32 Split Testing Ideas from 4 of the World S Leading Conversion Experts

32 Split Testing Ideas From 4 of the World’s Leading Conversion Experts

November 19th, 2014 by Lindsay Marder

Ok, I’m super pumped.

We tapped the minds offour of the world’s top conversion rate optimization (CRO) experts to tell us what they would test on this landing page for Rosetta Stone.

Here’s our panel of experts…

• Chris Goward – Founder of Wider Funnel, author of You Should Test That! and Traffic and Conversion Summit 2015 speaker.
• Tim Ash – Author of Landing Page Optimization, CEO of Site Tuners and Founder of Conversion Conference.
• Justin Rondeau -Data Driven Marketing & CRO Trainer. Analyzed 2,500+ Tests Across Virtually Every Industry.
• Brian Massey – Author of the The Customer Creation Equation and Founder of Conversion Sciences.

You’ll get tons of split testing ideas that you can apply to your own landing pages.

Here’s what we asked our conversion experts to critique…

The Search

The Ad

The Landing Page: Rosetta Stone

The Questions

• What did you like on this landing page?
• What would you test to boost conversions?

Here’s what four conversion experts had to say about what was done well and what could be improved on this landing page.(NOTE: We’d love to hear how you would improve this landing page. Give us your thoughts in the comments section.)

Chris Goward,WiderFunnel Marketing Optimization

What I liked:

Value Proposition: Known Rosetta brand gives them some leeway to overcome some conversion barriers. But not enough.

Value Proposition: Good font and graphic standards add to credibility.

Value Proposition: Box shots and app screenshots add tangibility for an intangible service.

What I would test:

Unfortunately, the rest of the design, layout and copywriting leaves a lot of money on the table. From a conversion perspective, there’s a lot of opportunity for testing.

The way I always approach a landing page test is to analyze it from the perspective of the shopper and look for barriers to conversion. Categorizing them using theLIFT Modelgives structure and helps with idea-generation.

Here are some of the LIFT points I identified for Rosetta:

At WiderFunnel, we don’t just toss around test ideas willy-nilly. The structured process ensures we narrow down the best test variations that achieve two priorities: 1. Conversion rate and profit lift and 2. Insights from test isolations.

The next step after the LIFT Analysis is to convert these LIFT point problems into test hypotheses. For example:

There are a lot of potential things to test here. So much so that I’d recommend doing some dramatic A/B/n redesign and re-copywriting testing as a starting point before drilling into many individual hypotheses.

We’d get one of our conversion-trained designers to work with the conversion strategist to design smooth eyeflows and design treatments that will enhance the clarity of Rosetta’s value proposition.

Any of the LIFT points above could be isolated as a hypothesis, depending on the long-term conversion strategy. In other words, which isolations will lead to the biggest insights for further testing? That’s an important discussion to have when planning a test.

No optimization test is done with one test round. It should be viewed as an ongoing test plan.

Hear more from Chris at Traffic and Conversion Summit 2015.

♦♦♦♦

Justin Rondeau,Data Driven Marketing & CRO Trainer

What I liked:

This page has a great look and stacks the content perfectly. The main CTA is above the fold, followed by descriptive content, and finally persuasive content via social proof.

What I would test:

Though this is a great looking page, there are some things I’d suggest to test.

1) Match your ad with your landing page! If you know the visitor is looking to speak Spanish, have that language pre selected! Your ad did the hard work for you; if a visitor were looking to speak a language other than Spanish they wouldn’t be clicking your ad! Do yourself a favor and match each bit of content with the desired language.

2) The first part of the headline is highly branded, though I think the most compelling piece of copy is the second line, ‘Language Learning that really works’. Rosetta Stone is a well-known brand; they don’t need to add branded copy here. I’d suggest eliminating the first line and only use the sexier hook. This gives a succinct description about what sets your product apart and also helps reduce vertical space in the process!

♦♦♦♦

Brian Massey,Conversion Sciences LLC

WhatI liked:

Marketing 101 is “Show the product.” I think Rosetta does a good job of showing what you get. However, could they show the “demo” better? That is really what they are selling on this page.

The offer is made early. It is displayed in a high-intensity gold box that makes it hard to miss.

Good use of proof; 8000 Corporations, 9000 Agencies, 20000 Schools.

The testimonial here is one of the best I’ve ever seen. It’s a story. It has love. It has a great start and a fantastic ending.

What I would test:

The landing page does a poor job of keeping the promises made in the ad.

Promise 1: Learn Spanish

Promise 2: Free shipping

Promise 3: 30-day Guarantee

Promise 4: Chat with tutors

I would test:

1. Mentioning Spanish in the headline.
2. Having the “Select a Language” pulldown default to Spanish
3. Test the copy written at a post-graduate college level against copy explaining the free shipping, 30-day guarantee and chats with tutors.
4. Test current copy against copy that explains what a demo is. Is this a lesson? What will I discover? How long will the demo take?
5. Get rid of the designer that made (1), (2), (3) and (4) readable, but made “Sights and Sounds”, “Word”, “Sentences” and “Conversations” too small to read. Test making the numbers smaller and the words larger. We’ve seen good luck with steps on landing pages.
6. Get rid of the designer that separated the numbers from the metrics. “20,000”… why is this important? Test bringing the words “Corporations” etc. close to their respective numbers. This could be important proof for the page.

♦♦♦♦

Tim Ash, SiteTuners

What I liked:

1) The production quality (professionalism) of the page is pretty good.

2) The page uses social proof in the form of large numbers of corporate and institutional users

3) The page uses a third-party endorsement from the Wall St. Journal

4) The page uses a testimonial from an actual user

WhatI would test:

The page is a hot mess and does not deliver on the promise of the original PPC ad that led to it. There is a huge mismatch between the calls-to-action on the page and the one in the ad. The title of the ad is “Learn Spanish – Start Speaking Spanish in Minutes” and the URL is Despite this there is not a single mention of “Spanish” visible anywhere on the page. Instead we see buttons for “Ready for a FREE Demo?”, and “Get Product Details And Pricing”. The social proof and testimonial are also not Spanish-specific. In fact, the lone testimonial on the page is about someone learning Mandarin Chinese.

I would focus the redesign on matching visitor intent.

Here are some page elements that I would combine is an alternative redesign:

1) Page Title “Learn Spanish – Start Speaking Spanish in Minutes”

2) Enlarge the packaging (box shot) portion that clearly indicates that it is the Spanish version

3) Put in some common Spanish phrases as background in the top image (e.g. “QuePasa”, and “Como Esta”, Etc.)

4) Have a single call to action, “Try Our Free Spanish Demo” (I doubt anyone would buy it without first trying it, so the alternative call to action of getting pricing is very premature) – without the need for a language dropdown in the form (since they have already indicated their language of choice by clicking on the ad).

5) Have 15-20 testimonials from people that bought the Spanish product at the bottom of the page, along with a specific number of people who have taken the Spanish course (e.g. “Over 1,200,000 satisfied Spanish learners”)