Sentiment data is an increasingly common metric for large tech companies. How can it shape the future of charity comms?
As a charity, you’ve probably dipped your toes into the world of data in order to get better results from your communications and fundraising efforts online.
But what if the questions you really want to answer and the information available aren’t as easily quantifiable?
What if you want to more closely understand the emotions of the people interacting with your charity - what they really think of your organisation, its services and the cause you stand for? After all, empathy is a quality that every charity aspires to.
An estimated 80% of the world’s data is unstructured information that doesn’t sit neatly in a spreadsheet or database but is floating around in an unorganised, undefined way. Think social media posts and Twitter discussions, customer service responses, reviews and blog posts.
Sentiment analysis, otherwise known as opinion mining, is a method of extracting and analysing this kind of data in order to make sense of how people are talking about a brand or issue.
Using machine learning techniques such as natural language processing, it’s possible to interpret this kind of data on a large scale to gauge the sentiment or emotion behind what people are saying online - making it measurable in the same way that you might analyse clicks, follows or email opens.
While it’s nowhere near perfect, artificial intelligence is getting much better at reading and reacting to emotional cues. As algorithms become more sophisticated, they’re able to extrapolate emotions much more effectively than in the past, and so a blossoming industry around sentiment data analysis has sprung up.
In the online world of the charity sector, sentiment can change very quickly. It’s important to have your finger on the pulse of what the public and your service users want in order to stay relevant. When generating new support for a charity, we know that personalising the giving journey matters, but most charities feel underequipped to deliver these experiences and communicate with people in a way that’s tied to specific moods.
Using sentiment data can help charities:
Sentiment analysis is designed to detect ’polarity’ (a positive or negative opinion) in a piece of text: whether it’s a whole document, paragraph, sentence or clause. But it can also focus on emotions such as angry, happy or sad or even intentions such as interested or not interested.
Apart from simple polarity of opinion, types of sentiment analysis include:
This provides a much more precise level of polarity than just ’negative’ or ‘positive’ by breaking sentiment down into a scale (such as a 5-star rating system from ’very happy’ to ’very unhappy’). While automatically generated by a computer gathering the data, this is similar to the way a customer service survey might be carried out or someone might fill in a review.
Emotion detection picks up on certain trigger words that express a specific emotion, from happiness to frustration, shock, anger and worry. This can be a lot more complex because a machine learning algorithm needs to be trained to grasp the context of what’s being said, or risk making a wrong extrapolation – not always an easy feat for a computer, but this is where natural language processing gets clever.
Taking it one step further, intent-based analysis recognises the actions or intended actions behind a certain emotion, anticipates the next step they might take and responds automatically or prompts someone to step in and help them.
Aspect-based analysis looks beyond the surface level sentiment to the wider context of the specific thing (product, service, feature, service element, etc.) being positively or negatively mentioned, instead of just assuming they’re talking about an organisation or brand as a whole.
Social listening can be a great way to analyse the mentions you’re recieving online and better understand people’s feelings and emotions, with sentiment analysis as a subset of that.
The beauty of social listening is that there is a huge amount of data to analyse, and social media is the number one place to connect with communities. Knowing what inspires action and engagement across their target audiences can help charities craft messaging that truly makes a difference in the communities they hope to reach.
There are a large number of ’software-as-a-service’ (SaaS) social listening tools and platforms now available to run from the cloud, many built specifically for non-profit organisations including:
Awario – a social listening platform specifically for charities which includes social and news monitoring, volunteer outreach and sentiment analysis, with notifications about what’s going on in your specific niche and influencer marketing assistance.
Keyhole – A social media analytics tool designed specifically for charities, with advanced sentiment analysis that includes brand, campaign and influencer monitoring.
Other SaaS platforms specialising in sentiment analysis for general organisations and businesses include: