no-entry sign or when your x-box dies.

Green on the other side is totally the opposite, green means everything is fine and you’re welcome to proceed, whether this means on the road or on your x-box once again.

This part of the research has actually provided us with the opportunity to improve our own websites in the future as certain sections still use red buttons/links.

Interestingly this is a view challenged in the print space. Magazine and newspaper wisdom suggests that Green covers do not have stand out on the news stand in the same way red and other key primary colours do and is another example of how print design rules differ from online.

Text links

Once we had established colour as a key conversion factor we decided to look at what the main element of attraction behind a text link to a user would be. Our initial view was that contextual relevance would be the key factor here.

To test we picked the link underneath the ‘Live Roulette’ section on every review page which reads ‘Play Live Roulette at *Casino name* by Clicking Here’.

This link is clearly targeted towards people who have finished reading the live roulette section and will want to follow that up by playing the game. Mentioning the game name and the casino provides the user with reassurance that they’ll be arriving at the correct location.

Upon changing the link to simply ‘Play game now’ we saw nearly 50% (48%) of clicks on that link stop. This shows that the key factor behind the link is contextual relevance.


The backbone of this post is research based, however as applicable to our own sites we have given our personal opinions on several key findings of this research below.

Why are people clicking on buttons and text links?


Speak to any site conversion specialist and you’ll always find that they consider trust one of the most important elements when it comes to external linking. Put simply, the user will always click what they trust more.

A link within or just below text is trusted by users visiting the website as it was usually placed as a point of reference by the author/creator of the website.

However, the average internet user is savvy to banners and their use and they will usually discard this information, whether the advertisement is relative or not purely because they know it’s a third party advertisement.

Internet intelligence

Arguably one of the biggest reasons is the dynamics of internet user intelligence. Years of forced banner advertising and bombardment from advertisers, mixed with spamming have created a scared-to-click mentality which has almost rendered banners useless to an ever growing population of banner-blocked-out eyes.

Whether you’re new to the web or not, you’re more than likely to have heard about banners and whether they should be trusted.

The average Facebook savvy teenager, with no real knowledge of computers, has been growing up in the last few years learning about the general do’s and don’ts whilst surfing online– whereas his aunt Pauline down the road is still confused as to what this e-mail attachment is.

So what are banners still good for?

It’s important when reading a post like this that you must interpret it properly and understand that very few data experiments are without flaw. Everything we did here is open to questioining. Of that there is little doubt. But the findings certainly backed up previously held beliefs.

What it also did was help the site improve its own key conversion metrics and also made it very clear to the team that key banner inventory space could soon be put to better use.

Our current view is that if your site hosts banners they are still very important and shouldn’t be removed. Banners still play a significant role in advertising and have secured their place on the web.

Live Casino would be 6% down on clicks and the associated revenue that goes with that if they were removed right now, but how long it will be until that position becomes unfeasible is one we will continue to monitor.

Here are a few other reasons to keep displaying banners above and beyond pure clicks:


Many users on the web know exactly what they want, and many are shrewd enough to know when a banner will lead them to a product they want over a spam ad which may infect their computer. For these users banners are handy pieces of equipment and will be clicked.


Whether a banner is clicked or not it is usually seen and acknowledged by the brain. This is a form of advertising in itself, usually improved with repetition.

Your brain stores a lot more information than you might think, and when the time comes that you’re looking for something related to that banner, your brain will scan itself for similarities, which will return the brand you saw once on a banner.

There is, therefore, definite brand value to be had.

Complimenting content

In many cases, banners will complement the content shown surrounding them. In our case, we are advertising William Hill casino and therefore display William Hill casino banners.

Though not many users necessarily click on the banners compared with the buttons or text-links, the banners add a sense of branding to the page and reassure the consumer they are on the right track. It also helps instill trust.

This does directly relate to targeting, but works a treat for branding purposes too. Much like when site takeovers are implemented and you get a real feel for what the adverts in coalition are portraying.