Eric Goldschein is an editor and writer at Fundera, a marketplace for financial solutions such as small business credit cards. He has nearly a decade of experience in digital media and has written for a number of outlets including Volusion, Marketo, BigCommerce, and RetailNext on digital marketing.
It’s never been easier to start an ecommerce business, thanks to the number of hosted ecommerce platforms and plethora of marketing apps (check out our post on the best Shopify apps as well as the best BigCommerce apps). And with reports saying that ecommerce retail is an industry worth hundreds of billions of dollars, and B2B ecommerce sales cresting $1 trillion, it’s no wonder lots of people are trying their hand at selling online.
But many people get stopped in their tracks right away—by a lack of market need, a misunderstanding of their target market, bad marketing, and other factors. That’s why the failure rate for ecommerce businesses reportedly ranges from 80% to as much as 97%.
To help avoid that fate, if you’ve got an idea for an ecommerce business—or, if you’ve already started one but want to roll out new products, services, or marketing campaigns—you’ll need to test your hypothesis. That way, you don’t sink a ton of time and money into an unproven concept and go broke in the process.
That’s where A/B testing comes in. You can use A/B testing (aka split testing) to try out a new business ecommerce idea, a new product or service—or even just a tweak of your existing site’s copy, call-to-action, or other important feature.
If that sounds overwhelming, we’ll go over what A/B testing is and why you should use it to help guide your ecommerce business decisions.
What is A/B Testing?
A/B testing is when you create multiple versions of the same page, headline, newsletter, copy, etc., and see which one is more effective at garnering attention, clicks and conversions (more on these terms below), and other actions that hopefully lead to you selling more of your stuff.
In a simple A/B test, you would have a “control” version (your current iteration) and a “variant” (with one variable changed). You can also run tests with multiple variants.
To most users, a web page or newsletter subject line appears static. With an A/B testing tool, however, your website, newsletter, or other digital presence can have variations—and if you find that one variation performs better than another in terms of driving people towards your conversion goal (whether that’s more sales, more sign-ups, more survey responses, or something else), you can identify what it is about that winning variation that is so effective.
Those variations could be your descriptions, the number of fields a prospective customer needs to fill out to sign up for more information, page layout, color scheme, pricing, and more. Typically, you would choose just one variable to change per test—that way, if one outperforms the other, you don’t have to guess whether it was the color scheme or the pricing that enticed people.
You can, and should, A/B test throughout the life of your ecommerce business, especially once you have steady traffic. The process never ends—there is always something you can optimize and improve.
What A/B Testing Can Tell You About Your Ecommerce Idea
In the context of this article, an ecommerce idea is any new idea you have that could help you make more money selling online. That could mean a whole new business, a new product/service, or a single new concept like a rebrand of your website or new conversion goals (such as getting people to sign up for your email newsletter).
Whether your ecommerce idea has been failing to gain traction, sales have plateaued, or you’re happy with how business is going, there are many benefits to A/B testing what you’ve built. The same goes with your newsletter subject lines or social media ad copy. With A/B testing, you can:
Improve CTR and CVR
Click-through rate (CTR) and conversion rate (CVR) are two of the most important metrics in ecommerce. The former is the percentage of people who click your ad after seeing it; the latter is the percentage of people who complete your call-to-action once they land on your site.
If you already have an ecommerce business: Are you underperforming on any of these numbers? When you’re not getting clicks, or when people are immediately moving away from your site after arriving, it’s time to A/B test and see if there’s a flaw in your business idea, or just in its presentation.
Unearth the True Needs of Your Customers
Pay close attention to the copy of your best performing A/B test variations. What pain points does that copy address? Does it focus on the speed of your solution? The price? The way it integrates with existing solutions?
One way to figure out what your customers want is to ask them. But as useful as online surveys are, they can be a bit misleading, because what someone says they want may be different from what they end up actually committing money towards. By measuring how many people convert on certain types of copy, you get a real sense of people’s actual needs.
Learn Your True Price Range
What are you charging, or planning to charge for your product? Is that number based on market research as well as what competitors are charging? Have you considered what charging an extra $5 per sale, or $10 per subscription, could do for your business?
A/B testing landing pages with different prices and seeing how far you can push your price range before there’s a precipitous dropoff in conversions is an excellent way to maximize your profits.
Improve Results From Ad Spend
Nobody likes throwing money away on advertising that doesn’t work. But that’s what you’re doing if you’re settling on just one social media or search display ad and not looking to optimize it in any way.
Some ad models charge you for each person that clicks on your ad from the third-party platform. Why are you paying for people to visit your website and then leave? Granted, you won’t get a 100% conversion rate—but converting just a few extra customers a day or month could be the difference between turning a profit and needing a business loan to stay afloat. A/B test both your ad copy and the landing pages your copy leads to, and not only will you get a better return on your ad spend, but you’ll feel more confident in increasing your ad budget.
Avoid a Costly, Unproven Redesign
Redesigning your website can be a hugely expensive endeavor, especially if you are doing so blindly, on a gut feeling that your web presence isn’t cool, modern, attractive, interesting, or whatever-enough.
Instead of redoing everything all at once, A/B testing allows you to change things piecemeal, over time, as needed—identifying what specific aspects of your site aren’t working for customers.
