How Experimentation and Measurement Can Help Win A Little Bet with Facebook Promoted Posts | Beth's Blog

How Experimentation and Measurement Can Help Win A Little Bet with Facebook Promoted Posts


As part of my work at the Packard Foundation,  I have the honor of working with a cluster of grantees working on children’s healthcare coverage on a networked capacity building project called “Friending the Finish Line.”    designed to help grantees institutionalize social media and networked nonprofit skills and strategies and integrate them into their broader strategy for communications on children’s health.  A networked capacity building project is where a cluster of grantees learn and practice new skills and knowledge together, skills and knowledge that are needed to achieve longer-term program  impact.

But it isn’t just about the presenting content, it is about applying the ideas in a series of small experiments or action learning experiments and sharing insights with peers.   This philosophy is very much in tune with what Peter Sims, author of “Little Bets: How Breakthrough Ideas Emerge from Small Discoveries”  calls “Little Bets”     One message in the book is not start with a big idea with it all planned out in advance, but to make a methodical series of little bets about what might be a good direction and learning from lots of little failures and from small wins to find unexpected strategies and tactics that lead to big and extraordinary impact.     This is exactly the thinking that a nonprofit needs to institutionalize to be successful with networked approaches and social media.

Each month, this peer learning project provides an opportunity for an organization to share the results of a “small bet” or experiment and what they learned.   Not all experiments are big wins or perfect, but they become the jumping off for deeper insights and a try it and fix it mindset.

This month the Arkansas Advocates for Children and Families presented the results of their experiment with promoted posts on Facebook for their Medicaid Toolkit Rollout.  The toolkit includes information about how Medicaid works for Arkansas and offered some tools for their advocates to help spread the word.    The content was high quality and based on the insights they have gleaned from their audience, highly relevant.     They created materials that could be shared on Facebook and other social media channels, including some nifty infographics.   All of these materials were available on a special landing page:

Many nonprofits, including the grantees in this group have been feeling that paying  for promoted pays was unfair, but with a goal to get greater reach and engagement.   And, their content posted on Facebook getting an average of 15 likes and 6 shares on FB – they decided to place a small bet.     What if they spent  $50 to promote a post about the toolkit – could they get more advocates to consume and share the toolkit?     The results?   Their reach metric exceeded their most popular post by 11,000,  they got the more meaningful comments and more likes/shares since they started their page.     In addition, the newspaper used their toolkit content to write a feature.

For me, what was most exciting was the group discussion about after reporting on the experiment about to how to design and measure the next experiment with a promoted post.

Key Takeaways:

  • It is important to think about having a small amount of in your budget for promoted posts
  • Use promoted posts to accelerate the engagement of good quality content that you know from previous measurement insights performs well
  • Iterate, Iterate, Iterate – it isn’t about just one experiment, but a series that will lead to the greatest impact

Understanding the Recent Facebook Algorithm Changes and What It Means for Nonprofits

Source: Tweet4Ok Blog

In early October, nonprofit social media managers started to notice a change in Facebook’s Edgerank algorithm that was causing big drops in reach for many branded pages.  The Edgerank algorithm is what determines how posts are displayed on the newsfeed and in what priority order.  (If you want more details about how it works, a very good resource for the technical details is:    Both Edgerank Checker and Post Rocket published good posts on the details, but Facebook has changed the way it displays posts in the newsfeed to prioritize posts that have the greatest number of interactions – a cumulative figure determined by the number of clicks, likes, comments and shares a post receives.  The new algorithm also factors in negative feedback and penalizes posts that have been actively hidden by viewers or reported as spam.

In October, I invited Meghan Keaney Anderson @MeghKeaney at HubSpot to write a guest post on my blog called “How To Use Measurement To Adapt to the Changes in the Facebook Algorithm” she offered some excellent advice about how to use measurement and learning BEFORE investing in promoted posts.

  • Do a content audit
  • Post at the optimal time to drive engagement
  • Understand the behavior and content posted that leads to unlike and hide
  • Balance your content and review performance in insights
  • Keep Calm and Inspire On

I can’t emphasize the importance of doing a content audit which requires looking at your Facebook insights and getting a good understanding of what content performs well or not on Facebook before you begin experimenting with promoted posts.       While you do this, it might also be useful to do a little benchmarking.

What is the average percentage of total fans reached?  This post, based on data from Socialbakers,  gives you the run down on average percentage of total fans reached.   The larger your fan base grows, the harder it is to scale engagement.    So, use this as a guide – on an average month what is the percentage of total fans that you are reaching?   If it is below the average,  go do that content audit.

Understanding Hide and Unlike

John Loomer has a good post explaining why hide and unlike are the silent killers of reach and engagement on your Facebook page.  He also explains how to extract and analyze the Facebook insights data.    Page lever has done some benchmarking of the hide and unlike activity on Facebook and it is described in detail in this FastCompany article.  Here’s the relevant bits:

There are four types of negative feedback on Facebook ranging from undesirable, to worst for a marketer, starting with:
● Hide: hides a single specific post from the user’s newsfeed
● Hide All: hides all the posts by that page from the user’s newsfeed. This used to be known as “unsubscribing.”
● Unlike Page: “unfan” the page
● Report spam: user thinks your page is spam

What they found was rather interesting: 98% of post views generate no negative feedback. But the other 2% of the time a fan responds with some form of negative feedback.

Results from the data:
• 1 out of 50 post views gets a negative response
• Facebook fans are most likely to block ALL your page stories when they take a negative feedback action, 60 times more likely than unfanning your page. Which means that just because your brand has a lot of fans doesn’t mean all those fans are seeing the page content. Some fans may have just hidden the page.
• Fans are more likely to report a post as spam than to unlike the page.

So, benchmark your unlike and hide percentage and if it is higher than 2%,  evaluate the tone of your posts.  Are they spammy?      (And if you are wondering how to calculate this percentage,  this tutorial explains how to do it – it isn’t hard to do in Excel – if you are lacking a data nerd in your organization – read this.)

Promoted Posts:  Best Practices

In November, there was a big backlash against Facebook brewing – the question.   John Loomer has a good post explaining the different views with “Facebook Pay To Play:  Fact or Fiction?”   In the aftermath of Hurricane Sandy, nonprofits were asking, “Why is Facebook asking us to pay to reach volunteers?” (and not offering a grant program of donated promoted posts).    Right before Thanksgiving,  Adage published a piece about a research study with some good news and bad news.  Reach is down, but engagement is up.    The result is that while fewer fans are seeing posts in their news feeds, those who do are more likely to have a real affinity for the brand and engage.   The conclusion was that brands may need to invest ad dollars to maintain the reach among fans they were accustomed to.   Okay, but how to go about that in a way that doesn’t waste your nonprofit’s limited resources?

Set up Experiments! Here’s a guest post from Global Giving that describes how they used experiments and data to under best practices for Facebook Promoted ads.

How To Resources

Social Media Examiner has an excellent guide to using  promoted posts and John Haydon has offered some great tips here.     To summarize:

  • Understand how your content performs in real time and if a post is getting “lift” consider promoting it
  • Always be analyzing, testing, and measuring
  • Get a budget for promoted posts
  • Promote high quality content
  • Promote posts that help get you to overall results
  • Promoted posts don’t replace a robust content and engagement strategy that has a measurement system in place to constantly improve

The metaphor that stuck in my mind as we had this discussion and reviewed the advice is that it is like dieting.  You won’t keep the weight off unless you make a lifestyle change – eat healthier and get exercise.  There are no quick fixes or magic pills.    The same holds true for your communications and marketing strategy — you have make a lifestyle change and incorporate measurement and making sense of the data.

Is your organization placing a small bet on promoted posts?  What did you learn?

12 Responses

  1. […] hard to keep up with all of the changes that Facebook regularly makes, so it’s nice to have a review of how things stand now from Beth Kanter (who […]

  2. […] hard to keep up with all of the changes that Facebook regularly makes, so it’s nice to have a review of how things stand now from Beth Kanter (who […]

  3. Hi Beth–lots of great content here! I like the basic premise of the “little bets” approach (beyond the social media application!). We haven’t been quite as methodical as some of the experiments you describe, but we do assess how our posts are performing on a monthly basis at With our modest (but growing and pretty engaged) fan base, it costs us just $5 or $10 to promote a post. It has certainly seemed worthwhile to do this for our top couple of posts each month, helping reach thousands instead of hundreds. It also has seemed to help grow the overall fan base.

  4. […] Post: How Experimentation and Measurement Can Help Win A Little Bet with Facebook Promoted Posts – As part of my work at the Packard Foundation,  I have the honor of working with a cluster of […]

  5. Beth says:

    Hi David,

    Thanks for your thoughts here! Did those fans or views turn into an action?

  6. Yes, we’ve had some additional volunteers generated through that visibility.

  7. Gabi says:

    The true happiness comes to a house with the first cry of a nebrown baby. Congratulations!!!Hahahaha@#$$$**This Baby is Your Blessing,I wish Happiness in every way,Good Luck God Bless, I sayAnd Many Blessings and Wishes,To Welcome Baby Into Your’s Life .Warm Regards,RamaUdaya.K

  8. […] How Experimentation and Measurement Can Help Win A Little Bet with Facebook Promoted Posts […]

  9. The Ifster says:

    You ask ‘Are they spammy?’

    Very simply, if they are unsolicted, irrelevant to the user and on their newsfeed rather than under the ‘sponsored’ column then yes, they are spammy.

    And every ‘suggested post’, ‘sponsored link’ and ‘page you might like’ will always be reported by me and thousands of others as spam when it is in the newsfeed. Curiously, when the ad/spam is under the ‘Sponsored’ column on the right hand side of the page I never report it – it’s in it’s correct place and not invading mine.

    So, by all means keep promoting companies use suggested posts and sponsored links and we’ll keep reporting them as spam.

  10. […] in particular, cracks me up). Also check out a recent post that is chockfull of information on using Facebook promoted posts (don’t dismiss it without reading […]