MarketingExperiments Blog: Research-driven optimization, testing, and marketing ideas

It’s August, the sounds of sleigh bells are in the air, and the menorah lights are twinkling. Well, in the minds of consumer marketers at least. This is the time to nail down your holiday marketing plans. So to get your marketing juices flowing, I’m sharing an e-commerce optimization test we ran last holiday season.

Which order process leads to more conversions? A simple, fundamental question every e-commerce site (and, frankly, every lead generation site) needs to answer.

And the primary research question of the test we’re sharing in today’s blog post…
Read more…

E-commerce is a lot like magic. According to my in-depth research into the art of illusion (hey, I watched the movie “The Prestige”), there are three basic elements to the classic magic trick:

• The Pledge (set-up)
• The Turn (the trick itself)
• The Prestige (the reveal)

Much like a really good magic trick, you must also have your audience join you on a convincing journey, and deliver real value to them in the end without them feeling, well, tricked. For David Copperfield, that value is entertainment without feeling lied to by the illusion itself. For the classic e-commerce purchase funnel, that value is a quality product without being shocked by an unfair price at the end of the journey.

To that end, you might say e-commerce offer presentation has three stages (which may, in fact, overlap):

• The Interrupt (gains attention)
• The Value (establishes appeal)
• The Price (the reveal)

But when do you reveal price in the cart process? That was a question we received in our recent Web clinic, Shopping Carts Optimized: How a few tweaks led to 12% more revenue across an entire ecommerce website. I discussed this topic with our Director of Sciences, Bob Kemper, and here’s what he had to say… Read more…

In our most recent Web clinic, Shopping Carts Optimized: How a few tweaks led to 12% more revenue across an entire ecommerce website, Dr. Flint McGlaughlin shared our recent discoveries from our consumer marketing experimentation, set out a strategic approach to shopping carts and gave a few helpful fishing tips to boot.

As usual, we received more questions than we could possibly answer live during the Web clinic. A few were simple and straightforward (to which I say, “Yes,” “Maybe,” “One form field for name instead of three,” and “By the pier in Jacksonville Beach using Mayport shrimp as bait.”)

But one question particularly caught Dr. McGlaughlin’s eye…

How long should a shopping cart be? Is it better to have a long page or many short steps?

I passed this question around the lab, and here’s what our researchers had to say. We’d love to hear what you’re learned from your tests as well…

It depends on your product

I think this really depends on the product.  If you have a very simple product, like a DVD, you know what you’re getting as soon as you click “Add to Cart,” so I would get them through the process as quickly as possible.

If you have a more customized process, like ordering flowers with different vases and greeting card variations, I have no problem breaking them each out to their own special page so we can hammer home the value of each step without over-cluttering the long form. This also allows us to better track which specific step someone is dropping off on so that we can more easily determine the leaks in the funnel.

In the end, you have to test checkout process length.

– Tony Doty, Research Manager

The two optimization factors that you must balance

To the customer, shopping cart page length may be irrelevant unless the length is driven by unnecessary information.

I ran a test and discovered that reducing unnecessary fields on a single page inside the checkout funnel resulted in an increase in finishes, whereas including these fields in a similar process outside of the cart resulted in more conversions.

To the client/site, it is preferable to measure in multiple pages so they can track where the leaks are in the funnel.

If everything is one big page, it makes it much more difficult to track where or what causes a visitor concern enough to abandon.

In this case, showing the customer where they are in the process (progress indicator) helps keep the balance and alleviate the effects of that type of process friction (perceived process length).

– Jon Powell, Research Manager

Reinforce the value proposition

Optimizing the shopping cart path – including its length, sequencing of steps/forms, etc. – should conform to our foundation landing page optimization/conversion index analysis tenets. For instance:

• Not asking for any more information than you need
• Not asking for information you do need before you need it (to complete the process step)
• Managing form length and eye-path
• Avoiding ‘visual barriers’ such as horizontal bars across the page, etc.

The emphasis shifts slightly upon transitioning from ‘offer’ phase to ‘cart’ phase, shifting from ‘expression’ of the value proposition towards ‘support’ or ‘reinforcement’ of the value proposition to sustain (rather than build) cognitive momentum toward conversion.

