Not Valid How relevant, non-specific “crisis” language affects conversion

Date Added: April 14, 2020 Research Partner: CaringBridge Element tested: Advertising

When COVID-19 became a worldwide pandemic, the importance of CaringBridge became greater. As people were sheltering-in-place at their homes, the digital connections CaringBridge helped facilitate were more significant—which also increased the urgency to support CaringBridge as a platform. They wanted to find the right way to integrate this into their fundraising message, hoping that it would increase conversion rate. They created messaging that didn't address COVID-19 specifically, as not to be exploitative, but addressed the larger sentiment. Then, they tested this in their highest revenue placement, the tribute widget.

View the experiment »


26.2% lift How a inbound widget affects conversion against a donate button

Date Added: April 14, 2020 Research Partner: CaringBridge Element tested: Advertising

CaringBridge had used their tribute widget throughout the site as a fundraising mechanism. They had tested this widget on their "tributes" pages, where people leave supportive messages to their friends and loved ones, and found that it increased revenue. However, due to a technical setup issue, the mobile version had a different experience. CaringBridge decided to re-test this to determine if the same results held for the mobile experience.

View the experiment »


126.7% lift How pre-checking an email opt-in box affects email opt-ins

Date Added: April 14, 2020 Research Partner: CaringBridge Element tested: Name Acquisition Form

As part of their new user registration process, CaringBridge had always given people the option to opt-in to marketing emails, new content, stories, and helpful articles as they created their account. However, that check box had always been unchecked.

View the experiment »


30.2% lift How using a dedicated landing page instead of a homepage affected new site signups

Date Added: March 26, 2020 Research Partner: CaringBridge Element tested: Name Acquisition Design, Name Acquisition Headline

CaringBridge launched a Facebook marketing campaign to try to attract users to create new sites. Since the homepage was the primary place that new site traffic landed, they initially directed visitors there. However, there were lots of additional links, calls to action, and other content on the homepage. They worried that these additional items might reduce the likelihood that someone would create a new CaringBridge site on that page. So, they developed a standalone landing page on Unbounce, and created a radical redesign with the sole focus call-to-action being to start a new site. They did this because the Facebook traffic call to action was to start a site, not to engage in any of the other content on the page. Therefore, they thought it would be beneficial to focus the call to action on the intended conversion to a new page since they knew the intent of the traffic being driven from the ad. However, they wanted to test this to ensure that it actually did improve conversion. If they targeted this landing page test towards users who had two factors: they were coming from Facebook ads, and they were not logged into CaringBridge.

View the experiment »


33.0% lift How directly asking the visitor to give with a question affects conversion rate

Date Added: March 26, 2020 Research Partner: CaringBridge Element tested: Donation Page Copy

CaringBridge received most of their donor revenue through an area on the site called the tribute widget. This tribute widget asked visitors to make a donation in honor of their friend who they were visiting, and then gave them the opportunity to write a message along with their donation. Over the years, this tribute widget had been tested repeatedly to find the ideal combination of language that would motivate donors to give. Upon reviewing the most recent control, the CaringBridge team noticed that the call to action was a command, rather than a question: "Help keep this site online for them and for you". They were curious to find out if asking a question convert more people than a command. So they drafted a new treatment that asked the visitor: "Will you make a gift to ensure that this site stays online for them and for you?" It was a very slight change to the copy, but a very significant change in how it addressed the visitor. They launched an experiment to see the effect on donor conversion.

View the experiment »


12.1% lift How using quantified social proof impacts someones interest in making a recurring gift

Date Added: March 20, 2020 Research Partner: CaringBridge Element tested: Advertising

CaringBridge was running a popup for certain visitors that had previously visited a CaringBridge site. These visitors were shown the popup when they went to leave the site and it prompted them to become a recurring donor. We had previously tested a personal signer and personal message on the popup without success so this time we decided to run an experiment to see if people were more likely to click through to the recurring giving donation page when they knew how many other people were a recurring donor (social proof). We've seen in other experiments and research how social proof can often increase giving so wanted to see if it would apply here to CaringBridge donors and for a recurring gift ask.

View the experiment »


-46.7% drop How adding a gift array affects conversion on a monthly giving page

Date Added: January 7, 2020 Research Partner: CaringBridge Element tested: Donation Page Form

CaringBridge had begun testing a program to acquire monthly recurring donors. One of the questions we were asking involved how much to ask for on the donation page. Historically, Caringbridge had found that a gift array improved conversion but, since we didn't have an idea of what level recurring donors would give at, we decided to test a form without a gift array to help learn what people would give.

View the experiment »


-29.0% drop How adding a signer to a recurring ask popup affects clickthrough rate

Date Added: November 8, 2019 Research Partner: CaringBridge Element tested: Advertising

CaringBridge was running a popup for certain visitors that had previously visited a CaringBridge site. These visitors were shown the popup when they went to leave the site and it prompted them to become a recurring donor. We decided to run an experiment to see if making the ask relate more to an individual instead of the generic "website" would help increase traffic to the donation page.

View the experiment »


Not Valid How using a brand name in donor acquisition affects conversion rate

Date Added: October 31, 2019 Research Partner: CaringBridge Element tested:

CaringBridge had long acquired many donors through their “tribute giving“ pages, which allow people to make a gift in honor of a friend or loved one using CaringBridge. The language on the donation page let the donor know that their gift made “this site” possible. The CaringBridge customer care team hypothesized that using the name “CaringBridge” instead of “this site” would increase clarity and decrease refund rate. The development team was open to this, but wanted to make sure that using the brand name didn’t negatively affect conversion. They developed a page with the same creative, but using “CaringBridge” in place of “this site”. 



View the experiment »


18.0% lift How personalization and gratitude affects conversion rate and average gift

Date Added: October 28, 2019 Research Partner: CaringBridge Element tested: Advertising

CaringBridge was running a matching grant as part of a fundraising campaign. They had previously tested and found that a sticky bar on the main donation page increases results, as it reminds donors of the limited-time match. They hypothesized that they could further increase results if they made that message come from a person (in this case, Kelly Espy, a CaringBridge employee), instead of appearing like a site-generated message. They also had Kelly’s message thank the donor for their support, presumptively showing gratitude for what they hoped the donor was about to do.

View the experiment »