Online Testing: 6 business lessons I’ve adapted to website optimization

I never intended to become an online marketer. In fact, I never planned to go into anything business related.

My big life plan was to quietly finish my history degree and curate at the most famous museum that would take me. Sometimes, however, things turn out better than we planned and a bad job market pushed me into an MBA.

This is how I came to find myself with the coolest job in the world: optimizing everything I can get my hands on at MECLABS.

Unfortunately, without a marketing or e-commerce degree, I sometimes I feel like a fish out of water around my colleagues. As a result, I have learned to adapt what minimal business savvy I do have into concepts I can apply to optimization and testing.

 

Look for bright spots

A bright spot in the business world is a successful element in an otherwise unsuccessful situation. (If you haven’t read the book Switch by Chip and Dan Heath, I definitely recommend it as they nail the concept of “finding the bright spots” as a catalyst for change.)

And for CEOs and executives, finding the bright spots can mean identifying successful employees or programs and then finding a way to spread those practices or patterns company-wide.

We often use a similar method to the one Chip and Dan advocate when working to optimize a page or site.

Bright spot identification in optimization can come in the form of competitor or internal analysis. At the beginning of a MECLABS Research Partnership, we perform an analysis of the Research Partner’s competitors to identify things the competitors are doing well that we could repurpose for our Partners.

Consider for example, if every competitor in the marketplace is promoting its product with large dynamic images and the Partner is using smaller ones, we might identify testing larger images as an industry bright spot opportunity. Many of these bright spots can create big learnings and substantial lifts.

For internal analysis, we often review the entirety of a Partner’s site looking for pages, traffic segments or functionality that is outperforming other options. To do this type of analysis, we usually employ the data team to pull and analyze engagement and conversion data from throughout the site and perform a Conversion Index Analysis.

Sometimes, the internal bright spots can come from the most unexpected places. For example, we had a Partner once who was seeing a much higher clickthrough rate from one of its landing pages as compared to all of the others.

After running data analysis and reviewing the page, we determined the well-performing page had a video that was not present on the others. We tested the video on the other pages and saw a lift across the board.

 

Be proactive, not reactive

In the business world, it pays to be nimble.

Diversification and flexibility will always be buzzwords for the CEO crowd because being able to stay ahead of rapidly changing industry standards is a coveted skill. In years past, Apple has been hailed as a company with such skills.

When asked about its ability to define the industry, Steve Jobs famously said, “People don’t know what they want until you show it to them.” 

There are many symptoms of a site that needs testing. For example, falling conversion rates, high bounce rates and visitor complaints.

Yet, many marketers will wait until one of these symptoms appears to begin looking at ways to improve their site – but this is a reactive response to problems that could have been prevented.

Performing analysis proactively on how you can make your page better is a step toward a better user experience overall. Steve Jobs did not wait for the marketplace to ask for an iPhone. Instead, he looked at the industry, identified an unfilled need, and made a product that became the new standard for smartphone design.

As online testers, we should be inventing iPhones all over our pages. Don’t wait for your customers to tell you something is wrong. Actively search out ways you can make your site a better user experience and then test until you have something capable of changing the industry.

 

First give value, then get value 

“First give value, then get value,” was a favorite saying of an old business professor of mine.

He used the phrase to mean you have to provide value in a workplace before you should expect to get any back. He also used it as a rather diabolical system of extra credit, but that is beside the point. When it comes to applying business concepts to optimization, this is one of the most cut and dry examples.

We must provide our site visitors and customers some sort of value before we can expect to get any back.

The value we provide to a visitor often comes in the form of a well-crafted value expression. It should be one highlighting all of the major elements of value: appeal, exclusivity, clarity and credibility.

We can also provide value via incentives in the forms of a free download or extra product offerings. Whatever you use as “value,” you must remember it has to have value in the eyes of a customer. Irrelevant or useless downloads may not provide enough value to get their information.

For example, a short e-book or buyer’s guide in exchange for a long or multi-step lead gen form including a phone number might not be an equal value exchange. At the same time, asking for a name and email address in exchange for the same incentive might be just right.

 

 

Too many jams

“Too many jams” isn’t so much a business concept as it is a marketing experiment turned catchphrase.

The term is based on a study done by Sheena Iyengar about the number of choices a consumer can realistically think through before making a purchase. At the end of the study, it was determined if a consumer is offered too many choices, they will often shut down and not make a purchase decision at all. 

Too much of anything on a website is never a good idea. We do not want to present too many options to the visitor at once because they might have a hard time choosing between them.

Closely related, we also do not want to provide too many exit points. This goes double for any kind of landing page optimization. If we have a specific goal in mind for our page, let’s try to make that clear by eliminating unnecessary links and navigation.

The goal of the page (or macro-yes) should be painfully obvious to the visitor. Giving the user too many product options, extraneous links or CTAs can lead to confusion, distortion of the eye-path, distraction or even failure to convert.

 

A loss can be an education 

One of my favorite business anecdotes is about Thomas Watson, the founder of IBM.

The story goes that an IBM employee made a multimillion dollar mistake. The next day, when called into Watson’s office, he attempted to offer his resignation.

Watson’s reply to the resignation, “Are you kidding, I just spent $10 million educating you!”

The lesson Watson sent into the world with this meeting is not only useful in business situations, but also in testing. Losses create excellent opportunities for education and learning.

At MECLABS, we always aim to increase conversion, but when we don’t, we never dismiss the test as useless.

Sometimes, we spend even more time pouring over the results of a losing test than those of a winner. We do this because sometimes a loss can provide deeper insight into a visitor’s motivation and behavior than a win.

In short, leverage your losses as a learning opportunity.

 

Use humility to set the right expectations

Humility is a character trait more than a business lesson, but I propose we should view it as critical to success in any situation.

This is also true when testing.

Just as a loss should not be discarded, a win should not be considered the end of testing.

Sure, you need to pat yourself on the back. Yet, no matter how large the victory we should attempt to remain humble and remember there is always room to improve when testing.

Value proposition development testing is a great example of this concept. In our value prop training, we teach that value testing is ongoing because there is always a different expression that might better resonate with the customer base.

Consequently, setting expectations of humility with our Research Partners is an important part of all testing plans. Partners are often so excited by a win they don’t want to risk a loss with a new test or are not willing to devote more resources to the same section of their site.

Reminding them a victory can always be improved upon usually helps us move past this stalemate. In testing, having the humility to say, “Yes, I won, but I can do better,” is a valuable trait.

 

Related Resources:

A/B Testing: One word will unclog your conversion testing

Marketing Careers: Why marketers and media professionals must never lose their wild spark

Online Marketing: 3 website optimization insights I learned from baking