Sequential Visual Storytelling and the Carebot

NPR have been working on an interesting style of storytelling which they call ‘sequential visual stories’. The technique combines audio, photography, text, video and illustrations into a compelling ‘slideshow-like format’.

Wes Lindamood, senior interaction designer for NPR has written in detail about how and why the format evolved;

As an experiment, we tried to think of how to take the audio, photography and text (for this story) and not treat them as separate content types, but instead integrate them into a storytelling approach that felt native to the web. For example, we wondered if we could combine audio, photography and text in the browser to achieve the immersive feel we wanted without sacrificing the ability to disaggregate the content and share or remix it in different ways. Also, knowing that users have an expectation of control over online experiences, could we introduce the right level of interactivity into the story without disrupting the momentum of a compelling narrative?

Life After Death’ was published in February this year, and covers the impact of Ebola on one village in Liberia. The story combines photography with minimal text and a background of ambient audio, punctuated by the voices of villagers telling their own stories. The combined effect of the concentrated visuals and personal spoken word, produces a tremendously intimate viewing experience.

Surrounded by the sounds of the village, a survivor tells how the local custom of washing the dead before burial, led to the outbreak of Ebola following the first victim’s death. Later, a farmer talks of how the grief her husband feels at the sudden death of both his parents has impacted on their capacity to grow food and feed their family. A young man describes how he contracted the disease and recovered, but cannot return to his studies in Monrovia as he is now the sole carer of his six brothers and sisters.

The goal of including audio quotes in this story is to minimize the distance between the user and people of Liberia whose story we are telling. We wanted users to hear the emotion in voices of the villagers of Barkedu. If this were a story told purely in audio, we would have needed to hear the voice of an interpreter in lieu of the subject. By integrating the audio quotes into a text narrative, we not only give control to the user over the pacing of the story, but we provide users with the ability to hear this emotional story in the voice of the subject.

In an NPR blog post published this week, entitled ‘Do Visual Stories Make People Care?’, Tyler Fisher of NPR explains that making people care is the ultimate mission of their visuals team. As well as working on ways to tell a story in a compelling and engaging way, they are also building a tool which they hope will be able to measure caring.

Carebot, is being funded by the Knight Foundation and has yet to be built. Fisher’s blog represents some thinking out loud about how the tool might work and analyses the click and completion rates of the sequential storytelling articles to determine how successful different aspects have been.

The research provides some interesting feedback on what encourages (or discourages) a reader to begin a story;

For the stories where begin performance fell flat, we can point to a clear reason: “Put on your headphones” prompts or similar notices that audio will be a part of the experience. Of all users who saw a titlecard without an audio notice, 74.4% of them clicked to the next slide. If an audio notice was on the slide, only 59.8% of users faced with that titlecard moved forward.

However, when it comes to engaging the user, the audio element performs well;

Our average engaged user completion rate across stories was 50.9%. But the data gets more interesting when we start dividing by story subtypes — particularly the divide between stories that integrate audio and those that do not. In that divide, the average engaged user completion rate for stories with audio is 54.5%, compared to 48.5% without.

The findings so far seem to show that measuring caring in numbers is very far from straightforward, and is perhaps too open to subjectivity to ever be an exact science. At face value, the metrics seem to point to success for the sequential format, and Fisher concludes that is down to a combination of a strong story and powerful visuals;

Ultimately, making people care is about the quality of the story itself, not about the format in which we tell it. But I think that, with stories where text plays a large role, we are capable of making people read stories longer than they normally would because of how sequential visual storytelling allows us to pace the story. Of course, this is not an argument for telling all stories in the sequential visual story format. Sequential visual stories work when the visuals are strong enough for the treatment. Not all of our stories have worked. But when they do, we can tell important stories in a way that pulls people through to the end.