I saw a billboard the other day at a coffee shop in my neighborhood: “One day, or day one?” I don’t often like pithy sayings, but this one resonated with me. Sometimes we put something off because it seems overwhelming or intimidating, when in reality we can achieve our goals by just taking one step at a time.
This saying is a nice way to think about your startup’s testing strategy. When we talked to startups in a May 2021 webinar, they told us one of their biggest marketing tech pain points is feeling that they don’t have enough customer data to work with.
And while A/B testing and analytics might feel intimidating if your startup has low site traffic or you don’t think you have the technical skills, it doesn’t have to be that way. You can start somewhere, and then build on that initial step. Rather than thinking about testing as something to do one day, make today your day one.
If you’re reading this article, you know how important analytics and data are to help you learn about your customers.
You likely know there’s no silver bullet, and that "best practices" can sometimes lead you astray. There’s no one-size-fits-all approach to your audience and customers. If you want to grow your business and be successful, you have to learn about your particular customers so you can optimize your sales funnel.
You also can’t rely on your tool to do all the thinking for you. You’ll have to put on your thinking cap and spend some time analyzing your data.
For instance, you might want to look at your sales funnel to see how users interact with your ads. That way, you can learn about the behavior and demographics of certain user groups—such as those who come to your site from Facebook vs Google vs Instagram.
Focus more time on your customer behavior than the bells and whistles of the testing tool itself. No matter how sophisticated or advanced your testing tool, you’ll likely still need to set aside an afternoon to dig into the data.
If you have low website traffic, you can look at other ways of learning about your customers. Let’s talk through a few of them.
One way to understand more about the user experience on your site or app is through usability testing. While this won’t necessarily give you statistical significance or quantitative data, you can still glean important insights from individual interviews, especially if you do frequent usability tests (say, once a month).
How to test? It’s as easy as recruiting participants—friends, colleagues, customers—and having them perform a simple task on your app or website. Incentivize them by offering to buy them a coffee, lunch, or their favorite snack.
Make sure to keep it minimal and focus on one specific thing (like book a yoga class, buy some merch, or place an online order). You also want to make sure it doesn’t take too long—a max of 20 minutes would be a great place to start.
If you have low site volume, you might not be able to do split testing, but you can still observe marketing analytics on your site through a free tool like Google Analytics. With Google Analytics, you can observe site behavior through metrics such as time on site, scroll depth, and traffic to particular pages.
You might not be able to test alternatives, but you can certainly look for user patterns to learn things like what content resonates most with your audience. Then, you can create more of it.
This tip isn’t about an A/B testing alternative, but rather how to get the most bang for your buck when you’re A/B testing and have limited audience data. According to Shiva Manjunath, a conversion rate optimization marketing manager at GetApp, when you have limited data, you want to go big and bold with your test design.
This is because small changes make it harder for you to reach statistical significance. It’s always best to run tests based on a strong hypothesis, so Shiva recommends you start with an idea of what you believe will happen, then run a test to see what actually happens.
Statistical significance is basically a model that allows you to determine whether any increase in results (such as higher conversion rates) is due to random chance or a true impact. The worst thing you can do is see a five-percent increase in conversion rate in a test and roll it out, only to realize what you saw was actually due to random chance and end up seeing a five percent decrease in conversions. In this case, you'd have rolled out a loser because you didn’t reach a statistically significant threshold. On the flip side, if you have a good sense of statistical significance, you can ensure you only roll out tests that actually help improve your bottom line.
“There’s a difference between what you can and should test,” says Shiva. “You can test green and blue buttons all day, but that won’t necessarily help you learn anything about your audience.”
Instead, Shiva recommends focusing on things that will generate insights about your audiences. He suggests running tests that will lead to insights, rather than to try to generate revenue. Gaining more insights over time will help you figure out how to drive revenue long-term, rather than simply always focusing on the short-term gain of running tests to “win.”
For instance, let’s say you’re running a test to understand how your audience reacts to a new video. You could test putting videos as a hero image above the fold to get the most attention, and you’ll have a more direct understanding of how that video affects audiences, compared to displaying the video in the middle of the page. You can always iterate in future tests and move the video around, but you’ll know immediately whether the video is a net positive or net negative to your user experience.
Regardless of what you choose to test, focusing on bigger-ticket changes about user behavior will be most helpful when you’re dealing with low site traffic.
A/B testing is really for optimizing, but if you need more traffic to do your testing, you could probably benefit from more traffic overall to boost your conversion rate and sales volume. If this sounds like your business, you might want to think about pivoting to focus on building your audience base.
“Don’t fine-tune your site for 10 people,” says Shiva. “Instead, focus on putting fliers up everywhere to get people in the door.”