Best Practices for Split Testing

A vs B – Best Practices for Split Testing That Really Works

Paul Rose Jr Content Marketing, Email Marketing, Get Strategic, Growth Marketing, Inbound Marketing, Small Business Marketing Leave a Comment

You did your strategic work. You built a great website, even pulled some information together to create your own personalized, laser-focused, Lead Magnet that should work to start driving people into your sales funnel. You’re watching your Google stats, and people are finding your website, but…

Maybe you’re not getting the response you were expecting. Or you spot a logjam in the funnel at some point. Or something else is happening. For some reason, the website visitors are not turning into the paying customers you were expecting. Luckily, with software like Infusionsoft you can pinpoint exactly where the process is succeeding and failing. But the question remains…

Why is it failing?

One thing we’ve found over the years at Grow the Dream is that sometimes as small business owners, we focus too closely on the end result and not enough on the process of getting there. And that’s no surprise. As a small business owner or solopreneur, you don’t have a huge marketing department, or advertising group, or sometimes even an office manager to keep track of some of the smaller issues.

If you’ve followed our recommended marketing approach – key word strategy – you know where to begin – with your strategic work: determining your ideal customers, the questions they’re asking that you can answer and how to answer those questions in a way that they ‘hear’ and gain value from the answers, even before they start paying you. A crucial part of the strategic work, though, is on-going testing. This isn’t ‘set it and forget it’ marketing any more than it’s ‘cookie cutter’ or ‘one size fits all’ marketing. So once you’ve established a starting point, A/B testing is a great way to refine and develop your focus to reach more of your perfect customers.

It may help to think of online marketing as a cross country road trip to your small business success.

Your initial strategic work has helped you map out a route to your perfect customer. But no route may be the perfect, or only, one, and, in doing your strategic work, you may have also found some side paths, shortcuts or alternate routes that also seem to lead to your ideal customer. Some of them may be minor route adjustments and some more extensive. There may be an alternate route that can get you there a little quicker, but speed isn’t always your goal. What if a different path could provide free accommodations, with friends or family? Or, to really stretch the analogy, lets you check out the world’s biggest rubber band ball? If those things are important to you – or, rather, your ideal customer, how do you choose the best overall route?

The Key is A/B Testing

Much like a map legend, A/B, or split testing takes a portion of your audience and exposes it to an alternative – whether it’s a different landing page, a different email headline, a different call to action, a different offer for your products or services, or even just an adjusted color scheme.

You can either test two options on similarly sized samples simultaneously, or you can test a new approach on the entirety of a group, comparing it with the results of the initial campaign under similar circumstances.

Although I recommend that everyone start small with A/B testing, the beauty of it is that it’s never truly done, and split testing should become not an occasional adventure, but a regular journey you take with your online marketing materials, always looking to improve your impact and narrow your focus.

A/B testing helps you do one of two things, which are themselves intricately linked – it improves how you approach your ideal customer, by gaining a better understanding of who your ideal customer is and what they truly want.

We want to better understand the customer, so we can better serve the customer.

Let’s start with how you approach your ideal customer.

There are six key factors to taking a complete stranger who found your company on Google to making a sale and beyond. Each one can be improved and developed using A/B testing, with one exception.

Motivation

Unfortunately, so far there is no way that marketers have found to increase motivation. Your prospects either have it or they don’t. And when we try to encourage an unmotivated buyer into making a purchase, we come off as the slimy, pushy used car salesman everyone loves to hate, no matter how innocuous our approach. But a good thing to remember, is if the person found your website by searching for the key terms you are marketing, there’s a good chance they’re at least somewhat motivated to buy. So we want to cement the beginnings of that relationship as quickly as we can.

The best way to do that? Find a way to:

Increase Perceived Value

This is where the beauty of the Lead Magnet comes into play. By offering something of true worth to a prospective customer long before they start paying you, you are raising the perceived value of your company. If the content of the Lead Magnet is also relevant, it raises your profile in the eyes of the prospective buyer.

If your Lead Magnet lacks value, relevancy and urgency, the prospective customer will perceive your business as less worthy. Even the exchange of an email or the time to fill out a simple form should be rewarded with a benefit, so as to keep increasing their perception of your value.

But don’t stop there. Build value to:

Raise Customer Incentive

Keep building and reinforcing your business’ perceived value by offering better incentives as the prospect inches their way down your sales funnel to becoming a customer. The more incentives you offer that are of value to the prospect, the easier they will move, and the less you’ll have to worry about the other side of the sales equation:

Reduce Customer Friction

Friction is defined as psychological resistance to something in the sales process – whether it’s obvious or in the prospect’s subconscious. Sometimes it’s cost, but more often there is just something about purchasing from you that elicits fear for whatever reason.

The easier you make it for the customer to purchase, while continuing to build credibility, the more likely they are to convert to a sale down the line. Because you never know what can get you to:

Reduce Customer Anxiety

As much as we like think of ourselves as businesses, the fact is, even if your product or service is B2B, you’re still selling to a person within that business. And people have irrational fears. You’ve heard this before – we buy based on an emotional reaction, then justify our decisions.

So there’s not always a clear cut way to reduce anxiety, because even the customer might not know why they’re reacting that way. They’re trusting their gut without recognizing that multiple factors that have nothing to do with this sale are affecting their judgement.

But if you can overcome the anxiety and friction, you’ll be able to:

Increase Rate of Conversion

While this is our bottom line, we have to remember that’s not our customers’ goal. While it’s difficult sometimes, the best way to increase conversions from prospect to sale and then to evangelist, is to put ourselves in our customer’s shoes to master the first five marketing and sales factors.

So at this point, you’re probably wondering ‘Why the Marketing 101 lesson?

The more you learn

about your ideal customer,

the more you can test

and refine.

For one, it never hurts to review the basics. And… Each of these five factors is a sign post on your road trip that can lead to a new path, a worthy detour or an actual shortcut. Plus, by digging down into each factor and how you can adjust it with A/B testing, you’ll be learning more about how your ideal client thinks, reacts and makes decisions.

Everything Can Be Tested

  • Branding
    • While you will want to try and keep your branding as consistent as possible, it’s okay to veer off your main focus to test both the branding and how it works in each piece of your marketing.
  • Email/Newsletter blasts
    • Headline & subheading/Subject line
    • Personalization
    • Contents
    • Call to Action
  • Website
    • Design
      • Functionality
      • Ease of navigation
      • Ease of getting assistance
      • Value to customer up front
    •  Copy
      • Value to customer up front
    • Landing Pages
      • Headline/subject line
      • Copy
    • Testimonials?
    • Mobile traffic vs. desktop traffic
  • Blog Posts
    • Headlines
    • Graphics
    • Copy
    • Voice
  • PPC
  • Lead Magnets
  • Telephone Call Scripts
  • Social Media Ads

I know that all of that might seem overwhelming at first, but the good news is that you don’t have to tackle it all at once. In fact, you can’t. And if you want to do it right, you shouldn’t!

Test One Thing at a Time

Remember what I said about small and large changes? To truly get in deep and get effective results from A/B Testing, you can only make one change at a time. How else can you measure the impact of each test and sample results? Even small or subtle changes can make a difference in the long run.

Go For the Greatest Impact

It’s sometimes crazy what some businesses discover while split testing. Even to the point of slightly changing the punctuation – a comma instead of a semicolon (true story) – can alter conversion rates.

However, you have to start somewhere, so test the bigger, more apparently impactful things first, then work your way down to the smaller stuff.

Know what you want to test and why

This might seem obvious, but you wouldn’t believe how often even larger businesses leap into A/B testing, having no idea what they’re testing or why. Another good reason for the Marketing 101 lesson.

Know your desired outcome

Yes, converting a prospect to a sale is the primary goal, and the ultimate indicator of the success of a split test, unless you’re testing why there’s a traffic jam on the third level of your sales funnel. While conversions are the ultimate goal, you need to get the prospects moving again to accomplish that. So always be clear on what you hope will change in the process that will ultimately lead to a successful outcome for whichever piece you’re testing.

Create your hypothesis, but don’t be married to it

By the same token, you should start each A/B test with a desired outcome and a theory of why whatever it is you’re testing could be improved. However, you HAVE to trust the data. Don’t try to fudge or manipulate it to fit the conclusion that you’re hoping for – you’ll only cost yourself results in the long run.

Use a large enough sample

While there’s no numeric indicator of how sizable your sample needs to be to get an accurate measure, it does have to be significant. Otherwise, how can you be sure that the testing is giving you the correct results? A good rule of thumb would be to test at least half of your regular visitors, email subscribers, etc.

If you finish the test with too few prospects, you might confuse results with random chance. To truly learn from the test, you need the high level of confidence that comes with statistical validity – you want to make sure your sample group is a true representation of what would happen if instead of testing, you launched the change for your entire audience.

Timing is Everything

Along with making sure you have a large enough sample, it’s important that you let the test continue until you have sufficient data. Don’t get impatient and end the test early. Let it run its course so you can accurately judge the difference between plan A and plan B.

Some marketers recommend ending on the same day you started, so an A/B test that started on a Tuesday will last two weeks – 15 days, to also end on a Tuesday.

Also, you’ll want to run the test when it’s reasonable to assume that you will actually get a sample to test.

Document, measure, adjust and repeat

Just like shampoo instructions, Split Testing is a process that grows value with repetition. Like any other marketing effort, you need to track its effectiveness, with measurable metrics. Keep track of each test, what was changed and the final result. Also keep track of when you are doing it – weekdays, weekends, what month, what stage of the sales cycle.

You want to be a connoisseur of data, so that you fully understand what it means and what steps to take next, whether to test some more, implement the changes, or just leave it alone.

Never stop… Testing

Really, A/B testing is never really complete, because there are always more things to test, more ways to narrow your focus and find the best route to your ideal customer, and let the marketing take you there.

In the end, even if you discover that the best route to go is the one you were already on, exploring the side paths and shortcuts will help broaden your creativity – leaving you with more options to draw from in the future. And a better understanding of the customer you’re hoping to serve.

About the Author

Paul Rose Jr. started his career as a TV journalist, writing and producing the highest-rated weekend newscast in Southwest Florida. In the 25 years since, he’s written in just about every medium and format available: radio, TV, print and online journalism, commercial and promo copy, op ed articles, press releases, website copy & blog posts. Story is his passion and it’s at the heart of all he writes. All good writing involves story – it connects with people and opens their mind and heart to hear what you have to say. Paul is a voracious reader, and draws from a wide-range of research and life experience when writing. He is currently in Los Angeles, CA, and in addition to the work he does for Grow the Dream, he is hard at work developing screenplays and teleplays.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.