Remove Friction Points
Finally, where are you losing your customers? Do they abandon their shopping carts just before purchase? Do they balk at talk of shipping, or when asked to complete a profile? Does your site take so long to load on certain devices that they don’t stick around to see what you’re offering?
A/B testing allows your customers to highlight your friction points for you, and gives you the greenlight to address those friction points without disturbing the rest of what works about your business.
How do You A/B Test an Ecommerce Idea?
Let’s run through how you can use A/B testing to make better decisions when either starting a business, rolling out a new product, or even making a simple tweak to your model or appearance.
A/B Testing a New Business
Typically, A/B testing is only useful and significant when you have enough traffic flowing through your website to obtain statistically significant results. An A/B test that tests the behavior of 10 people can’t be considered valid, as the chance for outliers messing with your results is too high. A/B tests usually pit thousands of experiences against thousands more.
That being said, A/B testing can play a useful role in determining whether an ecommerce business is a viable idea. It’s most helpful when that business is a B2B, software-as-a-service style company.
Why? B2C companies need thousands of customers—honestly, more like hundreds of thousands or even millions of customers—before it can become viable. Because businesses can pay more for a service than an individual, you’ll only need a few hundred customers to become viable if that’s your market.
Once you have an idea for a business—as well as a good sense of who your customer is, what the product will look like, a name and logo, etc.—it’s time to start testing and see if your idea has traction. To do this, you need to set up landing pages.
The landing page test is rather simple. Use a service like Unbounce or Instapage to set up a landing page that describes your service, and asks visitors to enter their email address to learn more, gain early access, or receive an alert when you go live. To send traffic to these pages, you can buy a few targeted ads on Facebook and Google, the main ad platforms on the web today. Even a minimal amount of investment—say, $100 each—can deliver results.
Then you measure your conversion rate and see if there’s enough interest to build your product.
Where does the A/B test come in? Well, you’re certainly not going to build just one landing page. You can build as many as you want, changing an aspect of each. By A/B testing these pages, you’ll get a better sense of how engaging audiences find your business idea, design, price points, copy, and more.
What if you build 10 versions of your landing page, and your conversion rate still stinks? That’s a problem—and means it's time to go back to the drawing board. If you get decent conversion on a couple of iterations? Clearly, you’re on the right track.
A/B Testing a New Product
You can use a similar tactic for testing out a new product or service that you want to roll out in addition to your current offerings. The landing page test works the same way—and even better, you should already have enough traffic coming through your ecommerce platform that any sort of product test (whether it’s consumer-facing or enterprise-facing) will be viable without targeted ads.
When considering a new product, you can create landing pages just as you would for a brand new business, and use your existing channels (including paid and organic social and blog posts) to direct traffic to them, gauging interest by means of asking customers to “Drop your email here to learn more” or “Sign up to get an alert.”
Landing page tests will tell you whether the idea is a popular one. Where A/B testing comes into play is that it can help you make important decisions about your new product before going live. For example, you might find that customers are willing to purchase it at a price point above what you expected, increasing your revenue.
You can also play around with your descriptions in your landing pages, and based on the conversion rates of each description, you can tweak your product to better address the needs of your customers. Perhaps your customers have a different pain point than what you assumed based on your research.
Testing a new product before rolling it out is important. A sure way to tie up your capital is to assume that a similar or related product offering to what you already have will be a hit, without getting a sense of whether you’ll get a decent return or not.
Cash flow crunches doom even the most successful businesses—so you need to make sure that your business idea has legs before moving ahead with it.
A/B Testing a Tweak to Your Business
Within the context of your existing ecommerce store, there are seemingly endless ways you can A/B test the user experience in order to increase conversions—including, but not limited to:
- Page layouts
- Body copy
- CTA copy, location, and design
- Form length and style
- Product images and infographics
- Different price ranges and tiers
- Offers such as free shipping or discounts
- Pop-ups and prompts
- Checkout process
- Additional payment options
- Reviews and testimonials
- Related or similar product recommendations
And many more.
As mentioned above, the best way to conduct an A/B test is to test one factor at a time, choose the best performing variant, then make that your control and continue testing other variants or move on to the next factor.
You’ll need a clear roadmap of what’s working and what isn’t on your page to understand what to test first, and why. You can use surveys, heatmaps, analytics, and other tools to identify what might be holding up conversion.
This might sound like a lot of work to you. What if you’re running a profitable ecommerce business, and don’t see the point of optimizing further?
That’s not the right mindset to have, however. If there’s a way to improve profits and boost the bottom line, you should look into it. You never know when your expenses will go up, or your customer base will decline—and suddenly you’re facing a much slimmer profit margin than you planned for. Never leave conversions—and thus, money—on the table.
The Bottom Line on A/B Testing
A/B testing is about finding what works and helping you do that better—it’s as simple as that. Don’t let your vision for your business and your biases about what your site should look like or what a customer should behave like hinder you from growing your business. You need to let your customers help you help them.
It’s easier than ever to test nearly every aspect of starting and running your business. Take advantage of these tools now—future iterations of your site, and your higher-than-expected conversion rate, will thank you later.