Net Affinity Blog - Say hello to the future of hotel booking technology

How to Create Higher-Converting Special Offers with A/B Testing

Written by Net Affinity | Nov 9, 2015 12:00:00 AM

A/B testing (sometimes knowns as split testing) is comparing two versions of web page to see which performs better. To find the better version, you split traffic between two versions at the same time and have a success metric in mind. After you’re done, you take the successful version and keep it for real-world use.

While many hotel sites simply don’t get the traffic to run a statistically significant A/B test, larger ones will (here’s a quick calculator to find out if you qualify). If you’re a hotelier with a large volume of traffic to your website, you need to be A/B split testing. Why?

Just like the fashion industry, the hospitality industry is constantly evolving. New consumer trends and marketing practices always emerge. The question is:

Can those trends and practices help you drive more bookings?

A/B testing helps you find the answer to that question.

By using real-world data, it eliminates guesswork from the equation. You get actionable insights you can use to maximize hotel bookings.

In today’s post, you’ll learn you how to use A/B split testing to create higher-converting special offers.

Let’s get started.

(Image source)

Why A/B Split Testing?

(Image source)

Have you ever heard of Aikido? It’s a martial art that focuses on using minimal energy for maximum results. Instead of trying to force your opponent into submission, you simply step back, and use the momentum from his incoming attack to put him on his back.

So what does this have to with A/B testing?

Quite a lot, actually.

A/B split testing is like the Aikido of marketing, because it lets you use the momentum from your existing traffic to get way more. Think of it as the dedicated hotel marketer’s shortcut to more direct bookings.

You can spend thousands on acquiring more traffic to increase bookings – or you can boost bookings by tweaking your conversions with existing traffic.

This type of conversion-oriented testing has led to rapid growth and millions in revenue.

The A/B Testing Process

(Image source)

Following a scientific process for testing is essential. It’s what leads to undisputed data that spikes conversions, so you don’t need to wonder if you’ve tested for long enough or if your results are significant.

Now, a “scientific process” may sound a little intimidating…but it’s not. The process is pretty straightforward and rewarding. Here’s what it looks like:

  1. Finding leaks to plug: This is where you target an element for improvement. It starts with asking a question, like “Where are people leaving during the booking process?”
  2. Doing research: Time to peek behind the digital curtains with analytical tools. This step is all about understanding your visitor behaviour.
  3. Creating a hypothesis: Here is where, based on your research, you come up with a way to plug your leak. For example, simplifying design by adding a progress bar and removing unnecessary add-ons can reduce bounce rate in the booking funnel.
  4. Finding out the number of days/visitors you need to run the test for: You can use VWOs test calculator for this.
  5. Put your hypothesis to the test: This is where you’ll pitch your existing variation (the control) against the one with the theorized improvement, variation B. You’ll be measuring the bounce rate and how many people complete the process.
  6. Analyze and conclude: Did the changes result in a lower bounce rate? If so, you can conclude that decluttering design lowers bounce rate. But what if there’s no difference between variations? Then you go back to step 3 and test a new hypothesis.

What Exactly Can You Test?

Every A/B test will be unique, but common elements for testing are:

  • Headlines
  • Descriptions
  • Call-to-actions
  • Form fields and form length
  • Offers or packages
  • Layout and design
  • Images
  • Long vs short copy

Testing Your Offers for Maximum Impact

(Image source)

Before you A/B test your special offers (or anything else) keep the following in mind:

  1. Start with your biggest, costliest leaks and high traffic pages. Need help finding those pesky leaks? Start with Google Analytics reports.
  2. Run tests for at least a month.
  3. Never stop monitoring results and testing. “Every single day without a test is a wasted day.” Peep Laja, Conversion Xl.

We’ve covered the basics of split testing. Now let’s look at how to use it to boost the power of your special offers.

Power Up Offers by A/B Testing Different Designs

Your design plays a huge role in the conversion process, especially since we’re in the hotel business. It helps potential guests subconsciously form judgements about:

  • What your hotel is like
  • What type of service you provide
  • How they will feel when staying with you

A case study from Mindvalley shows how powerful a simple design change can be. In their A/B test, they tested this header image:

Against this variation:

Not exactly a huge change, is it?

Well, this small design tweak boosted conversions by a huge 230%.

If you want to amp up conversions for your special offers, design is a good place to start.

Improve Credibility by Testing Social Proof in Your Offers

Your offer may be attractive, and it may appeal to your guest, but that doesn’t mean you’ve won your guest over.

This is where social proof can be a great aid.

Social proof tips the scales in your favour, because people are likely to trust impartial reviews and testimonials. When testing social proof in your offers, be sure to test different types of social proof for maximum impact.

Are guests more responsive to written testimonials with pictures? Or are they more receptive to a Groupon-style numeric representation of people who’ve taken a specific offer? Or would a star rating work better?

You’ll have to test to find out.

Ramp up Conversions with a Stronger Title or Headline

Ask any copywriter and he’ll tell you that having a powerful headline is essential.

This is because your headline is responsible for hooking attention. If it fails to click with your potential guest, he won’t be motivated to look at your offer.

Your offer headline needs to convey value. And the only way to find out what your guests value the most is to test out the different options. In a case study by VWO, they switched the headline from “Men’s Clothing” to “Order Men’s Clothing Easily for Bargain Prices”.

The results?

A 127% increase in conversions.

Can you guess why the headline worked better? If you said “because it communicates more value,” you’re right.

A good example of a hotel offer headline comes from The Fitzwilliam Hotel Dublin:

To help yourself come up with better headlines, start by placing yourself in your guest’s shoes. Ask yourself: “Does this seem enticing?” “Would it catch my attention if I was casually scrolling down?” and proceed from there.

Boost the Appeal of Your Special Offers by A/B Testing CTA Copy

(Image source)

“The biggest change on the page doesn’t necessarily result in the biggest lift.” -Michael Aagard, Conversion Rate Optimization Consultant

Your call-to-action is one of the most important page elements. It may be a small one, but it can yield big changes in conversions.

For example, in this case study by Dmix, simply changing CTA button colour from green to red yielded a 34% increase in conversions.

(Image source)

In this study by contentverve.com, changing one word in CTA button copy boosted conversions by 90%.

When A/B testing your CTA button, you can test:

  • CTA button copy
  • CTA button colour
  • The size of your CTA button
  • “My” to “Your” button copy (like the example above)
  • Different phrases: Like “Continue” or “Get my special offer” or “Find out more”

Conclusion

A/B testing is a powerful way to engage potential guests and drive more bookings, but it isn’t a one-off event. You can’t expect major results from your first few tests.

Instead, think of A/B testing as a process. You have to be constantly testing new things to grab the most out of your traffic.

If you are to take one point from this post, it would be this:

Always be testing!