• NZ 03-441-3223

The Last 3 Steps in Wine Landing Page Optimization

This is part of a series on Wine and Landing Page Optimization (LPO). The previous posts outlined why you should do LPO and the first steps of how to go about testing. This post will looks at installation and the LPO process.

Step 4 Install Google Website Optimizer (GWO)

GWO logoYou have two choices of the sort of testing you can do: A/B Split test or Multivariate test.

A/B

Splitting the traffic between 2 or more published pages and tracking which one gets the most conversions.

Multivariate

Carving up a page into a few sections, then trying different content in each section to create different versions of the page.

If you are a small retailer, then you shouldn’t do multivariate tests because:

  • you probably don’t have the traffic to complete the test in a reasonable period of time
  • it is much harder to set up
  • you’ll get the best performance increase by concentrating on a few big things not lots of smaller things.

Here is an excellent Google video giving you an overview of how to set up a simple A/B split test.

GWO has lots of other videos and resources to help you get started.

What website software has this integrated

Most web developers and/or ecommerce software can integrate Google Website Optimizer.

For a simple website you or your web developer can paste some code into the individual web pages. However many software packages make it simpler by providing fields in the administration backend of the website where you paste the code – a much more intuitive process to most people.

The real question is whether they are fully integrated or whether they need to be do some coding. One ecommerce package that I know well, Magento, includes ready-made fields. The software I use for this blog, WordPress, has a plugin you can use to make GWO work.

Cost and Time

That annoying response – “it depends“.

But as you can see above some packages are integrated, making the process simple. It also depends on whether you have decided exactly what you want to test or whether you need assistance on choosing test elements.

It is going to take some time if you do it yourself as you need to follow the steps in the previous post before you actually get to the implementation of the GWO. Then you need to wait while enough conversions occur for GWO to make a call as to the winner of the test.

AB testing

A/B tester starts to go a little crazy on his whiteboard...

Step 5 “Beat the control”

This blog is for small wine retailers. However it’s read by all sorts of companies so here are some rules of thumb.

Your conversion page

If you have more than 10 sales per day, then make this your eCommerce “thank you / payment confirmed” page.

If you have less than 10 sales then I (controversially) recommend your shopping cart page. I say controversially because I’ve implicitly traded off time of testing for accuracy of testing. In the business world we don’t have the time that statisticians would like us to have. You’ll probably check every week or perhaps month to see if GWO has made a call on a winner.

Your conversion measurement

To start out just use conversion rate % (conversion #s / unique visitors). It is preferable that you put dollar figures to this but let’s walk before you run.

If you’re technically able, are briefing a LPO company and/or have many sales per day then you can use revenue per visitor if your markup is the similar across products. Otherwise it’s preferable to use profit per visitor if your markup is not similar.

Decide on the page

Remember the 80/20 Rule I keep harping on about like a scratched record ;)? Apply it to choosing the pages you test. Use Google Analytics or your web logs to find the top product (or post) pages. If you’re running Google Adwords then the ad’s landing page is a good one to optimize.

I suggest you just choose one at first, but aim to do this for the 20% of products that make up 80% of your sales.

Decide on your testing elements

I’m a big proponent of blogs in addition to (and sometimes instead of) eCommerce sites. The most important part of a blog post is the headline (which is why I sometimes have crazy headlines!). So test headlines if you’re blogging.

If you have an ecommerce site then you may be able to change a sub heading. Otherwise check out Step 3 in the previous post for what to test, I suggest product description may be the easiest thing you can change.

Do it

  • Create the original page known as the “control”, “baseline” or “champion“.Wacky races - les fous du volant
  • Create the alternative page known “variant” or “challenger“.
  • Follow the GWO steps as per the video above.
  • Try to manage your impatience as the days go by and GWO still hasn’t made a call. At least manage it better than I do ;).

Got the results? Great.

The violent bit

Have you beaten the control? If so the challenger page becomes the new control.

If not, one theory down let’s try something else. Record it in a testing document so we don’t try the same test again, and move on to the next test.

Step 6 The Plan

Most people would have this as Step 1, but I think the best way to learn is put your “toe into the water” and then step back and plan.

Time to look at the forest not the trees. Make a list of all the pages and all the elements you want to test. Prioritise them and knock ’em off.

Do not just get lost in one simple test for one element on one page. You want to spend the next 12 months making lots of changes to lots of pages to get a big overall improvement.

It becomes just another thing you do every week or month.Testing Thursday

What’s your thoughts? Would you do it differently?

Photos courtesy of Testing Thursdays , A/B testing, Wacky Races

Speak Your Mind

*