How Compassion International built a social listening function from scratch

Despite the growth of social listening technology that claims to make accessing insight from social data easier, it is still difficult for organisations to develop global social intelligence from scratch. It takes planning, skills and an understanding of what social data can actually offer in order to extract actionable insight.

This is what Noel Schieler, Market Intelligence Specialist at Compassion International quickly learned. Her experience at The SI Lab’s Social Intelligence Growth Certification helped her to understand what to consider when setting up a multinational social intelligence function to have the greatest chance of success in supporting the organisation’s programmatic and fundraising functions.

About Compassion International

Compassion International is a multinational non-profit organisation that works across 40+ countries (either in a fundraising or programmatic capacity) and runs programmes to release children from poverty across 27 countries in the majority world. Their social data analytics function sits within the global insights team. Like many organisations, Compassion International kicked off its social intelligence journey by procuring a new social listening platform. Noel was then hired to develop and manage processes around it, with an aim towards applying the discipline as an on-going source of insight into global strengths, weaknesses, opportunities and threats.

However, Noel was also new to social data analytics. And, despite many technology providers presenting their tools as user-friendly to anyone - even those without experience in data analytics - she quickly discovered this wasn’t the case, and found it hard to achieve real return on Compassion International’s investment. As a result, she signed up to The SI Lab’s Social Intelligence Growth Certification, to learn more about how to use social data for her work.

“I realised from day one of the course that social intelligence isn’t just about mastering the use of a social listening tool. It’s a broader discipline that has so much opportunity for us as an organisation.”

She learnt that her job doesn’t end with producing dashboards for other teams. Instead, there’s real insight that can be gathered if done the right way.

Key learnings

After taking part in the course, Noel took away some key learnings that she has since applied to her work at Compassion International.

‍

1. Challenge your list of use cases. Social listening can be used for more than just brand monitoring.

Initially, like many organisations, Noel’s work focused on using social listening to track brand mentions. “Brand tracking was high priority, and still is, but we were focused on how social intelligence could diversify our mix of data sources for voice of the customer (VOC) collection. However, I’ve become increasingly passionate about applying social intelligence as a key mechanism for looking beyond direct brand mentions and known audiences.”

Since realising this, Noel has found that social data can provide valuable insights that helps them to achieve their core mission of releasing children from poverty. This includes analysis of adjacent non-profits, charitable causes and intervention types, as well as identifying partnership opportunities that elevate joint impact and gaps in the market to promote their cause.

‍

Beyond marketing, it’s important for Compassion International to understand the environment surrounding vulnerable communities at any given time in order to channel aid and resources effectively when needed. And this is where social listening comes into its own. “Due to the real-time nature of media and social monitoring technologies, I believe social intelligence could have increasingly direct programmatic or “field” impacts as the capability expands use cases and internal reach.”

Dashboard: Child labour. Image provided by Compassion International. Source: NewsWhip[/caption]    

‍

2. Plan your tech stack carefully

Noel inherited a social listening technology that could manage narrow use cases, but she quickly realised that it didn’t fit with Compassion International’s organisational structure or vision for democratised insight. As a team of one, she wasn’t able to benefit from the user-based pricing model, and the vendor wasn’t able to switch to volume or query-based pricing. At a certain point, the add-on costs associated with scalability were forecast to significantly outweigh the benefits of switching tools. So, Noel decided to seek a new technology provider.

But, before engaging with any vendors, she wanted to understand exactly what capabilities they required from a social listening tool. “We had an incomplete understanding of the needs of social intelligence end users and a lack of pre-definition surrounding which insight gaps social intelligence can complement or fill and which processes social intelligence might make more efficient.”

To address this, Noel undertook a series of cross-functional interviews and ideation sessions to understand the different use cases - current and future - that the wider organisation envisioned being supported by social data.

“The process shone a light on several use cases we hadn’t thought of previously and pin-pointed unique opportunities for future scalability across a wide variety of functional areas.”

Throughout this process, Noel documented all the findings she gathered from the different teams. This not only clarified how social intelligence could help solve the organisations’ challenges, but it also highlighted the value that social data can bring. “This information was invaluable to have on file as we were able to evidence clear alignment between stakeholder needs and the features and capabilities of our frontrunners [technology providers]. And, by grounding ourselves in what could realistically be built and activated in the short term—while still positioning ourselves for growth— we achieved a 40% annual cost savings on our SI tech stack.”

‍

3. Understanding nuance isn’t always something a tool can help with

Whilst social listening technology can provide invaluable support in areas such as gathering and making accessible large volumes of data from multiple sources, it does have limitations. Noel quickly learnt that manual analysis is a crucial part of extracting insight from social data. And this is particularly important in the non-profit sector.

For example, when it comes to sentiment analysis, technology can only provide semi-accurate sentiment classification, and this issue is often amplified for non-profits. “At Compassion International, we work with children living in extreme poverty. A piece of supporter user-generated content (UGC) may reference both a devastating circumstance - such as a natural disaster - as well as an expression of gratitude for the opportunity to intervene and positively impact children in need. This would likely receive a mixed sentiment tag and skew measures of brand favourability.”

Dashboard: Ebola. Image provided by Compassion International. Source: NewsWhip      

Another area that requires special attention is non-English data. Compassion International’s social listening function sits within a global-facing team, and can be called on to provide insight across the 40+ markets they operate in. As well as this being practically difficult to do alone, she learnt that it’s hard for her to extract relevant insights for each market. To be able to translate local languages (particularly slang), it requires local analysts with understanding of the cultural nuances of the market. Very few social listening tools are able to do this automatically.

Setting the groundwork for the future

Now that Compassion International has a better understanding of the value social listening can bring to different parts of the business, and what is required to extract this value, they can build a social listening function that works.  

As they start building relationships with their new technology providers, Noel has identified a couple of focus areas. “Firstly, migrating existing monitoring assets into our new platforms and, secondly, establishing additional, multi-language analysis topics.” In particular, she’ll be looking at producing global reports, circulating case studies, providing localised training, and facilitating the addition of more complex use cases over time. “Generally speaking, we plan to take more of a “centre of excellence”, hub-and-spoke approach as the capability expands—identifying local advocates & enabling as much regional ownership as possible.”

‍

Main takeaways

When developing a social listening function within your organisation for the first time, it’s important to consider the following:

• It takes time to set up a social data analytics function properly, but it will bring benefits in the long run as you will get actionable insight more quickly.
• Planning is essential. You need a clear idea of what insights you’re looking for, and the data sources that can generate this, before beginning a project.
• Understanding cultural differences is important. For example, someone in the US cannot fully understand the nuances of another culture. To get true and relevant insight, you need analysts who do.
• Scaling a social listening function across a global organisation is challenging, so it’s important to develop a structure that works for the needs of your business.
• Join the social intelligence community and benefit from the group wisdom of those who have already gone through what you’re going through!

“In my experience, something unique and special about the social intelligence community is professionals’ genuine desire to connect with and help one another. Especially because building a social intelligence function is so challenging, and there’s a lot of commonality across the barriers and roadblocks we face.”