I'm a huge fan of A/B split and multivariate testing. It simply works. And that means more money at the end of the day.

But because this can be a complicated undertaking depending on the scope of your testing strategy, there are often pages that get left behind.

While the usual suspects (eg. your homepage, your registration page, etc.) are hard to forget, one page that I've personally left behind when I first started employing A/B split and multivariate testing is one page you don't want to ignore: your order form.

I was reminded of this when I read that ClickBank, a popular affiliate program provider, is going to be launching a new order form. The announcement explains why:

As part of our recent update to the look and feel of the ClickBank website, we are making visual changes to our order form. A brief but statistically significant split test was recently conducted, and the modified order form outperformed the current order form. Due to these positive results, we will begin using a modified version of our order form on Wednesday, June 17th.

Playing around with an order form seems like risky stuff, especially when you're running an established business. That's why I was initially hesitant to apply multivariate to my order form pages. But depending on the type of business you run, all of your good work elsewhere can easily be diminished if your order form isn't optimized.

Here are some things that I've personally found to be useful when experimenting with different versions of order form pages:

  • Layout. It should go without saying, but how you lay out your order form can play a huge role in closing a sale. Something clean and professional looking is obviously going to perform better than something cluttered and ugly. Hopefully you're closer to the former but little tweaks and improvements can make a world of difference.
  • Messaging. What you say and where you say it on your order form can mean the difference between a completed sale and a lost sale.
  • Imagery. On an order form, there are two areas I've found that can influence a sale when it comes to imagery. First, utilitarian imagery (logos of accepted payment methods, security seals, etc.). Second, design imagery. In the latter, you may be surprised to see how the use of different colors or image-based buttons instead of HTML buttons can actually make a difference.
  • Help. The order form is typically the most important page in the checkout process and if the user becomes confused or frustrated, they will often bail. So make sure you're helping them get through the process as quickly and efficiently as possible. Experimentation with providing explanatory information (eg. how to locate the CSV code on a credit card) and error checking (eg. making sure the user knows how to complete the order form right the first time) can be useful.

While experimenting with your order form may seem dangerous at first, if you do it well and you do it early your bank account will thank you for it.

Photo credit: mil8 via Flickr.

Patricio Robles

Published 12 June, 2009 by Patricio Robles

Patricio Robles is a tech reporter at Econsultancy. Follow him on Twitter.

2642 more posts from this author

You might be interested in

Comments (5)



A/B testing is a great way to actually check the efficacy of your style and layout.  The problem with most web projects is that unless they are an on-going concern they have a budget for one design, and even then they don't have the volume of budget to make sure that the site is designed to it's most efficient point.  The result is often half-cock sites that don't fullfil potential.

about 9 years ago

Tim Watson

Tim Watson, Email Marketing Consultant at Zettasphere

Testing is a vital part of website, email and digital marketing.

For an online transaction developing trust at the point of order is important and as with all optimisation colour, font size and position all impact results.

Don't forget to check results are statistically significant. t-Test formulae are a common approach to this. Without checking the stats you may make wrong choices as to what is performing better.

about 9 years ago

Matt Isaacs

Matt Isaacs, Founding Partner & CEO at Essence

@Patricio always pleased to hear people actually doing A/B testing (or indeed Multi Variate or other forms of testing).  I find organisations talk about it a lot and do it very little - often because it is Commercial/Marketing that want to do it but Tech teams that have to implement and it's always at the back of the queue.  But when you've done it a few times and realised that it really is pretty simple it seems extraordinary that folk struggle with it so much.

Anyway, I have a few other practical checkout/order form tips (sorry if I'm blurring the lines here between checkout and order form):

  • Progress Bars can be very useful to show the user where they are in the order form process, particularly where the order form is quite long by necessity (e.g. for sales that require a credit check such as contract mobile phones)
  • Product Images have been repeatedly shown to improve order form conversion (really applies to basket not order form)
  • Delivery Information - while you may have already had this information available earlier in the process there are always folk who only think about it at the last moment - ideally without clicking off the order form page (e.g. an AJAX overlay).  (again, basket more than order form but could be both)

about 9 years ago

Matt Isaacs

Matt Isaacs, Founding Partner & CEO at Essence

@Tim good reminder about stats tests.  I come across clients who say 'we are buliding our own split testing and multivariate platform'....now why would you do that?  Google gives you a pretty good one for free and if you want something more advanced the commercial platform options are relatively 'cheap' these days.  Whatever platform you choose, statistical skills are scarce in our industry but crucial to ensure you are 'doing it right'.

@Vincent also right about the issue with one off redesign efforts.  I still feel such efforts are a valuable part of the design cycle but best in class organisations will adopt A/B testing and MVT as a core part of their business and use them to test anything/everything such as:

  • page layout
  • page design
  • customer journeys
  • cross site content (content which is best displayed across multiple pages)
  • product pricing
  • customer promotions
  • checkout and order forms (thank you Patricio)
  • ....

about 9 years ago


Alasdair Bailey, Solutions Deployment Manager at Maxymiser

Matt - some very valid points there, to put some numbers to your top two, the design and layout of progress bars in checkouts or insurance / financial application processes can have up to a 10% impact on completion.

The addition of photographs to basket pages can add 5% or sometimes more to conversion rates to our experience. There are the outlyers so it's by no means fair to say that it will be the case every time (hence the 'testing' part) but they're good points to look at after the landing pages are addressed.

about 9 years ago

Save or Cancel

Enjoying this article?

Get more just like this, delivered to your inbox.

Keep up to date with the latest analysis, inspiration and learning from the Econsultancy blog with our free Digital Pulse newsletter. You will receive a hand-picked digest of the latest and greatest articles, as well as snippets of new market data, best practice guides and trends research.