A study from 2012 found nonprofits do a better job of ‘engaging’ with supporters over Twitter than with their websites.

That’s probably a bit like saying there is more engagement in a phone call than a press release.  After all, it’s written right there on the tin, Social Media.

Nevertheless, one could always treat it as just another one-way broadcast channel.   So, what does nonprofit tweeting look like?

A follow on study analyzed the tweest of a subsection of the “Philanthropy 400” but purposefully selecting a cross section of sectors and pulling from the top of the top and the bottom of the top 4oo list.  Example charities from the top of the top include The Nature Conservancy, American Cancer Society and Goodwill Industries City Harvest from the bottom.

An entire month of tweets was analyzed with a high of 298 (Audubon Society) and a low of 12 (Goodwill).  The content analysis was done by machine algorithms and multiple, human coders.   The tweets were tagged and coded in several ways.  This coding scheme, in and of itself, has value to all those social media managers out there:

  • Functional Purpose:
    • Informational
    • Community building
    • Action (e.g. asking people to do something)


  • Framing:  positive, negative, neutral.
    • Examples include “We’re honoring & thanking heroes…(positive)” and “forced to close food pantry…(negative)”


  • Type of communication: 
    • one-way: broadcasting facts or information
    • two-way asymmetrical: when the tweet asks followers to do something (e.g. click a link, share)
    • two-way symmetrical: retweeting or when the organization responded to a follower’s tweet


  • Goal attainment: Did the tweet include information about attaining a goal such as level of support or number of votes or donations received?  (You can imagine if this sampling included Giving Tuesday or end of year – it didn’t – that this coding would become redundant, all “1”s for presence of it as the faux matching gift bonanza hit full tilt)


  • Persuasion Type: In predictable academic fashion, the study authors drone on about Aristotelian rhetoric and using ethos, pathos or logos, which we’ve stripped away for your reading pleasure.

Suffice to say, messages can use emotion or facts but also, change the voice/speaker to establish credibility, garner attention or create connection.

    • Voice:  Did the message include a celebrity reference, a political authority or reference to a person like me (e.g. citing survivors, members or victims)?
    • Emotion: this includes motivational messages or those using humor, guilt, love etc.
    • Facts: e.g. facts or statistics or survey results

We promise not to turn the rest of the post into the equivalent of a 6th grade book report and rattle off descriptive statements for what you can plainly see for yourself.  Instead, one comment and one additional study finding.

  • It’s good to see the infrequent use of negative framing.  Study after study show that optimistic wins for long-term growth.  This doesn’t mean you can’t start with a bad situation but don’t leave the story there.  It only induces guilt or other negative emotions that don’t foster the kind of motivation that keeps on giving.


  • There was a noticeable difference in tweet content between the top of the Top 400 and the bottom.  It would seem the bottom groups either don’t have a full-time social media manager or don’t have a good one because their tweets had far less two-way symmetrical activity.  The top of the Top were 3x more likely to use two-way symmetrical.