Get the best website conversion optimization results with the right process

Harness the value of marketing insights from controlled testing

Are you truly leveraging the potential of your traffic?  I’m going to show you why most companies aren’t and how you can today, with two phases.

Editor's note: Chris Goward, the author of this post gave a well received webcast on website conversion on 28th June 2013. If you didn't make it, we hope you find his advice useful. His deck on conversion for mobile sites on Slideshare is also worth checking out.

Phase one: commit to conversion optimization as a strategy

To tell you that your website is under performing is a pretty bold claim. I may have never even seen your website, much less analysed your performance metrics. Nevertheless, I can confidently tell you exactly that: your website is under performing its potential.

I can say this because I have never seen a website that couldn’t be improved. In fact, the best online companies in the world are committed to continuous improvement on their websites.

Some of the best-known examples include Google constantly tweaking and testing its algorithm and website design, Facebook testing, releasing, and modifying new features rapidly, and Amazon, which is well known for evolving its website through testing.

More important, let’s think about you. What is your conversion rate for new visitors right now? 1%, 3%, 20% or 30%? Whatever it is, I’m willing to bet that you don’t have a 100%  conversion rate.

Promises can be dangerous, but I can safely say that the average conversion rate lift we’ve seen at WiderFunnel for our clients’ tests is 23.1% for e-commerce and 49.0% for lead generation websites.

Using the LIFT model to improve your conversion rate

The secret to getting great test results is one thing: creating great hypotheses!

One of the tools we use is called the LIFT model™ conversion rate optimization framework to construct task-based hypotheses. I’ll walk you through how to evaluate your webpage with the LIFT Model and create your test hypotheses for improving your website.

 The LIFT Model has six parts

•  1. Find your best Value Proposition

The full set of perceived benefits and costs, in the prospect’s mind, of taking your call to action. Your value proposition is the vehicle that provides the potential for your conversion rate, making it the most important of the six conversion factors in the LIFT Model. The other five factors are either conversion drivers or inhibitors.

• 2. Create Urgency

The degree to which your prospect feels a need to act now, based on her internal drivers and the external influences you introduce, is your urgency factor. You should test the best way to create urgency. In a test of a site-wide call-to-action, for example, we found that adding an urgency message lifted conversions.  Can you test something similar?.

• 3. Improve Clarity

Communications with high clarity are understandable, cohesive, and accurate. They communicate the value proposition and call to action quickly. Even very effective pages probably aren’t perfect. Is the copywriting effectively communicating your value proposition? Do the images support the message? How obvious is the next step call-to-action?.

• 4. Improve Relevance

The relevance of your page is a function of how closely it matches your prospects’ expectations and perceived needs. Ask yourself, 'Does your page relate to what the prospects thought they were going to see?'.

• 5. Decrease Anxiety

Anything in the conversion funnel or missing from the pages that creates uncertainty in the prospect’s mind. Ask yourself: 'What are potential misgivings the visitor could have about undertaking the conversion?' If your visitor has to stop to consider their safety they may never start back up. But, too much emphasis on security can also hurt sales. We tested an example where placing a security symbol too close to a shopping cart actually reduced e-commerce sales by 2%. You should test that!.

•  6 .Decrease Distractions

Components of the page that redirect attention from the primary value proposition message and call to action. If an element isn’t important to your business, tone it down. How many things are you asking your customer to do on a single page? What is the most important element on the page? Stay focused on the page’s purpose.

By evaluating your landing pages and website through each of these six lenses will help you create true problems-solving hypotheses. This will get you out of the rut of testing random tips that miss your unique context.

Campaign Study with Iron Mountain

Here’s an example of some of the LIFT points we identified for one of Iron Mountain’s landing pages.

We can then find the best ways to eliminate problems by turning them into hypotheses for improvement. In this particular test for Iron Mountain, our winning page produced a 404% lift in lead generation conversion rate. (Yes, you read that number right.)

But, testing for conversion rate lift is just the beginning.

Phase Two: aim for marketing insights

There’s more to gain from your conversion tests than just conversion rate lift. If you’re only testing the latest 'tips and tricks' or so-called 'best practices' you’ve picked up from blog posts, you’re missing the potential.

Sure, the incremental gains are great. Who wouldn’t want a 10 or 20% (or greater) lift in their conversion rate, right?

But, what if that’s only a sliver of your potential?

 The most successful online businesses use a continuous improvement cycle where tests are planned to give marketing insights that lead to even bigger improvement. This ongoing process takes pre-planning, discipline and perseverance to make the biggest advances.

At WiderFunnel, we use a 7-step structured process to prioritize test strategy and feed learning into future tests. It leverages great thinkers who have gone before us and adapts the scientific method for website practicalities.

This process is intentionally designed with Graphic Design and Copywriting in the fourth step. This ensures that website test iterations are driven by solving customer problems identified in the LIFT Analysis rather than just creative aesthetics.

Using a cyclical process like this ensures that we build on learning and keep developing better hypotheses for the next rounds and for other websites in similar situations. By feeding the insights from the results of previous tests into the following rounds, you can make better gains.

You may have heard other conversion optimization advocates say that, if you’re not testing yet, you should 'Just start testing!' Well, I believe that’s bad advice. If you start with a poor process, you can waste a lot of time and organizational support. You’ll get better results if you just start testing the right way.

What do you think?

Do you run a continuous improvement testing program? What have you found to give the best results? What are the biggest barriers you face?

Thanks to Chris Goward for sharing his advice and opinions in this post. Chris is Founder & CEO of WiderFunnel, the conversion optimization agency where he’s developed conversion optimization programs for clients like Google, Electronic Arts, SAP, Shutterfly and more. His new book, You Should Test That!, redefines CRO and shows how to create dramatic business improvements. You can follow him on Twitter and connect on LinkedIn and Google+.