Questions? (877) 643-7244
FutureNow Post
Wednesday, May. 13, 2009 at 2:21 pm

Conversion Makeover: LandsEnd.com – Part 2

By Bryan Eisenberg
May 13th, 2009

In our last post, we looked at LandsEnd product page and showed you how we would change the flow of the page to maximize conversions. Today we’ll look at a scenario for LandsEnd, that starts at a pay per click ad for “Polo Shirts” and ends up in the cart.

Our search on Google started with the term “polo shirts” but provided us with an ad for “Polo Shirts for Man.” I guess women must never search for “polo shirts” even though LandsEnd sells women’s polo shirts on the site. The ad copy talks about the variety of fabrics and colors and LandsEnd does a good job of providing scent from the ad to the landing page.

My initial observation for this page was that people get to this page and would tend to narrow down their selection by size first. If you look closely at the main “hero” shot on the page you’ll see this blue horizontal line and inside it options to narrow down by size.

My suspicion was that most visitors were not seeing it. So the first thing I did was run a heat map on the landing page.  The results were the first sign that I may be correct about this.  Now, if I had access to the web analytics, and if those links under the “hero” shot were tagged properly, we would know for sure whether those size links were being clicked or passed over unseen as visitors eyes jump down the page toward the product photos.

The other thing the heatmap showed is the importance of testing which shirt or product on the category page placed in each of the top positions on the page.  As you can see, high-intensity colors draw the eye more effectively than desaturated colors and neutral tones.

To confirm our suspicions and heat map findings, we decided to use another important tool in our toolbox, usability testing. UserTesting.com was kind enough to provide 3 free tests for us to look at LandsEnd.com at the conference.

The first thing I needed to do was provide the task for our testing participants to complete. This is a critical piece to setup correctly in order to get maximum insight from the test. Here is the scenario as I described it to them:

Purchase 2 different polo shirts and matching pants/shorts for a trip you are taking to a hot tropical location next week. You’d prefer clothing that are low maintenance and are going to feel most comfortable in the heat.

Go through until the end of placing an order but do not submit the order. If this wasn’t a test, would you have made a purchase? Why or why not?

The link provided took them directly to the Men’s Polo Shirt landing page that the pay per click ad above did. Each usability tester recorded their thoughts out load and we were able to see their screen exactly as they did while navigating the site.

Here you can see one of the testers as she goes through making her purchase for a trip to Jamaica. If you want to see the other tests and the test participants notes you can see them on this video summary page. Analyzing her test confirmed what I had seen in the other tests as well and validated my concerns on the page.

Please enable Javascript and Flash to view this Viddler video.

These tests confirmed that testers were missing the sort by size links in the blue bar under the hero shot and additionally some of the same issues that we identified in part 1 of the conversion makeover.

1. Conversion Challenge: People just don’t understand some of the jargon/terms that LandsEnd uses. In terms of size what does “Big” mean? Is XL big?

I also asked the users to look for shirts that had fabrics that were good in tropical heat, and every participant went straight to cotton (because they knew that) and missed the other special fabrics (like mesh and moisture wicking fabrics) LandsEnd offers.  See a sample screen shot of the selection process below:

Recommendation: LandsEnd looks at different ways to label these terms or to explain them to make them more intuitive for their non-catalog buyers (in the catalog they explain many of these terms).

They could do this in the hero shot area using visuals as well for the sizing options.

Priority/Likely Impact: High

Resources Required to Execute:

  • Fixing the wording or labels in the pull down menu will probably take 2-3 hours, with most of the time spent on coming up with and agreeing on terms.
  • Providing a more intuitive understanding of the sizing categories within the hero shot, would likely take an additional 2-3 hours.
  • Total time = 4-6 hours, with a possible hour or two to actually implement the changes depending on their technology and platform setup

2. Conversion Challenge: The upsells on the product pages were virtually invisible to most of the participants until they went looking for them, but even then, the recommendations still weren’t entirely relevant for what the visitors were looking for.

Recommendation: As part of the product page redesign we did in part one of our makeover the recommendations were moved to a more visible location where there would also be additional room for 3 to 4 recommendations per product.  This is an improvement, bu LandsEnd still needs to ensure that the recommendations are as relevant as possible.  I would look to have a combination of recommendations for each product; one that is a similar product in the same category, one or two complimentary products from different categories (when looking at a shirt show pants or shorts), and possibly a complimentary accessory. Occasionally you can find a product were they have this kind of combination.

Priority/Likely Impact: Medium

Resources Required to Execute: Variable depending on LandsEnd product recommendation technology, which we are not privy to.

