Archive

Posts Tagged ‘a/b testing’

Email Marketing: Test ideas for five types of email

July 1st, 2016

Testing your email marketing can help power some pretty impressive results – like a 100% increase in clickthrough or a 114% boost in revenue.

But … let’s be real … it is harder to test your email than to just send a single version of the email idea you come up with.

One challenge with email marketing – to keep your customers clicking and coming back for more, you need to endlessly come up with new messages and ideas for every email you send or set in an automation platform.

However, when you test your email marketing, you don’t get to create just one email for each campaign, you now need an A and a B (and a C and a D … etc. … etc. … depending on how many treatments you have and your list size can support).

To give you some new hypothesis ideas for your next email test, I interviewed Mike Nelson, Co-founder and Head of Marketing, ReallyGoodEmails.com, in the MarketingSherpa Media Center (MarketingSherpa is the sister publishing brand to MarketingExperiments).

He brought examples of five key types of marketing email from his site, which is described as a “modern-day museum” full of emails.

Read more…

Digital Marketing: 3 test ideas to optimize your incentive offers

June 13th, 2016

Coupon. Free download. Discount code. Gift card. Complimentary ebook.

These are just a few of the countless types of incentives marketers use to influence customers to say “yes” at the final macro-decision – whether that’s making a purchase, filling out a lead gen form or some other form of conversion.

Incentive can be just what some customers need to commit to an action you want them to take. But how can you ensure your incentive offers are having optimal effect on conversion?

Here are three tests you could use to optimize your incentive offers.

 

dollar-versus-percentage-incentiveIncentive Test #1. Percentage off vs. Dollar amount off

This first test idea is definitely a numbers game.

What do customers see more value in: a percentage discount or a defined dollar amount discount?  One case study from deep within the library of our sister site MarketingSherpa tested this question.

Evo, an online retailer of outdoor gear and fashion apparel, offered a coupon for certain product packages that had an average price of $333, and a minimum price of $250. With these numbers in mind, the team determined that $50 was the ideal discount point.

In addition to a traditional $50 off coupon, the team elected to test a 15% off coupon as well, which roughly equaled $50 off the average price of $333.

The test produced these results:

  • The $50 off coupon produced a 72% higher conversion rate
  • The $50 off coupon generated 170% more revenue than the 15% off coupon

It’s important to measure both conversion and overall revenue with this test. Even if one treatment results in more purchases, it may have lower average order amounts. If the average order amount is lower overall, the treatment with fewer orders of higher amounts could be your best choice moving forward.

  Read more…

Email Marketing: 7 (more) testing opportunities to generate big wins on your next email test [Part 2]

May 2nd, 2016

Does your email audience prefer short or long emails? How about images versus GIFs?

If you don’t know the answer to any of these questions, it’s OK. All you need is an A/B email test. 

Testing allows us to better understand our customers, and determine ways we can better engage them.

Last week, we detailed nine experiment ideas for you to try on your next campaign. If those weren’t your style, we have seven more for you — for a total of 16 testing opportunities.

Today, we’ll be reviewing opportunities in your body messaging, calls-to-action and design.

Email Body Messaging Testing

Testing Opportunity #10. Messaging tone

In this test, from the Web clinic, “Email Copywriting Clinic: Live, on-the-spot analysis of how to improve real-world email campaigns,” researchers used two treatments to increase total lead inquiries from visitors who abandoned the free trial sign-up process.

The first treatment was designed based on the hypothesis that visitors did not convert because the copy didn’t engage them enough, so it took a direct response tone. The second treatment was based on the hypothesis that visitors experience high levels of anxiety over potential high-pressure salespeople or spam phone calls. This treatment took a more “customer service”-oriented tone.

Read more…

Email Marketing: 9 testing opportunities to generate big wins on your next email test [Part 1]

April 28th, 2016

Email is a great medium for testing. It’s low cost, and typically requires less resources than website testing. It’s also near the beginning of your funnel, where you can impact a large portion of your customer base.

Sometimes it can be hard to think of new testing strategies, so we’ve pulled from 20 years of research and testing to provide you with a launching pad of ideas to help create your next test.

In this post and next Monday’s, we’re going to review 16 testing opportunities you can test around seven email campaign elements.

To start you out, let’s look at nine opportunities that don’t even require you to change the copy in your next email.

 

Subject Line Testing

Testing Opportunity #1. The sequence of your message

Recipients of your email might give your subject line just a few words to draw them in, so the order of your message plays an important role.

In the MarketingExperiments Web clinic “The Power of the Properly Sequenced Subject Line: Improve email performance by using the right words, in the right order,” the team reviewed several tests that demonstrate the importance of thought sequence in your subject lines.

Try testing point-first messaging. Start with what the recipient will get from your message and the email.

Read more…

A/B Testing: Cut through your KPIs by knowing your ultimate goal

February 4th, 2016

Marketers often struggle to know what metrics to use when trying to decide on the positioning of their marketing collateral. This can lead to many problems. At MECLABS Institute, the parent company of MarketingExperiments, we have run experiments and tests for over 20 years to help answer this question.

Customers take many actions when moving through the funnel, but what is the ultimate goal the company is trying to achieve with their marketing collateral? By answering this question, companies can best determine what the most important KPI is to measure.

To best illustrate this point, let’s walk through an experiment that was run regarding metrics. By reviewing this experiment we will understand how important it is to have a clearly defined idea of what the ultimate goal is for your marketing collateral.

 

The Experiment:

Background: A large newspaper company offering various subscription options.

Goal: To determine the optimal regular price point after the introductory discounted offer rate.

Research Question: Which price point will generate the greatest financial return?

Test Design: A/B split test

 

Subscription services often offer a discounted introductory rate for new subscribers. This gives potential subscribers a low-risk opportunity to try out the service for a period of time before the cost defaults to the regular full price. In this test, The Boston Globe team hoped to determine the optimal price point for a monthly subscription after the introductory offer rate expired. 

Read more…

This 1960s Statistician Can Teach You Everything You Need to Know About the Fundamentals of A/B Testing

January 21st, 2016

I did a training on selling training for the sales team today. It was what Millennials call “meta.”

I was talking about how our training uses scientifically valid experiments to back everything we say in our training rather than best practices, anecdotal case studies or just “expert advice.”

The question naturally arose: “What do we mean when we say ‘scientifically valid experiments’?”

When I answered the question in the meeting, I immediately thought it would be a good idea for a blog post. So, with that said, here’s the answer:

In short, it means that we use the scientific method to validate every piece of knowledge we transfer in the training (and also in our Web clinics and on this blog).

I found myself trying to explain what I learned in high school about the scientific method, and while I was able (I think) to get the basic gist across, I don’t think I did it justice.

Fortunately, after doing a little searching online, I found this guy.

His name is J. Stuart Hunter and he is one of the most influential statisticians of the last half of the twentieth century.

Fortunately, back in the 60s, he recorded some rad videos around experimental designs in a business context. If you can extrapolate a little bit from the industrial context and apply this to a marketing context, it should be everything you need to know about the scientific method, or “what we mean when we say ‘scientifically valid.’”

 

 

Read more…