Experimentation ideas for nonprofit organizations

  • Updated
  • Optimizely Web Experimentation
  • Optimizely Performance Edge
  • Optimizely Feature Experimentation
  • Optimizely Full Stack (Legacy)

A thoughtful, well-designed experimentation program can help a nonprofit organization increase donations, drive awareness and engagement, and connect more strongly with the communities it serves. This article provides a few quick test ideas that have worked for several of Optimizely Experimentation’s nonprofit customers in the past, and that might be easy opportunities for you to get started experimenting today. Use them to learn about setting up experiments and to explore ideas for your own site.

The following ideas are built in Web Experimentation using the Visual Editor. You can implement the ideas using Optimizely Feature Experimentation and Optimizely Full Stack (Legacy) by creating an A/B test and implementing the Decide methods.

Idea 1: Increase the visibility of your CTA with an image

Sam has noticed that newsletter signups for his organization have been steadily decreasing. Since email is a primary source of donation generation, as well as a powerful means of connecting to donors, his department wants to make a change to the newsletter CTA to reverse this trend. Since they want to keep the email signup box in the same location (near the top of the page), he thinks improving its visibility might be enough to drive visitors to engage with it more.

  • Hypothesis – Replacing CTA text with an image will make the email signup button more noticeable and engaging, which will increase email submission signups.
  • Pages – the URL of the page where the email sign up field is located
  • Audiences – Everyone
  • Metrics – Clicks on form submission button (primary); views on the form submission confirmation page (secondary)
  • Required Technical Skills – None

Build the experiment in the Web Visual Editor

  1. In the experiment, select the variation you want to work with.
  2. Click Changes, followed by Insert Image.
  3. Using the selector tool (_), select the CTA text you want to replace.
  4. From the drop-down menu, select either Before or After to specify where you will place your image in relation to the existing text. You can play around with this positioning after you upload your image.
  5. Upload your image.
  6. Click Save.
  7. Click Changes, followed by Element Change.
  8. Using the selector tool (_), select the CTA text again.
  9. Under the Visibility section, select the remove icon (_).
  10. Click Save.
  11. Use the Preview tool
  12. When you are ready, click Start Experiment.

Interpret your results

  • If the variation is a winner, Sam has confirmation that visitors responded to the image. He will want to direct all traffic to the winning variation, then consider other images to test.
  • If the variation is a loser, this could mean that the image deterred people or that there is a different problem he has not considered. This will require more hypothesizing and testing around how to increase the newsletter subscriptions.
  • If the test is inconclusive, this could mean that a new image was not a bold enough change. Sam may want to test different images or CTAs next, or consider changing the location of the signup box while still maintaining its high level of visibility.

Idea 2: Re-focus your CTA messaging on your community

Breana’s nonprofit works to expand participation in computer science education by making it part of the core curriculum in education. She wants to increase signups for an upcoming one-day learning event called An Hour Of Code. She thinks their current messaging for the event is a bit detached, and would like to experiment with a message that resonates more with visitors, so they feel like they are going to be a part of something important.

  • Hypothesis – By adding a more community-focused message and button for the event CTA, visitors will feel more connected to the event, which will increase participant signups.
  • Pages – the URL of the page where the event CTA button is
  • Audiences – Everyone

  • Metrics –

    • Pageviews on the sign up confirmation page (primary)<
    • Clicks on learn more and join us buttons (secondary)
    • Pageviews on the information page (secondary)
    • Clicks on the sign up form submission button (secondary)
  • Required Technical Skills – None

Build the experiment in the Web Visual Editor

  1. In the experiment, select the variation you want to work with.
  2. To edit the CTA, click the text in the Editor to select it.
  3. In the HTML box, edit or replace the text with your new CTA. Do the same for the button if you wish to change its text.
  4. Click Save.
  5. Use the Preview tool to QA your experiment.
  6. When you are ready, click Start Experiment.

Interpret your results

  • If the variation is a winner, Breana’s hypothesis has been supported: Visitors responded positively to the more community-driven tone of the messaging. Next, she should consider how to apply this insight to other aspects of the site. She can then test these ideas for both returning and new visitors.
  • If the new messaging and CTA is a loser, this does not necessarily invalidate the hypothesis. Breana might want to test other community-focused messaging, perhaps using multiple variations in a single experiment. She could also segment the Results page to see if one subsection of visitors performed better than others; if so, she could target more community-driven experiments to those people.
  • An inconclusive test suggests either the hypothesis itself is wrong or the change is not bold enough, because visitors continue to behave in the same way. Breana can do more research and brainstorming on issues the community is most passionate about, or techniques that have driven visitors to sign up in the past. From there, she can come up with a new community-focused message, or try changing the layout and positioning of the CTAs.

Idea 3: Change the pre-filled suggested amount in the donation input box

Contributions to Lakshmi’s nonprofit have been stagnant over the last year. Almost all their donations come through the website’s Contribute page, which pre-fills a suggested donation amount in a text box for the user’s convenience. This amount has been set at $25 for some time.

Lakshmi has noticed that most donors give only once, and that the vast majority of donors simply accept the suggested donation amount. She thinks her organization can increase that amount without reducing their overall conversion rate.

  • Hypothesis – It is possible to raise the pre-fill average amount without reducing conversion rate, thus increasing their overall contributions.
  • Pages – The URL of the page where the donation box is
  • Audiences – Everyone
  • Metrics – Conversion rate (primary)
  • Required Technical Skills – None

Build the experiment in the Web Visual Editor

  1. Select the variation.
  2. Click the field containing the pre-filled donation amount.
  3. Under Selector, click the element selection button. A drop-down list of all the HTML elements attached to the donation field will appear.

element_selector.png

  1. From that drop-down list, select the element that includes only the donation field itself. You will know you have got the right one when you see the donation field, and nothing else, highlighted in blue in the Editor.

donation_box_before.png

  1. Under HTML, find the placeholder text and update it to reflect your new suggested donation amount.
placeholder-update.png 
Your change should be immediately reflected in the donation field:donation_box_after.png
  1. Click Save.
  2. To QA your experiment, use Optimizely Web Experimentation's Preview tool.
  3. When you are ready, click Start Experiment.

Interpret your results

  • Since the goal of this experiment is to find the highest suggested donation amount that does not negatively affect conversion rates, Lakshmi will likely have to try several different amounts before she finds it. Depending on her site’s traffic, running several variations at once can help her do this more quickly.
  • It is possible that the current suggested donation amount is already at a conversion-maximizing point. If this is the case, Lakshmi will see conversion rates falling for all her variations, and she will have to come up with a different strategy for boosting donations.

This example was inspired by a real Optimizely Experimentation customer case study.

Idea 4: Reorganize your site design to drive more traffic to your donations page

Asok noticed that the click-through rate from his organization’s homepage to its donation page had been lower than the average for nonprofits in his sector for the past six months. The homepage contained a large number of different elements, all competing with each other for attention. Asok suspected that if his organization redesigned the homepage with a simpler layout that highlighted the link to the donations page, traffic to it would increase, and so would contributions.

  • Hypothesis – A clear, user-centric homepage design will help more people find the link to the donations page and contribute.
  • Pages – Home page
  • Audiences – Everyone

  • Metrics –

    • Clickthroughs (primary)
    • Bounce rate (secondary)
    • Conversions (secondary)

  • Required Technical Skills – None 

Build your experiment in the Web Visual Editor

  1. Select the variation you want to work with.
  2. In the Editor, delete any page elements you wish to remove from the page.
  3. Reconfigure the remaining elements to match the layout your design team came up with.
  4. Ensure the link to the donations page is prominently displayed and appropriate for the overall design of the page.
  5. Click Save.
  6. To QA your experiment, use Optimizely Web Experimentation's Preview tool.
  7. When you are ready, click Start Experiment.

Interpret your results

  • If the variation is a winner, Asok’s hypothesis has been confirmed: the homepage’s original, busier design was impeding visitors from finding the Donations page. He will want to direct all traffic to the winning variation, and then consider other potential design improvements to test.
  • If the variation is a loser, this could mean that the root cause of the problem is something other than the homepage layout. Asok will need to do some more hypothesizing and testing around how to drive traffic to the Donations page.
  • If the test is inconclusive, this may mean that the new layout just was not a bold enough change. Asok may want to try a more radical redesign and run the experiment again.