How to measure emotional response to content marketing
It has been said that journalism has had a huge effect on content marketing. But as things move forward, it seems just as likely that content marketing will have a huge effect on journalism.
An area where the two differ is in terms of ROI. Content marketing needs to do a lot more than bring in adverts and subscribers.
Content marketing ROI is key for marketers. Just like other forms of advertising, companies need to understand where their budgets are being spent and how effective they are. And in this quest for ROI there is an increasing desire to measure the emotional impact of content.
As part of the Content Marketing Association’s recent Effectiveness report it was found that more than 70 per cent of respondents felt it was important to gauge emotional response to content as part of measuring ROI.
The question is how? At the end of the day, your reader will either do what you want them to – that is purchase or sign up – or they won’t.
But that seems to be changing. Here are some initiatives which will help tell you what the viewer or reader is thinking.
Facial recognition
There are a number of application programming interfaces that allow for the recognition of emotion through facial expression and eye tracking.
Examples include Affectiva, which to date has analysed around 3.5m faces and can be integrated into another program with separate analytics to track engagement. This is a test demo. It scans the face for response in terms of dislike, surprise and other expressions. It can give some intriguing results such as the apparent peak of interest at what appears to be a plug socket (see below).

Analysis from viewing of a Doritos advert
Emotional text
Another area of interest to companies looking to understand or rate copy that has been supplied from an agency, for example, is text-to-motion software. It could also be used to analyse comments or input from social media.
One example of this is IBM’s Watson supercomputer. It uses an application called Tone Analyser to detect emotional nuances such as anger and other social indicators from plain text. It can be applied to email, speech, customer service chat or just text.
It is available and open source so can be integrated into another application. Go here for a demo.
Voice detection
There are a number of speech-recognition systems on the market. Examples include EMOSpeech, which is used in call centres, and Vokaturi, whose makers say can understand emotion in a voice in the same way as a human.
Conclusion
While these technological advances are worthwhile, possibly the biggest challenge is actually obtaining reference input from people in the first place.
There are bound to be questions about how reasonable it is to carry out such a large volume of the necessary tracking. There is already a trend for blocking off microphones and built-in cameras on desktop and laptop machines. Possibly the most notorious example of this was Facebook’s chief executive Mark Zuckerberg, who was shown to have plastered electrical tape over his laptop camera in a recent PR shot.
At the same time, privacy concerns have never really deterred marketers. And ultimately, do the majority of us really care? Understanding how people appreciate content could rub off on journalism, changing the way editorial is produced.
In the immediate future, expect to see ongoing efforts to check for the required emotional response. Be it a clothes brand wanting to initiate joy or a pharmaceutical company wanting to provoke concern, the quest to quantify emotional response could change the way we view writing.
How to measure emotional response to content marketing is part of Content24, the blog for London content marketing agency FirstWord.