By Anne Holland, President
As I reported in a recent column, 3,637 Marketers Reveal Which Email Tests Worked Best
, MarketingSherpa research into 3,637 real-life marketers' tests showed landing page copy tests were the big ROI winner.
On the one hand, this didn’t surprise us because landing page tests have ranked as the highest impact tests in other online media for years. In fact, a December 2006 MarketingSherpa study of the world's heaviest online advertisers revealed that 56% had budgeted significantly for landing page A/B tests in 2007.
However, let's face it -- when you think about email tests, traditionally you think about stuff like subject lines, copy and offer, and text versus HTML.
It's not that the old tests don't work -- 60% or more of marketers say these classic tests are still worth the effort. It's just that landing page copy tests have higher impact. And if you're hoping to increase your email marketing results (much less impress upper management in order to grow your budget for next quarter), why not focus on the email test that will give the biggest bang for your buck?
But, what specific elements should you test for landing pages and email? My top two suggestions are:
#1. Headline copy
Once you have picked the perfect wording for your offer or newsletter lead article headline, *don't* get creative with it. Keep the exact words in your subject line, email headline and landing page. By this, I don’t mean words that mean the same thing, but the exact same words. That way the recipient's flicking, restless eye will immediately see clear visual cues that they are in the right place.
This may not be easy because you have a different amount of space in each -- especially short subject lines versus possibly longer landing page headlines. But if you use the exact same words as the foundation of each, you'll probably get better results than if you had not.
#2. Remove distractions
Pare your landing page to as few visually distracting items as possible so each clickthrough will focus on the path you want them to take. This may mean eliminating extra columns, sacrificing spare navigational bars/buttons and/or cutting unrelated graphics. The fewer distractions, the more chance clickthroughs will spend more time on the page.
If you have a hotlink on that page or an ad or anything that's not what they clicked through to see, then it should be there on purpose. And if it's there on purpose you should be measuring its value. Does your organization make more money in the long run if that extra element is there?
A practical suggestion for running tests if your Web department can't help you:
Now all of this can be tough because your email landing pages may be using Web page templates dictated by your main site design. You may not have the power to change your Web templates on the fly for particular emails. In fact, if you're like most marketers, you'll have to get in a long line at the IT or Web department and wait. When you're running an email program, you need more flexibility and speed than your site's content management system can deal with at present.
My suggestion: try a few campaigns using super-cheap, alternative Web page applications. What are these? You can find blogging software, online survey form software and landing page testing software from a variety of providers online that cost less than $50 a month. With a little ingenuity, all of these can be set up to look very much like a page on your own Web site as you might want it to appear for email clicks.
If you want to collect registrations or opt-ins, then use surveying software. Instead of 'survey" questions, you're asking for contact info. If you want to post an article, then use blogging software. If you want to make an ecommerce offer, then use landing page on-the-fly software with a graphic for your "add to cart" button that ties into your main shopping cart. (Note: You may need a little help from the IT department with the latter.)
First, create a "template" that matches your brand on the new page:
- Stick your logo in the upper left corner.
- Pick a font and colors that match your site.
- Pop in your copyright and contacts in fine print near the bottom of the page.
Then, use the rest of the page for your content and hotlinks.
Once you have tested tweaking your copy and design on these low-budget landing pages, you can use the results to convince management to give your Web design requests higher standing. The goal is to ultimately get the budget and IT time to integrate your landing pages back into the main site content management system.
Why? Well, because it's easier for the marketing team to run a single system than multiple siloed systems. Plus, your analytics will be better if they are unified.
By the way -- yes, my suggestions are completely based on real life. Both I and many marketers I've met (including ones from far bigger brand name companies than you might expect) have conducted guerrilla-style landing page builds and tests on our own when the Web department didn't have any extra resources or time.Useful links related to this article
MarketingSherpa's Email Marketing Benchmark Guide 2007: