DISCUSSION: STORYTELLING ANDSUCCESS STORIES

I’ve not been able to keep up with all threads but it seems there have been a number ofinteresting discussions over the past few days covering various aspects of the role of datavisualisation and what we should expect from it. Thought I’d join the party late and throw in afew thoughts of my own as I was planning on writing something about these subjectsanyway.

By Andy Kirk | April 8, 2013 | Articles

4/3/22, 12:35 AMPage 1 of 10

Storytelling in visualisationFirstly, I would recommend you take a look at Moritz Stefaner’s post (http://well-formed-data.net/archives/868/look-ma-no-story) about the di!erent functions of visualisations -those that tell (or more specifically show) stories and those that don’t. I particularly suggestyou read the comment responses at the bottom of the post, I haven't read them word-for-word but skimming through reveals some good discussions in there. Interestingly, you cansee how often the nuances and semantics of the written word are at the root of manydisagreements about perspectives when they are actually the same views just articulateddi!erently.

I’m not going to get into deep discourse about what I believe a story is and how that relatesto visualisation but I just wanted to share my view on the distinction I personally makebetween the two main types of visualisation function: exploratory and explanatory.

Exploratory visualisations create an interface into a dataset or subject matter. They do notpropose a single narrative, nor actively draw out key insights or headlines. Instead, theyfacilitate the user exploring the data, letting them unearth their own insights: findings theyconsider relevant or interesting. It is a discovery process that could potentially lead to thefinding of many di!erent insights or maybe none at all, depending on the user’s context.Typically, we might consider exploratory visualisations to be interactive in format (andindeed often we are talking about a ‘tool’ of some description) but they are not limited tobeing interactive.

Explanatory visualisations are focused, editorially driven works that aim to surface keyfindings. Whilst they may contain several di!erent dimensions of analysis this doesn’t meanthey are exploratory in the sense of facilitating broad manipulation of the variables beingdisplayed. It is in these types of visualisation that we would most associate the function ofstorytelling with data, often attributed to how they are structured(http://vis.stanford.edu/papers/narrative).

In many cases it is hard to exactly identify or purely distinguish whether a project isexploratory or explanatory in nature. In some cases this is the failing of the design. It mightneither draw out the key findings nor let you explore the subject’s data. In other cases thedi"culty might be because the designer has created an e!ective combination of both(http://projects.nytimes.com/guantanamo).

The way I judge whether a visualisation is exploratory or explanatory is to ask myself ‘whodoes the work to reveal insights?’ (BTW, this is not a unique idea, I’ve heard of others whopropose the same).

If I, as the reader, have to do the work (either visually or interactively) to find insights it ismore exploratory than explanatory. By contrast, if the designer is the one who takes theresponsibility to present the main insights, I see that as more explanatory than exploratory.

Visualising Data 2009 to 2022. All rights reserved | Cookie Statement | Contact Us

(https://www.visualisingdata.com/about/)



4/3/22, 12:35 AMPage 2 of 10

Let’s look at the example of Moritz’s work on the OECD Better Life Index(http://www.oecdbetterlifeindex.org/). This is an exploratory tool to allow people toinvestigate the di!erent quality of life indicators and measures for a range of countriesacross the world. This project is entirely user-driven Is there a single or group of importantfindings? No, there are many possible insights contained within this dataset. These onlybecome insights once a pattern reveals something relevant to a user. With (I recall) 34countries and 11 di!erent quality of life indicators, there are many ways a user can slice,analyse, sort and explore this data.

image_3462OT_01_01

(http://www.oecdbetterlifeindex.org/)

Similarly, the next example (http://infobawards.s3.amazonaws.com/SPOTLIGHT-ON-PROFITABILITY_Krisztina-Szucs.png ) from Kristina Szucs is exploratory in nature. It is a staticproject that provides insight into the combination of film review scores and movieprofitability. Whilst the subject matter is filtered to show a specific view of the data (just the

4/3/22, 12:35 AMPage 3 of 10

top 3 profitable movies, by genre by year) this is still exploratory. You learn how to read thechart using the legend at the top of the display then apply that lens across the piece todiscover the films, genres, and shapes that most interest you. The design does not include aspecific headline discovery or conclusion, it simply acts as an interface for you to find whatstrikes you as an insight. Once again, you as the reader do the work to find what isinteresting to you.

Movie

(http://infobawards.s3.amazonaws.com/SPOTLIGHT-ON-PROFITABILITY_Krisztina-Szucs.png )

By contrast, this piece from Simon Scarr is contains a main narrative, the striking reverse barchart shows the fatal impact of the Iraq conflict. Accompanied by a few supplementarycharts for additional context, this is a focused story principally about one dimension of thesubject matter. The size of the chart, the colour scheme and the overall architecture arerepresentative of the designer taking responsibility to highlight a story for you to consume.

4/3/22, 12:35 AMPage 4 of 10

Iraq

(http://graphics-

info.blogspot.hk/2012/09/malofiej-20-look-at-our-participation.html)

Success in visualisationThe other issue that has been aired in recent days – but also discussed for a long time ingeneral – is the question of where are the visualisation success stories? It is something weposed in the Data Stories podcast 16 (http://datastori.es/episode-16-what-was-big-in-2012-and-what-is-coming-in-2013/) and I know it will be discussed in podcast 21.

There is always a strong appetite to find examples of where visualisations have made adi!erence: opinion changes, behaviour changes, money savings, great discoveries andmaybe even life saving.

It is natural for us all to seek a marquee and tangible example that indisputably providespopular evidence of the value of this subject. That’s perhaps why so many of us cling on tostory of the John Snow ‘Cholera Map (http://www.amazon.com/Ghost-Map-Londons-Terrifying-Epidemic/dp/1594482691)’. Regardless of the true nature of the impact of his dotplot map, we can use this to fit in with our agenda of wanting to show how visual portrayal isso important at influencing decisions and situations.

However, I’m increasingly aware of how ‘success’ of visualisation is rarely tangible nor grandin nature. And it doesn’t need to be. It is ultimately a form of communication. We don’t seemto put the same weight of expectation or desire for outcome measures on many writtenforms of communication, do we?

The main issue is what does success look like? Yes, that awful management phrase thatinvokes a chill down the spine. Well, naturally and boringly, we have to think first aboutintent and context: What are you trying to achieve and what are the expectations in terms of

4/3/22, 12:35 AMPage 5 of 10

outcome?

Using the first of two horse-related idioms or phrases, it is all about horses for courses(http://en.wiktionary.org/wiki/horses_for_courses).

Building on the discussion above, let's think about the distinction between success fromexploratory visualisations and explanatory ones.

It would seem to me that the ability to attribute ‘success’ is more tangible with exploratoryexperiences, either as an exercise in visual analysis (where it might be you as the audience)or as an interactive experience for others.

Suppose a particular combination of variables, filters or selections reveals an insight. That isa discovery. If this discovery was achieved as a result of the visual portrayal and somethingthat we may not have seen visible with statistics or data alone, then that success has to beattributed to visualisation.

Whether this changes the course of society is another matter. Someone might haveunearthed the discovery that a vaccine can cure a certain disease or it might be that youspent more money last year in Nandos (http://www.nandos.co.uk/) than you’d realised. Thecontexts are very di!erent and similarly the impact is likely to be very di!erent (unlessNandos was the cure…) but success has occurred: something interesting or relevant hasbeen found in the data and the expectation was fulfilled. 1-0 to visualisation.

Now the chances are that the Nando’s situation is a good deal more likely to be happeningday in day out than the Hollywood-esque discovery of the vaccine’s positive e!ect. However,the Nando’s example will be happening in very personal and private settings. We don’tnecessarily have widespread access to these stories because they are ultimately quiteminor and we don't really have them on our radar because we want to aim higher. But theyhappen to all of us almost subconsciously everyday and they are important.

On the other side of the equation, the explanatory visualisations, what should we expectfrom them?

I’ve recently been thinking about the possible parallels between visualisation and marketingcommunication theory. In particular, the concept of AIDA(http://en.wikipedia.org/wiki/AIDA_(marketing)), which describes the incremental events ore!ects of advertising.

4/3/22, 12:35 AMPage 6 of 10

AIDA

(http://ptclientsnow.com/blog/wp-

content/uploads/2012/05/aida-1200.jpg)

I’m no expert in Marketing Communications – as will be seen clearly over the next fewparagraphs – but having learnt about it a few years ago it has always struck me aspotentially transferable (with a few tweaks), perhaps into a AIIA model of Attention,Informing, Influencing and Acting:

Some adverts are intended to simply achieve 'Attention' or 'Awareness', the first ‘A’ in AIDA.Attracting the attention of the customer and making them aware of the existence of thatproduct/service is the limit of the ambition so achieving that is success. That’s why youmight see, for example, so many god-awful annoying adverts that exist to stand out fromcrowd. They might not convince you to buy anything but that’s not the aim at this stage.Likewise with visualisation, some projects are probably achieving success (in their minds) ifyou have achieved such attention, somebody has simply looked or engaged with them.Quite a low ambition, of course, but this might be exemplified by the setting of visualisationas a prop or artefact, maybe in a magazine to focus your eyes on that article.

The ‘I’ part of the AIDA acronym is about raising 'Interest' by demonstrating the benefits andadvantages of a product or service. Maybe the equivalent in visualisation is about informingsomebody, by providing a visualisation that addresses a gap in understanding about asubject.

'D' is for 'Desire', to convince the potential customer why they should need this product orservice. Perhaps in a visualisation context the equivalent would be something more aroundinfluencing, going beyond simply informing and starting to have an impact on somebody’s

4/3/22, 12:35 AMPage 7 of 10

belief system or choices they make. Maybe an example of this was the Drones Strike(http://drones.pitchinteractive.com/) or US Gun Deaths (http://guns.periscopic.com/)projects? Neither of these works are likely to have the ambition of stopping drone strikesovernight. But the more people who digest them and who are a!ected by the humanmessages contained, it is not hard to imagine how a chain of events layers up with otherinformation could ultimately lead to protest or opposition action to government policy?(Incidentally, not the attempt at evaluating success at the end of this post(http://www.independent.co.uk/voices/iv-drip/shocking-visualisation-of-drone-strikes-in-pakistan-8550548.html)).

The final ‘A’ is for ‘Action’, to take the customer towards the act of purchasing. Forvisualisation this idea of action could be consistent with the ultimate success measure: youhave su"ciently influenced somebody about a given subject matter that you have triggeredan action, a decision, a change. Is it the impact of Al Gore’s(http://www.youtube.com/watch?v=9tkDK2mZlOo) rise in his cherry picker to follow theprojected temperature forecasts? Or was that less about the visualisation and more aboutthe drama of the delivery? Who knows, but this is the holy grail, the type of ‘success stories’we’re all keen to find examples of.

This model is not a perfect fit, nor do I propose that these equivalent levels of successmeasurement are the best way of judging impact, but I feel it is worth the discussion abouthow other communication-related disciplines define tiers of success.

Perhaps we just need to wake up more to when they happen and notice them. This morningI have been looking on Kayak (http://www.kayak.co.uk/) at flights prices for a forthcomingtrip. I never even thought about the display that was influencing my decisions but I’verevisited it just now. Decision-support devices like this are everywhere but we take them forgranted. A visualisation success story is that I am now in position to buy this flight withconfidence and know that I’ve got the best price for the date on which I’m leaving.

Kayak

(http://www.kayak.co.uk/)

Final ReflectionsFinally, we need to acknowledge the things that have most impact. People.

People are irrational. They are inconsistent, moody, prejudiced, have tastes, have pressures,

4/3/22, 12:35 AMPage 8 of 10

are influenced by political factors. In my training materials I include the image below, relatedto my second horse-related phrase for the day: You can lead a horse to water, buy you can'tmake it drink (http://www.phrases.org.uk/meanings/you-can-lead-a-horse-to-water.html).

Horse

(http://www.redbubble.com/people/rebelpony/works/1073982-you-can-lead-a-horse-to-water)

You can only do your best to put forward a visualisation that most e!ectively serves theneeds of the subject matter, the context and the audience you’re reaching. Beyond that, theultimate success is out of your hands.

Regardless of the information Kayak provided me with I might have booked an expensiveflight in the mid-afternoon just because it suited me better, maybe I am superstitious andlike to leave at 12 noon precisely. An irrational choice given the information I had to hand. If Idon’t follow the implied advice the data visualisation presents to me, it's not that it failed it isjust because people are and can be be odd but they are such an important part of themeasurement of success.

I look forward to any discussions, comments in the boxes below…

4/3/22, 12:35 AMPage 9 of 10

(https://www.eventbrite.co.uk/e/masterclass-in-data-visualisation-classroom-training-with-andy-kirk-registration-199664230087)

SHOW COMMENTS+

4/3/22, 12:35 AMPage 10 of 10