Ecommerce: Applying Conversion Science Q&A

Here are several questions about applying conversion science to ecommerce sites. These questions came from the sponsors of the GP Ecommerce Summit in Bucharest, Romania.

  1. Can we consider Conversion Rate Optimization a real science?

What defines a science? The Scientific Method.

  1. Assume we know nothing about a problem
  2. Research it
  3. Develop hypotheses
  4. Select the most likely hypothesis for testing
  5. Design a test that isolates that hypothesis
  6. Run the test using sound statistical methods
  7. Evaluate with post-test analysis
  8. Draw a conclusion
  9. Use the new information to formulate new hypotheses
  10. Repeat

I’ve just described our six month Conversion Catalyst process to you. We “science the sh*t” out of websites. Without the science, we make bad decisions, emotional decisions, decisions based on superstition and myth.
There is also a component of sport in conversion optimization. We are in this to win. While we must be objective, we like to find revenue and hate when our tests are inconclusive.

  1. What are the first steps you have to take if you wish to increase your conversion rate on your e-commerce website?

My recommendation is that ecommerce sites focus on the value proposition their offering. This is a combination of your categories (what you sell), your shipping policy, your return policy and your brand.
Zappos built an amazing online brand by putting its value proposition front and center, “Free shipping both ways. 365 day return policy. Empowered customer support people.”
What is your value proposition? Fast delivery? Local manufacturing? Free installation? Donations to charity with every purchase? Emphasize it on your site, in your cart and throughout checkout.

  1. How do you create a good landing page and what are the best ways to test it?

The best landing pages keep the promise of the ad, link or post that brought the visitor there. They make an offer that matches the promise as exactly as possible. They show the product, even if it is a service or a PDF or a video series. Good landing pages provide proof points that are specific and supported by fact. Good landing pages build trust by borrowing from customers and customers. Good landing pages make the call to action the most prominent thing on the page. And good landing pages don’t add any distractions, such as social media icons, links to other pages or corporate site navigation.
This is the chemical equation for landing pages: Offer + Form + Image + Proof + Trust = Landing Page

The chemistry of the landing page

The chemistry of the landing page

  1. Can persuasive writing help you sell more online or do you need more than that? For example, how do you test a good headline?

Most of our biggest wins come from copy changes, like headlines. We are even testing different kinds of testimonials on one site to see which build the most trust. The words are very important. This is related to the value proposition I discuss above. When you learn the emotional language that brings visitors into your site, you learn something about your audience. This insight can be used anywhere.

  1. What is an important point you want to drive home?

There is a wave of ecommerce sites rushing to rebuild their sites using responsive web design (RWD). This is in part due to Google and Mobilegeddon, but few can ignore the growing influence of mobile devices on our revenue. This rush to RWD is a mistake for many businesses who will find themselves with a poorly performing mobile site and a lower conversion rate on their redesigned desktop site. Tragic.
You should embrace your mobile visitors, and there are alternatives to RWD. I’ve seen some redesign horror stories and some pretty amazing success stories. Mobile design is still too new for there to be best practices, but our testing tells us what successful mobile designs should begin to look like.

  1. How do you remember the ecommerce market in the USA from 10 years ago?

Ten years ago, we didn’t have the data tools we have today. We relied much more on qualitative research. Most of my work was building out personas, making content recommendations and working with “best practices”. Google Analytics was young. We had been using server logs to get unreliable data on visitors. Only a few years before I had written my own web analytics package to get an idea of what was working on my sites.
Today, we have amazing qualitative and quantitative tools to uncover problems with our websites. We enjoy powerful testing tools to help us determine exactly what effect our changes will have on our businesses. We are creating revenue in the laboratory using science and creativity. We have moved from the tool-building phase into the human creativity phase. It’s a very exciting time to be an online business.

Brian Massey
2 replies

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

Your email address will not be published. Required fields are marked *