Image via Scott Webb under CC0 1.0

What eHarmony Can Teach You About Testing Social Media to Prove ROI

We know our content creates awareness—but does it drive revenue? In this post, we’ll share three techniques to set up your first social media test. These techniques are based on what we learned from our customer, eHarmony.

In this piece we’ll cover:

  • How testing revealed eHarmony’s most profitable type of social content
  • How to test your social content in three steps
  • Tips for your first A/B test

How testing revealed eHarmony’s most profitable type of social content

eHarmony is the No. 1 trusted online dating site for singles for a reason: their match-making algorithm works. They also know how to improve campaign and marketing performance.

With testing, eHarmony tracks and improves content based on how people respond in clicks and conversions. “We have a longer funnel in a crowded and highly competitive market,” Kerianne Mellott, the Director of Social Media at eHarmony tells us. “Once you decide that eHarmony is a match, you take our relationship questionnaire and then subscribe. We’re challenged to find our target audience on social media and get them to click through to subscribe.”

Small tests allow eHarmony to segment their audience and find out what content performs best with who and why. Working with eHarmony’s IT department, Mellott integrated Hootsuite and Google Analytics with existing eHarmony code in order to monitor all data coming from social. By tagging all content shared from Hootsuite with UTM parameters, eHarmony can track multiple steps of conversions with analytics.  

“We know which ad or campaign they clicked on that landed them on to fill out our questionnaire and can track ROI easily,” says Shannon Gallico, the Social Media Specialist, at eHarmony.

Through testing, eHarmony discovered that “Powerful testimonials on social media drive more people to our site for signups,” says Mellott. “As a love-dating relationship company, people look for those success stories as an affirmation.”

3 steps to test and improve your social ROI

1. Establish a benchmark

Before you too can come up with a hypothesis like testimonials drive more conversions, you must first track and analyze your daily activity. This will help establish your social media benchmark. We recommend tracking your analytics in a spreadsheet. Consider what type of content you share, where, when, and to whom. Success could be measured in engagement metrics, such as likes and comments, or in deeper metrics, such as leads, traffic-to-lead conversion, and demo requests. Use analytics to monitor daily, weekly, and monthly performance to get a sense of what’s normal.  


2. Make a hypothesis 

Understanding how your social content usually performs will help you spot trends in your data. For example, you might notice that messages shared earlier in the day receive higher engagement. Or perhaps more fun, personable messages converts better on Facebook than on Twitter. Come up with several hypotheses to test (one at a time) in order to improve your social ROI.


3. Launch A/B tests on your social

A/B testing, or split testing, is a great way to test your hypotheses. Basically, various marketing variables are compared against each other to find the top performer. One element doesn’t change and is called the ‘control’ (or version A). Your hypotheses element is called the ‘treatment’ (or version B). A/B testing allows you to learn how to drive more traffic or generate more leads, for example, with often just a few simple tweaks. Tweaks could be as minor as the inclusion of a hashtag in your Tweet or as major as a separate landing page.

Let’s say you wanted to improve your click-through rates (CTR) by improving headlines. In a nutshell, you could share two headline versions as Tweets and track which one has the highest CTR. By adding a trackable link to your message, this more extensive tracking in Google Analytics is possible. Analyze your data and draw conclusions.

Tips for A/B testing

  1. Do one test at a time
  2. Test one variable at a time (for example CTA button color, headline, or images)
  3. Choose your social channels
  4. Consider your test timeframe
  5. Select sample groups at random
  6. Add UTM parameters to your links
  7. Always test your variables around the same time of day and week
  8. Only share one variation with each sample group
  9. Test your variation against your control
  10. Measure vanity metrics and deeper, event metrics such as conversion
  11. Record your results

Now that you know what content performs the best, use that as your new control and repeat the whole process. Never let your social strategy go stale.

Learn more about how eHarmony marries social media and ROI in our case study.

Read the Case Study