top of page

A/B Testing Best Practices for Your Website’s Paywall

Updated: Mar 16

With the increasing prevalence of paywalls, one of the questions that is at the top of everyone’s mind is, “Will a paywall work for me?” The truth is, creating a paywall strategy that works is not a one and done deal; it’s a process that includes the consistent effort to understand and target your users based on data-driven decisions. A/B testing is one of the best tools for publishers to hone messaging and improve ROI.

The Process of A/B Testing

A/B testing can begin before you even implement a paywall. Here are some initial steps to set up your A/B testing strategy for success. In our Launchpad Process, we measured the progress of 4 publishers over a 6 month period, and here’s what we’ve found to be effective.

1. Analyze

The first step is to start by running analytics in the background to gather data on audience behavior including engagement and frequency metrics. This helps determine subscriber likelihood. Subscriber likelihood score is based off of a number of factors and can include data on how long a reader spends on articles, how many pages they navigate to, or how often they return. To add some additional layers of complexity, geographical location can play a large role in the results, so testing that takes into account the variation between geographic regions is important.

2. Test

The next step is to test your hypotheses by targeting only a subset of your customer base with a paywall. Some important things to test with this paywall are price points, messaging, and targeting to determine what customer journey converts best. Generally, newspapers choose to start with metered paywalls while magazines opt in for premium content paywalls and membership perks.

3. Iterate

A/B testing requires constant finessing, and it’s important to test not just different variants, but different combinations of different variants as well. Setting aside a period of a few months to test different variations can yield significant results. After testing three different major areas, we found that each additional test improved conversion rates by an average of 400%.

What should you A/B Test?

The sheer multitude of variables that you can experiment with may seem daunting, but investing in the right technology lets you test a large number of variables easily and gives you actionable data to hone your paywall strategy. Here are some different things you’ll want to test with your paywall.

1. How Much to Charge

One of the most obvious tests you can run is optimizing for different price points. Especially if you’re just starting out with a paywall, you’ll want to make sure that you’re not pricing so high that you drive readers away or so low that you could be missing out on potential revenue.

Our studies found that the conversion rate between $2.99/month and $7.99/month were exactly the same. This means that with each transaction you could be leaving a lot of money on the table if you don’t understand what your audience is actually willing to pay for. Interestingly, the middle range of $4.99/month got the fewest number of conversions, so not only are you losing out on potential revenue at each purchase, you’re also getting significantly fewer subscribers as a whole.

2. What to Say

Copy and messaging are extremely important when you make your subscription offers. When A/B testing, think about what your potential reader base may be motivated to invest in. This goes far beyond the actual content or specific articles. For example, copy that includes messaging about a network, such as offers that include access to multiple sites or multiple offerings, got on average a 30% increase in conversions.

Likewise, copy that included community messaging saw a 25% hike in conversion rates. You can build community in a variety a ways. This might mean something as simple as playing a part in supporting publishing and high quality journalism. It could also include exclusive offers, perks that come with memberships, or specific access to limited groups. Giving subscribers something like commenting access, subscription sharing, or a better browsing experience are all ways publishers can create a sense of community with their members.

How you position your pricing is also important. Which subscription period are you offering? Are you breaking down the pricing description by telling them how much they’ll be paying in monthly, yearly, or weekly amounts? Thinking out each element of the offer and how it’s positioned or expressed, you can then tailor that for each variable such as geography, device, and demographic and test which has the most success for those specific audiences.

3. How Many Free Articles to Give

The number of free articles is an important metric to analyze since too high a number could have you giving away your valuable content for free and too low could drive away customers who might have converted if they had an opportunity to browse, share articles, and determine that your content is really what they would like to invest in.

Surprisingly, a tighter paywall with fewer free articles often leads to higher conversions. In our testing we found that 10 free articles led to only a 34% conversion rate, while 3 articles led to a 77% conversion rate.

4. Which Devices to Target

User behavior, engagement, and a host of other factors can vary by device type. It’s likely that users who typically access your site via a mobile device have a different customer journey than those who habitually enter via a windows or mac, for example. There could be a range of factors that influence this including the way these potential subscribers consume content, or more demographic-related reasons such as economic factors or age.

For example, those who enter through mobile devices might be more likely to purchase subscriptions to daily editions, but testing could also reveal that those accessing the site through laptops are more willing to purchase weekly editions.

You can also try testing various payment option defaults by device type, so if, for example, a user accesses via an iphone, offering Apple Pay instead of credit card may result in higher conversions.

5. Which Articles to Put Behind a Paywall

Every user will have varying interests and come onto your site for different reasons. It can be difficult to know which content to put behind your paywall so that you have enough outside the paywall to entice readers with topics they’re interested in, but also not give away so much of the content that they don’t need to come back for more.

Some users might pay for auto articles or political topics, for example, or other specific, niche content. Knowing which articles readers are willing to pay for can help you develop not just your subscription strategy, but your content strategy as well, allowing you to focus on quality content in the key areas that offer extensive value to your members.

6. Which Paywall to Show and When

Making sure the reader has a positive overall buying experience with your paywall is a must, and there are a couple of different factors that can contribute to this.

We’ve seen conversion rates that drop significantly after being taken to another page to subscribe. Having the whole process happen right there on the same page will greatly increase conversion rates and help your subscribers have a smooth experience.

Time spent on page before seeing a paywall displayed is also important to test. We find that displaying a paywall before user has scrolled to 20% of the page shows a significant drop in conversion rates while 40-55% has proven to be optimal.

7. How Often to Show the Paywall

How many times does it take to get your readers to sign up through your paywall? If you show a paywall too soon it might create a negative user experience and drive potential customers away. It might block them from really exploring and investing in your content or promoting it on social media. But if you show it after giving away too much you risk devaluing your content.

Part of the complexity of all A/B testing is that every one of these factors varies according to a variety of other factors, as we’ve seen, such as geographical location, device type, etc.

We’ve run something we call the “Stubbornness” test that ascertains the optimal amount of times it takes to get readers to subscribe after being presented with a paywall. The results varied wildly from region to region, again highlighting the importance of continuous testing.

In Canada the best conversion rates were seen after presenting the paywall twice at a 76% conversion rate (with 5 times being a close second). In the US the highest converting number was 5 at 85% conversion and in the UK 6 at 71%. The results showed that the common misconception that showing a paywall too many times will cause readers to bounce is not accurate since most subscriptions occurred after seeing the paywall multiple times.

There is always room to improve

Paywall success is not an accident. It’s a continuous process, but it doesn’t need to be a difficult one. Pelcro is built for testing multiple different variations so that you can easily create a multitude of tests and determine your conversion tipping points.

While the growth of reader revenue has led to some positive restructuring within organizations and the development of entire teams focusing on optimizing the customer experience, many companies still don’t have the time to do all of this manually or dedicate full time staff members to focus on conversion. Having automated technology that allows you to easily test, track, and gather data from these variations is key to having a successful A/B testing strategy.

Next Steps

Deciding where to start can get overwhelming. That’s why Pelcro doesn’t just give you all the tools you need. Pelcro’s 3-month Launchpad Process is designed to get you up and running in no time. Throughout the process, you will learn everything you need to know about the platform and get recommendations from industry professionals. To find out more, contact us!


bottom of page