Building the Perfect Landing Page: Q&A with Conversion Scientist Brian Massey
We covered a wide range of pressing landing pages topics including B2C and B2B strategies, key testing criteria, popular hosting platforms, data-driven designing, and much more.
In case you missed it, you can catch the full recording of the webinar here.
Webinar attendees took full advantage of Brian’s expertise, so much so that we ran out of time trying to answer the many questions asked. To rectify this, we caught up with Brian following the webinar to cover all open questions – and have condensed these topics into a final recap for your benefit.
Here’s your final round of Q/A with Brian Massey:
Thanks for a great webinar, Brian! First up in our final questions: how do you balance designing a landing page for mobile, desktop, and tablet?
We design the two separately. The mobile landing page often evolves separately from the desktop/tablet versions. Mobile visitors are just in a different context most of the time.
Mobile responsive templates only get you so far, but it’s better than not having a mobile version of your landing pages. You should look at reducing mobile load times by stripping out images and other unnecessary content that slows loading.
If you are advertising using Facebook, Pinterest, or email, you’re going to have more mobile visitors than desktop or tablet visitors – start with mobile. If you are advertising to businesses, you may see more desktop users.
Is there a different strategy or approach for B2B vs B2C companies?
B2B visitors are consumers working in an office. While they behave in human ways, they are spending someone else’s money, so their motivations are different. They are worried about their reputation and their job, so they may be more emotional than consumers.
You collect the same data but selling to businesses is often a longer process involving multiple decision makers. Your primary focus is list building. Get a name and email address so you can qualify them by email. Then pass the good ones onto sales.
Volumes of leads are usually small, meaning AB testing is difficult, but using a tool like Decibel to understand them is the same. Use scroll maps, heatmaps, and session recordings – this will make you better.
What’s your opinion on ‘the fold’?
Not every page has a fold. And for those that do, it may not be at the bottom of the screen. I recommend using scroll map reports to see where visitors’ attention is waning. This is your true bottom.
For your quick decision-makers, the content they need should be above this drop-off. For example, if you have visitors returning to the page ready to buy, you better have a call to action above the fold that tells them where to start. For new visitors, such as those driven by ads, you may have to sell above the fold before making a call to action.
What do you suggest for new companies that don’t have much data yet?
Start collecting data. In particular, make sure your analytics database isn’t blind to any user interactions, such as:
- Content that isn’t on its own page doesn’t generate a URL for analytics. Examples: popups, tabbed content on a product page, text that rolls out when the user clicks, like a FAQ page
- Site search terms
- Ecommerce products that are viewed, added to the cart, abandoned, and purchased.
- Lead forms that are completed, but have no “Thank you” page
- Live chat interactions?
- Phone calls from the site
Over time, you will collect this data and should be able to query analytics for answers to questions like:
- Are our popups working?
- In what content on our product page are visitors most interested?
- What are people searching for on our site?
- How many products do visitors need to see before buying?
- What is our cart abandonment rate?
- How many leads are we really generating?
- Do return visitors interact with our live chat?
- Does anyone call from the website?
What is the minimum number of unique visitors to a page to conduct minimum testing?
We all know that the larger the sample size, the more reliable the data. For example, let’s say you were comparing two cameras online. If one camera has a five-star rating with 5 reviews, and another has a 4-star rating with 250 reviews, which data point is most believable? The 4-star camera is more likely to be a 4-star camera.
For AB split testing, we want to compare the number of conversions, not the traffic. Of course, the conversion rate will determine the amount of traffic we will need to get a set number of conversions.
We like to see a minimum of 100 conversions per variation, including the original, which we scientists call the “control.” Testing one variation against a control would require 200 conversions. If the difference between the control and a variation is small, the number of conversions will need to be higher. Likewise, if the variation delivers a large increase in conversion rate, you can expect to reach statistical significance with a smaller sample.
So, shoot for 100 conversions per variation.
What advice do you have to convince the marketing staff of the necessity of testing?
My advice is to get better yourself. Choose a place to begin from the Data Tool Map and start getting good at data for yourself.
As you incorporate more data into your deliverables and presentations, you begin to change culture. The good ones will appreciate some data to back up their work.
If you are not a manager, you may need to enroll a champion to help you drive some of this analysis.
If you can’t move the needle, you will be much more valuable to a company that does have a data-drive design culture.
How can you incorporate content for search rankings without compromising on conversion?
There is a constant tension between the team writing content for the search engines and those writing for the humans that search is bringing.
Here’s the rub: search rankings are heavily reliant on the time people spend on your site after the click. Keyword stuffing is no longer going to win, and conversion optimization is very good for SEO.
The mistake marketers make on content pages is not bringing the visitor to choose, which is the second job of the landing page. You should treat content pages as landing pages by offering relevant calls to action – ask the visitor to take another step.
Here are some things to try:
In your presentation you mentioned the example of Automatic Lite. With their pricing options, did you try leading with the more expensive option to use the higher price as the anchor?
Yes, we did! It didn’t make any difference in this case, but this is a valid hypothesis that any business with tiered products should consider.
There are several “landing page” platforms out there. Are there any you recommend in particular?
I would choose the landing page platform based on these criteria:
- Can I track the visitor all the way through the process using my analytics database? For example, Leadpages proved unable to track a visitor to my landing page, then to their form popup, and through to the thank you page. I need my analytics to be able to do this so I can track segments and repeat visitors.
- How easy is it for me to create and edit landing pages? If you run into problems creating landing pages, you will lose momentum.
- How quickly do the pages load? Performance is important, especially on mobile devices. Setup a page and run a test with Webpagetest.org. Does the page load within 2 seconds? It should.
- Is it mobile friendly? Can I make the mobile page very different from the desktop page?
- What other systems does the service integrate with? If you can’t easily get your leads into your email automation system, then what’s the point?
We chose Unbounce for our landing pages at Conversion Sciences. They meet all these requirements, and many of our pages convert at very high rates.