Crunching Numbers for the Arts

The SumAll Foundation is partnering with Cultural Data Project to assess how social media engagement impacts the economic health of U.S.-based arts and cultural organizations.    

Screen Shot 2014-02-07 at 1.16.06 PM.png

Measuring the “return on investment” of social media is something we specialize in at the SumAll. Our for-profit arm, SumAll.com, is an analytics company that guides businesses as they calculate the impact of their social media/digital marketing on the bottom lines. But for-profit businesses aren't the only organizations who have to measure their bottom line--non-profits often assess their own bottom lines not just in terms economic health, but also social impact. At the SumAll Foundation, we’re partnering with the Cultural Data Project (CDP), a national organization whose mission is to strengthen non-profit arts and cultural organizations through data analysis, to investigate the relationship between an arts organization’s social media engagement and its economic health. For the first stage of this project we are (1) quantifying the effectiveness of social media in expanding an organization’s reach, and, (2) providing practical guidelines on best social media practices for organizations in the arts. 

METHODOLOGY AND DATA

Our mission at the SumAll Foundation is to give non-profits access to valuable technological resources usually reserved for organizations in the private sector. Our project with the CDP relies, thus far, on two specific methodologies borrowed from the business world: benchmarking and decision tree modeling. The first method, benchmarking, compares one organization’s performance with that of other organizations. Using it can help to illuminate both (1) “best practices,” and, (2) standards to measure an organization’s performance in comparison to other organizations. A decision tree model, our second method for this project, is a visual representation of an algorithm that helps organizations assess all potential outcomes of a decision. We are using both of these methods for this project. 

benchmarking.jpg

Thanks to the CDP, data scientists at the SumAll Foundation have reliable longitudinal data to track and assess social media engagement and organizational economic health. This data includes, but is not limited to, information about the type of organization, attendance levels, revenue and expense line items, website traffic, social media network engagement, and workforce/employment practices.

Data often comes to life through visualizations. Our team has used data visualizations to help arts organizations compare their social media activity against that of peers. The visualization below, which shows our benchmarking tool, allows organizations to assess their social media engagement in relation to others in the same region, with similar audiences, and of the same size. We think that having such competitive measure enables institutions to gauge their social media presence more accurately. The interactive visualization below illustrates the average weekly Twitter activity of small-sized arts organizations in New York area (*this is the benchmark) versus an anonymous small New York theater (*bold line). The benchmarking data comes from over 800 arts organizations across the U.S. 

VISUALIZING SOCIAL MEDIA ENGAGEMENT

When assessing data-driven insights, seeing is, often, believing. This is why our team relies on visualizations as a tool to make data accessible and engaging to a wide range of audiences. With the CDP, our data scientists using data visualizations to measure an organization's social media activity against that of their peers. The visualization below, which illustrates our benchmarking tool, allows organizations to assess their social media engagement in relation to others in the same region, with similar audiences, and of the same size. We think that having such competitive measure enables institutions to gauge their social media presence more accurately. The interactive visualization below illustrates the average weekly Twitter activity of small-sized arts organizations in New York area (*this is the benchmark) versus an anonymous small New York theater (*bold line). 

We all know that comparing the MoMA to a small gallery in Brooklyn won’t yield the most useful results. So our data scientists used a decision tree model to determine the best way to arrange and compare a national set of cultural organizations; they analyzed and grouped cultural organizations based on region, organization type, number of donors, revenue, and attendance. Based on the decision tree model analysis, our data scientists determined that the best way to group a diverse set of organizations is by their attendance numbers. Then, then used these subgroups to analyze Twitter and other social media activity. The results are shown below. 

decision tree - mes edits.png

Given the popularity of large institutions like the the Museum of Modern Art or even smaller organizations like St. Ann’s Warehouse on Twitter, it may seem surprising that the average number of Twitter followers for an organization with the highest attendance numbers--over 140,000 annually--is only 3,751. But for every organization with a high numbers of Twitter followers, there are other organizations with small number of followers or no Twitter account at all. And our data scientists have removed outliers with huge Twitter audiences, like Metropolitan Museum of Art. For example, one of the NYC groups reporting massive attendance numbers, the Lesbian and Gay Big Apple Corps with 1,701,500 annual attendees, has a tiny Twitter account with only 83 followers. This compares with the relatively small Wooster Group, with 135,500 annual attendees, who also happens to have have a relatively active Twitter account boasting nearly 8,000 followers. 

FROM ENGAGEMENT TO ECONOMICS

Our work doesn’t stop here--we have a long way to go until our project is complete. Specifically, we will try to answer question whether social media presence has an impact on outcomes such as audience numbers, donations, web site visits etc and which type of activity and platform should an organization focus its efforts on. We also plan to explore how user bases of specific social  networks impact an organization’s social media engagement. For example, are young audience members engaging with their favorite organizations via Instagram--and is that engagement different than engagement with older audiences on Facebook or Pinterest? Further, how are Black American cultural organizations leveraging the power of Twitter--which African Americans use at a much higher rate than any other racial group in the U.S.? The answers, in turn, will guide us in offering the most important activity benchmarks to arts organizations.