What to Test on Your Pages for Best Results


Testing often falls into the “boring but necessary” category.

Until you test something that heavily outperforms your previous best. Then you get excited and remember that there’s a good reason to be testing.

If you’re not sure what to test, here’s a few things that have worked since before the web was born and still work today:

  • Headline or subject line. Without doubt, the most important part of any communication. You know how many emails you skim past if the subject doesn’t immediately grab you. And you know that you’ll often click something other than the top result in a search if the headline looks more interesting. Movies know that the title (a headline by any other name) makes a world of difference but for a film the title is almost always short and snappy. For a subject line it can be a bit longer & for a web page headline you’ve got around 65 characters (maybe slightly more) to get your message across. If you’re going to test anything, test the headline first. Change Capitalisation. Change the colour. Change the font size. Then tinker with the wording or try something completely different.
  • Who you’re targeting. Yes, I’ve put this second to the headline. It possibly should be the first thing you test. In mail order, it’s thought that the targeting of the list is the biggest part of the equation and on the web that’s similar. The reason I’ve put it second is that a lot of your targeting will be done for you – either by Google if you’re lucky enough to show up in some relevant search results or by the places you choose to promote. Whilst there may well be a market for chocolate cake recipes in a weight loss group, you probably wouldn’t be very popular promoting them as you could be leading people away from the straight and narrow. But turn that round to a tasty chocolate cake without the calorie burden and you’d probably be onto a winner. So tweak your message for your target market – that comes back to headlines again.
  • The rest of the message. Once you’ve done your main headline and other testing, it’s worth taking a look at the rest of your message. It should already be congruent with the headline – hopefully that’s obvious. Read just your sub-headings and check that they lead people logically through your message. Check your call to action – Gary Halbert always taught that you should make this as clear as possible. Because most of his sales letters were before the internet, he often explained what would happen when you called in. You could apply the same logic on the web – “when you click the buy button, you’ll be taken to a page that asks you to enter your name & email address (so we know where to send the download link) and put in your Paypal details. As soon as Paypal verify your payment, you’ll be instantly given your download”. I’ve not tested that (yet) but having just remembered it as I type this, it’s now on my to-do list.
  • Pricing. This could maybe be above the previous bullet point. There’s no logic in pricing. A $5,000 Rolex tells you the same time as a $50 Sekonda. OK, there’s a design difference but not much and the main function (the hands that point to the hours and minutes and seconds) is basically identical. We use pricing as a short cut. If something is “too cheap” then we think it’s not as good as something that’s near enough the same but more expensive. In internet marketing, people often get much better results from near enough the same products but with a difference in price. That’s a conundrum for me – I like to make my products affordable but the trade off is that people don’t necessarily take them as seriously. I’d like to think that my Internet Marketing for Busy People adds a lot more value than the mere dollar it’s priced at. It’s very tempting to make it a $97 product with the only real difference being a reporting/accountability element (homework for want of a better word). My guess is that it would get much better results because people would treat it more seriously. And if I priced it at $1,000 – maybe with four weekly follow up sessions – the results would be even better for the handful of people who bought it. Price can act as a major differentiator (check Apple products if you don’t believe that) and it’s not always best to do a race to the bottom of the pricing ladder.

If you’ve not tested anything for a while, pick something to test.

Do a split test if you’re getting enough traffic.

Or a new squeeze page.

Or anything else that allows you to test.

Small differences mount up – I was taught this at an internet marketing seminar years ago. The speaker had tested hundreds of small – seemingly insignificant – changes to their sales pages. Everything from headlines to font colours and sizes through to where on the page the credit card logos were displayed. Each change was relatively minor. But combined they meant the sales more than doubled from the same traffic.

Not something to be sniffed at.