Conversion Rate Optimisation Case Study: Bible Timeline

How would you improve this page?

This is what the Bible Timeline homepage looked like before we started testing. Have a quick look down the page, and see if you can pick out any things you might change in an effort to increase the number of people who buy the product.

Identifying ‘Rocks’

Initially, we looked at the page as a whole, and came up with a list of ‘rocks’ – page elements which might stand in the way of the user reaching the desired goal of the page (in this case, clicking on the ‘buy’ link).

Our general approach to optimisation is to move the biggest rocks out of the path first. Then, we might go back and clear away a few of the next biggest.

In our review, the most important things that stood out were:

• Lack of a clear call to action. It’s not really obvious that the ‘order now’ link is a link.
• Main product image is not engaging, doesn’t give a clear picture of what the product is.
• A general ‘untidiness’ could make the page appear unprofessional at first glance (although a professional design is not always beneficial – it depends on your offering).
• The page container is very wide, especially on a large screen, causing very long lines of text, which are less readable than shorter lines.
• Headings are random sizes and dark blue, the usual colour of links, which could be confusing.

The old web design model

In the past we might have taken all of this information and come up with a new design, which aimed to fix all of these issues in one fell swoop.

The problem with this old model is that whatever design you come up with, it is merely your best guess at what will be most effective, and you have no way of knowing whether it is more effective than any other alternative. (Read more in "Web Design is Dead".)

The new model – Test and Measure

A much better solution is to test alternative designs against each other, and actually measure the increase (or decrease) in conversion.

Then you can refine the more successful overall direction by changing individual design elements. That way you can see conclusively which elements have a positive effect on conversion, which you will retain, and discard the ones that don’t.

Experiment #1 – Call to Action Buttons

Within a week we saw an increase in sales well over our normal 20% over last year. Throughout December 2009 – traditionally our largest sales month, sales were up nearly double last year.

Margaret Agard – Bible Timeline

What we tested

• Changing the ‘order now’ links to large orange buttons, which tell you exactly what they do.

Why

• If you want someone to take an action, make it totally obvious where and how they do it.
• While the large text and yellow background are visually bold, they didn’t stand out from the page.
• Nothing says "click me" like a 3D button, so we used a glassy button with a highlight colour not found elsewhere on the page, to indicate this is something special.

This shows that within 1 week of starting this experiment, there was an observed improvement of 146%, and the cilck-through rate had gone from around 2.1% to 5.2%

Experiment #2 – Revised Product Image

What we tested

• Added a closeup section to the main product image, and arrows indicating the size.

Why

• Potential customers can’t really get what’s on the chart from the zoomed-out image.
• The chart is actually almost four feet by over three feet in size (it’s big!), but you can’t tell from the image.
• We figured that showing more information about the product would help make it feel more tangible.

These results show -8.9% improvement after a week, but also a relevance rating of 0/5, which means that the new image was almost insignificant to the conversion rate.

We also tried fixing the width of the main container to 960 pixels. This also proved to be irrelevant to conversions.

Experiment #3 – Add a Human

What we tested

• Added a young woman to the image to show scale and show people using the product.

Why

• We thought this would make the product seem more real, and photos of people often help people to engage emotionally, which is what drives us to buy.

These results show a 38.8% improvement with the new image. Not a conclusive winner at this point, but still worth keeping.

Experiment #4 – Clearer Header

This was a multivariate test, which means we tested two things simultaneously. This allows you to see which combination of different page elements perform best, as well as the page elements on their own.

What we tested

• Changed the main heading and subhead to Georgia font, made them bigger and clearer.
• Left-aligned the benefits list with the headings, and changed the colour to match the product image.
• Also moved the ‘See Bible and World History’ line to the bottom, as the other benefits are more engaging.

Why

• Serif font may read as more authoritative (in large size heading).
• Larger main heading is more noticeable.
• Neater alignment suggests attention to detail, which is a positive attribute for this product.

This shows that Combination 3 (The new header you can see in the image above) converted 4 times as many visitors as the original, and is a conclusively proven winner, with a 99.2% chance of beating the original version.

Experiment #5 – Further Readability

Following this, we decided to carry the new neater layout through to the rest of the page. We standardised all the second level headings, increased vertical spacing to improve readability, and gave the testimonials a standard style. The result can be seen below.

Surprisingly for me, this change only showed an 8% increase in conversions, and no conclusive winner. Possible reasons for this are that the headings might be too light, making the page difficult to scan, and the ‘buy now’ area could be improved still, perhaps by combining the price section and the button into a coupon style callout.

What’s next?

We are still running tests on the Bible Timeline page in an effort to increase the conversion rate still further.

We have already more than doubled conversions, but there is almost limitless potential for what can be tested and measured.

Your site design will never be perfect (until your conversion rate is 100%!), so we would advise all site owners to test constantly in order to achieve the highest possible conversion rate.

Real World Benefits

It’s all very well saying ‘this combination performed 210% better than the original’, but what does that mean in terms of sales and profits (which is of course the real motivation behind all of this)?

Margaret and Bill from Bible Timeline were kind enough to share with us their sales chart for the last 6 years, and you can see that since we started testing in November 2009, their sales have seen a significant spike.

My husband and I have been marketing a self-published product, The Amazing Bible Timeline, on the Internet for 12 years. What do we want from that? We want security and freedom – is that asking too much from life?

We’re retired and are living on US Social Security, along with a corporate pension that’s in danger of disappearing thanks to the stock market plunge and our web business income. The web business is key to our security and freedom.

Freedom means being able to afford to pay other people to write our content, pack and ship our product.

We’ve watched sales grow an average of 20% a year but never managed to make the big changes that would take us to the next level of sales – double or even triple our current level that would give us the freedom and security we’re seeking.

For the last two years we’ve tried coaching programs – StomperNet, Perry Marshall’s Bobsled run, Internet Marketing Tips – and signed up for every Internet Marketing course, newsletter and audio training we thought seemed legitimate in an effort to increase those sales over the 20% a year that happens whether we do anything or not. We implemented all the suggestions we were given. None helped. Or if they helped, it was simply in making sure we didn’t lose ground.

Then we tried Scratchmedia. Bingo! In November 2009 the team worked on our sales page. Within a week we saw an increase in sales well over our normal 20% over last year. Throughout December 2009 – traditionally our largest sales month, sales were up nearly double last year. Since only half our sales come from the web page the team worked on we’re very happy. We’re rolling the changes out to our other sales page now. We expect this huge increase in sales will keep up since the changes are fundamental.

The extra sales have already allowed us to stash some funds in savings, hire a free-lance person to write content and another person to handle shipping while we take a month’s vacation.

Thanks Ben, Dan and team!

Margaret Agard – Bible Timeline