– Bob Kemper, Director of Sciences

Optimize the page or the path

There are two approaches you can test to see which works best with your customers and products. Either have a clear descriptions of the steps (breadcrumbs) to let customers know how many to expect and where they are in the whole process, or create one longer page that includes all necessary billing and shipping fields.

– Zuzia Soldenhoff-Thorpe, Research Analyst

Test on new and repeat customers

How long should a shopping cart be? It is better to have a long page or many short steps? It depends.

Every retailer should test as many different checkout processes as they can. Retailers need to know what their customer target group responds better to. Some visitors will prefer one single, long step and others will prefer a couple of short steps.

Probably for repeat customers, short checkout process (1-2 steps) will work well because they already trust the retailer and are familiar with the process. But even in this case, it is important to test.

– Gaby Paez, Associate Director of Research

Be brief and be thorough

I personally prefer a short cart, incorporating all of the steps in one with accordion-style sections. With this type of a checkout process, it’s easy to get back to previous steps with an ‘edit’ link and it appears short while still collecting all of the needed information.

My favorite checkout process is Gap/Banana Republic/Piperlime – it’s super intuitive and really easy to get back to any step to make a change

My biggest pet peeve is when a checkout does not work in a certain browser. I use Google Chrome, and the other day I was looking at something on the Hobo International site and I couldn’t select from a drop down in Chrome, but when I went to Firefox it worked. Most customers wouldn’t be that dedicated or might not think to check another browser.

– Gina Townsend, Director of Operations

Related resources

Free Web Clinic, May 18, 4-5 p.m. – Optimization Researched: Latest findings about effective LPO practices based on data from 2,673 marketers

Web Clinic Replay – Shopping Carts Optimized: How a few tweaks led to 12% more revenue across an entire ecommerce Website

E-commerce: How your peers optimize shopping carts and product pages

E-commerce Shopping Carts: How a redesigned checkout process led to 13% increase in conversion rate

Shopping Cart Abandonment: How not being annoying can get you 67% more cart completions

How often have you placed an item in your shopping cart in a bricks and mortar store, a can of green beans for example, and then changed your mind before you forked over the cash and put those green beans back on the shelf without buying?

How about online? Ever drop something in the ol’ virtual shopping cart, but change your mind and decide not to buy?

If you’re like me, it’s very rare for you to change your mind in a physical store, but I do it online all the time. Why? And how can you reduce shopping cart abandonment and improve product pages on your own site? There is no one right answer to these very difficult questions, but we’re going to be sharing our latest research discoveries today at 4 p.m. EDT in our latest free Web clinic – Shopping Carts Optimized: How a few tweaks led to 12% more revenue across an entire e-commerce website.

But first, we asked your peers these vexing questions. Below are a few of our favorite responses… Read more…

“What else can I test?” This is one of the most comment questions we get asked at MarketingExperiments. Perhaps only after… “What should I test first?” and “How do I test?”.

Before I help you determine what to test next, you must honestly determine if you really have exhausted your own test ideas. Testing is both a meticulous and iterative process, so there are no short cuts. There are no silver bullets. If “Page Z” is the most effective webpage possible (most effective at the moment, since the universe is most certainly not static), then you need to test A, B, C, D…and so on in order to get to Z. By going directly from A to Z you will miss small subtleties and large revelations that would have influenced your design of page Z.

In order to exhaust your test ideas, there are two important stages to consider in a testing process:

• Radical redesigns
• Fine tuning

A shrewd procedure to adopt as you begin testing is to “Go radical, then fine tune.” This can apply to landing pages, home pages, shopping cart processes, etc. It’s an effective way to gain decisive knowledge in a short period of time. Read more…

This weekend I was paying for the 10 gallons I had just put into my old 1997 Honda Civic, when I decided that I’d purchase a nice cold soda for the road. I pointed out the pump where my fueled-up car was located and then slid the cold beverage to the convenience store clerk. He informed me that my total came to $25.89 and then he stopped.

Looking me dead in the eyes, he asked me what my name was. “Austin,” I replied a little hesitantly. “Austin, are you sure you want to spend $25.89 for 10 gallons of gas and a cold soda?” he asked. I nodded and attempted to hand him my Visa credit card.

He denied my overture and informed me that he could only help me if I were a member of his store. So not wanting to cause a scene with the five people who were now behind me, I conceded. Read more…