3 Steps to a Killer Conversion Rate
Published: September 13, 2012
Author: Theresa Baiocco
Are you guilty of believing – or worse, spreading – these common Conversion Rate Optimization myths?
– Thinking a site merely needs an updated design in order to convert higher (Do you also think a fresh coat of paint will make your car faster, then?)
– Thinking landing page optimization is the same thing as Conversion Rate Optimization
– Thinking you just need to learn the mechanics of running an A/B or MVT test
The bad news is that there’s more to optimizing conversion than that. The good news? I’ve condensed it into three steps for you.
So what are the three steps you need to take for a killer conversion rate?
1. Define the business’ Key Performance Indicators (KPIs)
I can’t stress the importance of this first step enough, yet this is the first thing people miss when there’re merely talking about landing page optimization. There are two parts to this:
– Define what drives revenue
In order to determine which pages to optimize, you first have to know what drives revenue. Then you have to determine your site’s specific activity – or goal – that moves visitors toward your revenue drivers. Is it:
– A form submission?
– Joining a membership?
– Buying something directly from the site?
– Driving to your brick-and-mortar store?
Write out your site’s macro and micro goals. If they don’t relate to your overall business objectives, then you’ve already gotten off-track.
– Configure these as goals in analytics
Next, configure those goals in analytics so you can measure the completion rate. Most people who have installed analytics have not configured their goals.
If not, do that now.
2. Figure out what’s preventing people from doing the revenue-driving activities on your site
– Understand your site’s visitors
News flash: not everyone who comes to your site has the same motivation, objective, knowledge, or buying style. Some of them are just researching, some are ready to buy, and some landed on your site by mistake and will never convert. So the first step in figuring out what’s preventing people from doing the revenue-driving activities on your site is to get inside the head of your site’s visitors.
Online marketing pioneer Bryan Eisenberg teaches that there are four personas that your site’s pages should consider. The personas are categorized by whether they make quick or deliberate decisions, and whether their decisions are based on logic or emotion:
You not only need to address the specific needs of each persona, but you also need to consider what stage they’re at in the buying cycle: are they just starting their research? Are they comparing you to a short list of your competitors? Or do they have their credit card in hand, ready to buy?
(To learn more about how to appeal to the different personas in various stages of the buying cycle, I highly recommend going through Bryan Eisenberg’s 12-week Master Certification program in Conversion through Market Motive.)
– Use analytics to find problem pages
The next part to figuring out what’s preventing people from doing revenue-driving activities on your site is to find problem pages in analytics. You DID configure your goals, right? Good.
This is another massive area that gets overlooked by people focusing on landing page optimization, rather than conversion rate optimization.
Use a filter like the Goal Flow in Google Analytics to find the pages where people dropped out of your conversion funnel. This tells you how to prioritize your tests. In the example below, a lot of people are leaving the “Education Page” without completing the goal of signing up for a webinar, so this is an important page to improve:
Other good pages to add to the list are top entrance pages with high bounce rates, and pages with high exit rates.
– Do user testing to uncover the specific problems on those pages
Now that you have a list of pages that are causing people not to convert, you need to drill down on those pages to figure out why. Look at those problem pages from a conversion framework. (Recommended reading on a conversion framework: Your Customer Creation Equation, by Brian Massey.)
After analyzing a page in depth, however, you run the risk of getting tunnel vision. Take your opinion out of it by running the page through various types of usability tests. This is the sexy side of conversion rate optimization. Here are a few of my favorite tools that won’t break the bank:
– Do a cross-browser test to make sure your page renders decently in all the browsers. After all, if your primary call to action is invisible in IE, that might hinder conversions a tad, don’t you think? (Bonus: check in analytics first to see the browsers, operating systems, and screen resolutions that most of your site’s visitors use.)
– Run a quick attention map to see whether your call to action is obvious on the page. Two good and affordable sites for running attention maps are Attention Wizard and Feng-gui.
– Gather the first impression of random, anonymous people with a Five-Second Test. If they don’t know what your page is about and what they’re supposed to do on it, fixing that should become a high priority on your list of tests.
– Order a couple of user tests, where random, anonymous people record their screens and talk aloud while trying to complete tasks you’ve given them. Warning: it’s kind of like watching a horror film, where you shout at the girl to not go in the cellar: you’ll be yelling at the user video, trying to get the tester to click on that link – right over there! But the tester won’t be able to hear you, and neither can the rest of your site’s visitors. That’s why, as painful as it is, it’s critical to watch where people get confused or frustrated on your site. And if you sell something as complex as HIPAA compliance, you can ask for testers who have some familiarity with it. Don’t avoid user testing because you sell niche products or services.
3. Test for improvement
To recap, you’ve:
– Defined your KPIs
– Configured goals in analytics
– Gotten inside the head of your site’s visitors
– Used analytics to pinpoint the problem pages
– Found the specific problems on those pages based on a conversion framework and through various types of user feedback
NOW you’re ready to test!
Google has a free testing tool: formerly Google Website Optimizer, it’s now called Content Experiments, and you can get to it directly from within Google Analytics.
The nice thing about Content Experiments is that you can analyze the effect of the test on different segments in your analytics after the test has run. Many other testing platforms make you define segments in advance.
And it’s nice that it’s free, too.
But it has a lot of limitations. For example,
– You can’t name specific traffic segments/channels to include or exclude from a test
– You can only run up to 6 variations of each page
– You can only do A/B testing, not multivariate
Still, it’s a good place to start if you’ve never run an A/B test.
But when you’ve reached the limits of what Content Experiments can do, you’ll want to graduate to a more powerful testing platform where you can do multivariate tests as well, such as Convert.com. This is my favorite testing tool for ecommerce sites because:
– It integrates with Google Analytics to track revenue, transactions, and total number of items ordered.
– It has the most options of any of the testing tools in its price range for segmenting specific types of visitors, such as by time on site, geography, pages visited, traffic source, etc.
– I’m not a designer or a developer, so I love their WYSIWYG interface that lets me drag and drop elements around the page.
If you followed the steps above to align your tests with your business objectives, and you dug deep to understand how you could help your site’s visitors to achieve their objectives, you most likely had a winning test.
Sadly, you’re not done.
Now it’s time to go back up to step one and repeat the whole process again. Re-evaluate your KPIs. Gather more data. Figure out what else can be preventing your site’s visitors from converting. And test again and again and again.