3. Conversion Challenge: People’s momentum was broken when they went to narrow down their selection from a category page and then further complicated when they went to purchase more than 1 of the same or similar items.

Recommendation: There are two things going on here. First the current task of narrowing down one’s selection criteria from a category page could be less than intuitive.  The location of the narrow search items, the too-subtle reaction to clicking on a faceted navigation selection, and the jargon used in some of the pulldowns all conspired to confuse shoppers looking to find just the right kind of polo shirt/item.

Secondly, the ease of adding multiple items of the same product but in different colors should take fewer clicks. Right now you have to click to close the cart, click to reselect your size, and then click to select your additional color before adding the item to cart.

I would recommend taking a look at how the Gap (Banana Republic, Old Navy, etc.) websites allow visitors to more easily narrow item selections through their quick look on their category page, and also how their “virtual cart” makes adding multiple items easier once the visitor has arrived on a product page.

I would also place some or all of the narrow selection/faceted navigation pulldowns above the featured products instead of having them outside the active window, in the left hand column.  Most important is the narrow by size option.

Priority/Likely Impact: High

Resources Required to Execute: This is a highly complex change which would likely require 30 or more hours, but which would have a huge impact and ROI.

Conclusion: I love LandsEnd and their fabulous products, and despite the conversion challenges we’ve discussed, the participants in our usability tests still felt confident they would buy from LandsEnd’s website.  However, very few people only buy clothing from one manufacturer or brand, and it only takes a slight erosion or friction in the process that would cause someone to abandon and head over to their next favorite (and competing) brand.

Usability testing is just one of the arrows in our quiver that we use for our OnTarget clients but occasionally we do this type of analysis for free for teaching purposes. Would you like to suggest your website for this type of review and improvement? If so, then be prepared to have it published. Just let us know in the comments below.

P.S. If you enjoyed this post please consider Tweeting it please.

Add Your Comments

Comments (15)

  1. I am the online marketing manager for one of Norway’s largest websites, FINN.no. Currently we work with PPC and other online advertising initiatives to strenghten our sub-brand FINN reise (our travel section) – http://www.finn.no/reise. We have certain landing pages where I’d welcome advise on improving landing page conversion rates. E.g. to increase hotel/flight/charter comparison searches on http://www.finn.no/finn/travel/destination/search/?location=oslo&targetForward=travel&logFreetext=true and flight comparison searches on http://www.finn.no/finn/travel/air/search
    I hope this could be an interesting case to have a look at for you and I’d be prepared to have it published for teaching purposes.

  2. Thank you Brian and your team for this thorough analysis. A lot of people can learn from you and improve their business. Keep on the very good work !

  3. Another awesome post! Thanks so much.

  4. “polo shirts” search phrase is genderless, you can’t know whether prospect is men or women. what style of landing page do you recommend in that case? a gender switch page or a mixed product list?

  5. Got here b/c of a link from a German blog that focuses on inproving conversion rates. Awesome article, I’ll definately keep it in mind. Love the AdWords ad though, they’re letting you know what to expect and don’t focus on price (so sick of it by now..)

  6. Good article, Bryan. I feel like you’re right here with me talking through the Lands End site.

    I wonder if they’ve got carried away with technology like the ‘Virtual Model’ stuff and forgotten some of the basics. Like the tiny ‘add to bag’ button below the fold.

  7. John,

    I don’t think so. They were already looking at some variations of this page to test soon.

  8. Point: If you buy PPC ads, keep in mind that a percentage of clicks always come from researchers and competitors, who have absolutely no interest in buying your stuff.

    What percentage might this be? maybe 10%. anyone know?

  9. Great article and interesting tool for the heat maps!

  10. Loved your article! can we share it

  11. great post. I wish we could find more posts like this often. Great job

  12. good article
    i think i really learn sth from it
    It is gonna work for those who deal with their business

  13. [...] Conversion Makeover: LandsEnd.com – Part 2 One of my favorite things about this website optimization blog post by Brian Eisenberg is that it [...]

  14. I wonder if they’ve got carried away with technology like the ‘Virtual Model’ stuff and forgotten some of the basics. Like the tiny ‘add to bag’ button below the fold.

  15. Point: If you buy PPC ads, keep in mind that a percentage of clicks always come from researchers and competitors, who have absolutely no interest in buying your stuff.

Add Your Comments

 

Print this Article
Share

Bryan Eisenberg, founder of FutureNow, is a professional marketing speaker and the co-author of New York Times and Wall Street Journal bestselling books Call to Action and Waiting For Your Cat to Bark and Always Be Testing. You can friend him on Facebook or Twitter.

More articles from Bryan Eisenberg

Marketing Optimization Blog
FREE Newsletter Sign-Up
send it once every: