Giorgia Lupi – Information Designer

August 25, 2019 0 By Stanley Isaacs


(applause) – Thank you Alan, I’m very excited to be here, and I had to find myself
a designer, actually, but I also like the
description that Alan gave me. I hope that what I do is not
too specific, I am a designer, I work with visuals but I work with data, I don’t know how much
of data visualization you are familiar with,
but I’ll try to be very, I don’t know, simple in what I say. Just a little bit of information
of me before I start it. I am an architect as a background
so I didn’t study design, but I work as an information designer, so building information and
knowledge rather than houses, and in my day job, I work as a professional
in the data world, I design charts, data
visualization and interfaces, that in a way, let people
access data in a visual way. I run a design firm
that is called Accurat. I cofounded the firm five years ago, we are around 20, 22, now,
data-visualization designers, software developers and data analysts. We started out in Milan,
Italy, I’m Italian. Then two years ago I moved here to start the New York office. And as you can imagine, we work across very different industries, creating different kind of data, data-driven digital products for business and communication purposes. This is just a pretty fast composition of some of our latest
digital applications. So we do build a lot of visual analytics, which is a boring definition but, we work with banks and organization, and design and develop digital tools to leverage the data assets. Recently we also started to
work around data strategy, from a design perspective so, supporting the creation of data workflow before even visualizing the data. We also work with startups, building their data-driven product, from the very concept to the development. And we also design multimedia
data-driven experiences, so combining different media
such as video interviews, and videos and data. But today I just don’t
want to show you our work or our process, I’d rather like to start
a conversation with you around what I believe are
some of the key challenges for us as designers, especially
communication designers, and information designers, nowaday, and what I believe could
be interesting perspective to look at our data, and the data-visualization where it from. First of all I think we
should embrace complexity, and think at our data projects in general, as engaging ways to convey the richness of the data we analyze,
the richness of the context of those data rather than simplification. Because the world is
complex, it’s compound, more and more reach of information, that can be combined in
these very endless ways, and more and more given
the increase of complexity of the phenenomenon that we analyzed, like simple bar charts are not really able to communicate and render the complexity. I do believe that one of
the important challenge for data visualization
designers, nowadays, is to experiment and find proper ways to express the data complexity. Most of the times all of that runs around designing
layers of data explorations and so layering both our
analysis and the visualization. So, “overview first, zoom and filter, “and details on demand,”
says Ben Schneiderman, but how can we do that? So I’m gonna be showing you today now, a not very recent project of ours but I still think it’s
useful to be showcased, because it was at Accurat, our playground to find our way to address complexity. It’s an old-fashioned series
of printed data visualization in collaboration with La Lettura, which is the Sunday cultural supplement of the main Italian newspaper,
which is Corriere della Sera. It was a journalistic
collaboration where we worked side by side with the
news room of the newspaper for more than two years
and we produced more than 40 data analyses and visualization, and we sort of created
our method to create rich and dense visual
narratives with data, and in static pieces like that, is of course even more challenging because you can’t have interactivity like in a digital application and so you really have to try
to build these languages to address this complexity. La Lettura is how the
supplement is called, and it can be translated
like the very act of reading, like spending time reading, and the purpose of the
column where we published the visualizations is explicitly
to explore what can be done with data journalism
and data visualization. Also to perform a sort of stress test. How much of a complexity
a reader can absorb. And every time here,
this is still in Italian, but then I’ll show you some
translations in English. Every time here we chose our topic, sometimes it was just a
fascination that we have, sometime it was a compelling hot topic that we wanted to explore. We found our own data sets to visualize, and then we also looked for multiple tangential data sets that
can enrich and contextualize the main one that we found. Then we every time imagined and create a very unique and tailored visual models for the data that we were analyzing. We call what we do here a
multi-layered story telling, with a first story that
should be visually clear at a glance, but then
that can lead readers to get lost in details in
marginal possible exploration. Here it was when in our company,
we didn’t have developers, and so everything that
you see here has been done pretty manually from Excel
to Adobe Illustrator, so no coding here, like
a laborious process of working with the data. I’ll just explain you
one of the visualization for the first idea around complexity. This illustration explores
the Nobel prizes and laureates from 1901 to 2012, and the main data set that we found, and have been working with, contained only the names of the laureates, the categories of the prize, and the years they received the prize. But you will then see
how we really enriched the visualization and the main data set. Because we wanted to look for a context to our journalistic story, and we tried to layer the complexity to make it visually accessible. We visualized the prize categories along this timeline, with
this parallel square, with these main rows highlighted by colors and so we will always
have red for Chemistry, blue for Economic Science, green for Physics, yellow for Literature, purple for Medicine and orange for Peace. Each of the dot that you see here represents a Nobel Laureate, and each guy is positioned according to the year the prize was awarded, in this horizontal axis, but also his or her age
at the time of the award. Which we easily calculated
given the date of birth, so highlighting how old you should be to win a Nobel Prize. Then per each category
we drew a line indicating the average age for winning
the prize per categories, and also the average age
for the total of laureates, so starting to highlight
differences between the different categories and the differences
in ages through time. Then on the other part
we used the main access of the visualization to provide
some aggregated information. The arcs that you see here represent the principal university of affiliation of Nobel per each category, and so we took the total of the laureate, we picked the seven most
common universities, and then we went to see
how many of our geniuses attended those universities per category. And it’s interesting to see
that Chemistry which is red, Physics which is green and
Medicine which is purple, are quite spread among the
seven main universities, while Literature and
Peace’s Nobel Laureates, which are yellow and orange, quite didn’t attend the top university. Then this is even more interesting, those bar charts here,
within these brackets, represent the average
grade level per category, and so how many of the category have a PhD or how many of these wasn’t even degreed, and you can learn that if
you want a Nobel in Medicine, or Economics or Physics, you should really think about a PhD, but you can try without it
for Literature and Peace. The double, rounded dot
represent the women, out of the total, which
are actually not so many, and our design choice used
here was pretty clear, we wanted to reduce the clutter, visually, so we could have also
added the double circles in blue for example around the men, and painted them blue,
but we didn’t because you got the distinction
even if we only highlight the women out of the total. Just going back a little bit, on the whole piece before
exploring the other parts, so you can recall where we are. And you have to imagine
that these pieces were printed like that big and so really people could physically dig into the data. Imagine a double page
of the New York Times is as high as. Then the part below represent
the most frequent hometowns out of the total color per categories, and aggregated per 30 years, so again, using the already built
access of the visualization to provide this aggregated information. And yeah, interesting is that lately, if you’re born in the US, you’re more likely to catch the prize, while at the beginning of last century, that most of the scene was European. And so I guess you’re
starting to see how many contextual information we
added to the first data set, to suggest possible interpretations. We also had some curiosity highlighted, so some prime ones in
particularly relevant stories, like for example, Marie Curie
winning two Nobel Prizes, or the oldest winner that was 90, while the youngest was
24, personal stories, that are then recalled
in the visualization with the number, to show that these are human beings, behind the dots. And of course, this is
the how-to-read-it part, when we use again, the
visualization main shape, to give readers information
on how to play around with our designs and this
is very important to build a very consistent legend,
especially when you are dealing with so many
layers of information. Here we are back again with the
whole structure of the piece which is quite rich and
it renders the complexity of the many overlapping and inter-playable stories around the Nobel Prizes, and this is not the ultimate point of view around the Nobel Prizes
but as a journalist, here at least we provided multiple levels for readers to explore. I explained how we normally work with these many layers of information, in a paper I’ve been
asked to write for the Parsons Journal for Information Mapping, and so we realized
while writing this paper that there are really
different phases of layer in both our analysis
and the data visually, that we normally follow. The process is of course
is not solely in error, as I will be telling you right now, it’s a constant iteration
and seeing if it works, and coming back and forth
but if I should start to, talk about phases, I would
say that first of all we always compose the main architecture of the visualization, the basis to which the main story will be
mapped and displayed, and so is it a timeline,
a scatter plot, a map? Is the organization of the
layout of the visualization? Then we start positioning singular element within the main frameworks, so each singular data point
finds its location within it, and these are of course
diagram and abstract diagrams, so it’s not a real visualization, it’s just to explain the phases. This is also the process where
we test the effectiveness of the main architecture
of the visualization. So the placement of the elements reveals if there are weaknesses in our models, and we have to maybe
design a different one. Then we construct shapes, elements, of dimensionality and form, with quantitative and
qualitative parameters, so each element starts to assume a color, a feature, according to the characteristic that we want to highlight. Then if there are internal
relationship between the element, then we start to
elucidate this relationship. Then we start labeling and identifying, so the addition of explanatory labels and short text to provide,
sort of like clarity, throughout the visualization
and to identify our elements. Then this is the main point, after we know that the
first story has been told, we are starting to
supplement the greater story through the addition minor
or tangential tales elements, and we consider it a very important step to contextualize the phenomenon. And as we’ve seen before with the Nobels, we can consider this part
as linking the main story, to external ideas, to other
times or other places. And the element of what
I call a secondary story, of course should be positioned
where they best help to enrich the overall comprehension. Where they fit into the data. For example, where before
we used the main axis of the visualization to then
aggregate the information. Then of course providing
small visual explanations such as a legend or a key. The process of creating the legend, always involves the simplification of the general architecture for example, the x and the y axes, base timeline or a scatter plot, as well as minimal explicit
shape, colors and dimension of the singular elements. And in the end, fine tuning
and stylizing the elements. This is very, very important
when we have so many layers. We really have to work
on, like you see before, really to make the most
important information pop out. Here we really work
with capacity of layers, with thickness of lines,
something very small, and probably that not always the readers can really know that you’re doing, but it’s very, very
important for the overall, it really makes a difference on how readers perceive the piece. Again, this is not
linear as it could seem, it’s a constant iteration of explorations, and we made no claim to have created the universal method to layer complexity, but in general, I do believe
that it’s really important to keep on exploring around that. Another challenge or let’s say approach, that guides my personal work, is the chase for beauty. Because we know that design and aesthetics has a very deep role in how users perceive every kind of product, which it could be a product but
also a piece of information, and I do believe that a
pure, beautiful visual can be an extremely powerful trigger, to get people curious
and willing to explore, the content of the data
analysis that you performed. This is why I think that in certain cases, the aesthetic aspect
of a data visualization can be considered as
important as the data itself. To catch readers’ attention, to make them willing
to dig into the topic, and to trigger their
curiosity to explore more. And why a reader shouldn’t be able to find data visualization both
intellectually compelling, so very rigorous data wise,
but also emotionally rich, and I really like that sentence. And where I want to go with it, of course, is that beauty cannot
replace functionality, but beauty and functionality together, achieve incredibly greater result. and this is what we like at Accurat, and why I like the idea
of making people think, oh, this is beautiful and maybe even strange in the first place but then, hopefully, I want to
know what it is about. But again, how can we do that? I’m just sharing my
personal way and method of chasing beauty. What I try to do when I
build data visualization is, never get inspiration from
existing data visualization. I’d rather get visually
inspired from various fields, and I always suggest others to try. I suggest that as designer
we should learn how to see before learning how to design, and to learn what is that
we like of what we see. What are the features of the things that would naturally attract our eyes. Personally, I’m very, very attracted by abstract art, for the
elegance of the composition, and the juxtaposition
of different elements, but also from the repetitive aesthetics of music notation for example, or the layering system
of architectural drawing, and I really think that
it’s important to observe and then to be able to
translate the features that your eyes are attracted by, to core principle of our personal style, and you will see what I mean with the next visualization
that I’m showing. This is still part of our
collaboration with La Lettura, then I’ll move on to show
you different projects. It’s a data narrative that took a very, very deep visual inspiration and I will explain the
visualization pretty quickly, and then I’ll get to the core of my point. This visualization
explored the phenomenon of the Brain Drain, researchers
that decide to move abroad in another country to pursue
their research career. It is exploratory
visualization that displays the main income and outcome
of researchers’ flow, on 16 countries and
discovering color again, correlation with other parameters. As in the Nobel Prize, we combine very many data sets and again, you have to imagine that
these were print pretty big, so I know that this screen, unfortunately, it doesn’t really make
good to this visualization. So the main architecture
of the visualization. The countries here are positioned, contrasting two parameters
and so horizontally, the GDP that each country dedicates for the research and development field, and vertically the number of
researchers per million people, so the very first point
of view of the story, you have it with the
positioning of the country, and seeing how they are
performing quite differently. Then each country is
displayed by contrasting a lot of information that
helps discover its situation, in terms of how many
researchers go abroad, which is the solid histogram going down, the red solid histogram going down. How many enter the country, which is the blue solid
histogram going up, and also how many are coming back after a period abroad, which
is the orange thin bar. But also we wanted to
add many more information to understand what’s
happening in the country, and so we included data about
the regular populations. Emigration and immigration
with the red and blue, hollowed histograms
following the main one, the GDP per capita, the unemployment rate, the university rankings, and
also an interesting parameter that we found, that is the
female employment rate, with the pink bar. And we’d be spending a lot of
time understanding the data, finding the most proper
data set to combine. There were many, many possibilities for how to visualize the data, knowing where the patterns were. But actually the visual idea for the piece came to me after a visit to the MOMA’s Inventing Abstraction exhibition
which was very fascinating, in 2013, I guess, which
happened during the first day that we were analyzing
the data that we had found on the researchers in countries. I was looking already in my mind, to a visual way to correlate
this many parameters, about researchers per country, and while walking past Mondrian, Malevich, or Kandinsky’s art pieces, I started to envision each
country as a compound element. The parameters of the element could have been visually
related by the position, and the rotation and
the spacial correlation of those geometrical shapes
that I was kinda like, sketching down during my visit. Consequently we normalize the number, so we represented each value as a function of the country’s population. Thus, in the visualization that you see, we are displaying relative percentages, to let readers visually compare
the relevant information, which happen to be much more interesting, that the values persist,
so the actual numbers. There are a lot of patterns
that visually emerge, so for example you can see
from the piece in general, that researchers move around much more than the average person
because for each country, the red and blue solid histogram
which are the researchers, are always longer than the hollowed ones. But we have some exception like in leading countries like Spain and Italy, in fact they import proportionally
more regular workers, say with a lower level of
education rather than researchers. And many, many more other
findings that one can dip into, but this is not the point today. I guess that visually
you can totally say that, the modern-painting,
visual reference plate is draw on the piece and how
it help not only creating a visually unconventional
and unexpected piece, but informing the choices that we made on how about tell you the story. And I want to stress this
point around aesthetics because to be communication designer,
information designer, and so most of all
data-visualization designer, is you have to find new ways
to attract people’s attention. Even like through new
languages and new solutions that besides being functional,
accurate and appropriate, they also must be magnetic
and surprising in a way. And I believe again,
that learning how to see is really essential to
learn how to design. And learning how to
see means really again, asking yourself what is that
you like of what you see. It means building your visual vocabulary. Which leads me to the next point which is try to not limit to
standards, when we can, as you probably have
seen the visualization that I’ve showed you before, the visual that we produce with data, are not really standard,
are not really something that you can get out of
an Excel or out of a tool that can create a visualization for you. They do rely on standard like timeline, bar charts or scatter plot, but they are not limited to that. Also as you’ve seen, I
sketch with the data a lot, when I work on any kind of
data-visualization project, I produce tons of sketches, even before pulling the
data into any sort of tool that can return me to a draft chart. I sketch to understand how to
spatially organize the data, to define both the
architecture of the composition and the visual aspect of the tiny details. And why I think this
is a valuable approach, specifically in data visualization, one of the most common
approach is to start from what the tools that you
use can provide you with. Even like processing tableau,
or the basic Excel graph, from what the tools can really return you. Or also maybe from what
we feel more comfortable in doing with these tools. But when we sketch with data
in designing data visualization like sketching with data,
what comes into mind, and the fact that I can’t in a way, have data on my pen and on my paper, it’s very helpful to
explore visual features, and a visual aggregation
that really comes exactly from what you have in mind, after of course you
analyze the data itself, and I really see that as
a shortcut from your head to the final piece. Talking about data and
sketching with data, I’ve been asked an interesting question by Moritz Stefaner in
his Data Stories podcast, that I will invite you
to check out because he always have amazing guests,
it’s called Data Stories. So he basically asked
me, okay Giorgia but, how about the real data, like
you sketch possibilities, but when do the actual
data come to the table? I would sum up, again here,
by saying that there are about three phases that in
parallel of the data analysis, are helping me draft the visualizations, with the sketches. So a first phase, when
I am interested in say the main macro categories of data, like the kind of topics
we are talking about, the eventual correlations
of pattern that we find, or even just the number of elements, are we talking about 50
entries or 500, 5,000? So understanding the
macro categories to start the visual possibilities
about the macro organization, again, the architecture of visualization. Then there’s a second phase where after we test our designs
with the actual data, after we see that
especially, how it works, I would just focus on
the singular elements or the singular entry point. I would sketch and
resketch shapes and details to figure out which features
to use to better represent them according to the variables that we have. To conclude I would
generally have a final phase where I would structure
everything that I expect to finally have in digital, but in paper. And here is again when
I’m pretty sure that we can go ahead with the
design that we created. This is just the final
piece that we’ve been doing before the sketches and I
think that it’s very important also to have refined
sketches to have something that is very sharable with clients. I always like to share
sketches with client in the first phases as
opposed to wireframes, because I realize that most of the clients take what you do digitally
even if you write like Draft, right in caps, they take it as more final as a sketch. So there’s less room for a discussion if you just share a wireframe,
than if you share a sketch. They also get more defensive
in general with wireframes. So, I mean a sketch is good. Then another thing that I
would like to talk about today that it might seem obvious but
it’s worth saying out loud, is that we should always
take into consideration what data stands for. What numbers actually
represent in their context and keep this in mind always. It varies at every stage of every project. And in my profession I really realized that when it comes to data, a common approach is to throw
technology at the problem. Sometimes without even
spending enough time framing the real issues
and the real challenge, like, “we have big data we
figure is due for a spark,” or whatever, this is really
something that I hear a lot from my clients but
these are just instruments, and over the last year
a lot have been done in research around what
technology can bring to the data world and this is amazing, but I do believe that
it’s the true convergence and hybridizations of science and design that can open new perspective and take the conversation to a next level. Because when we work with information it’s really so easy to get fascinated by the quantities, the
number, the variables, and it’s really easy to
lose track of what numbers really mean and what it’s
important to display. And with my company we do work with business and organization
of different nature, and for them we do build
software, custom platforms, we do really experiment with
different kind of technology, but we also every time, try to look at any of their data problem using a design approach
from the very early stage. A design approach that actually
limit the possibilities, to increase the opportunitites. To focus on framing the right questions and ultimately to reconnect
numbers to what they stand for. Which is people and behaviors. And to this regard, last year, a very particular challenge
was posed to my team by this woman, Samantha Cristoforetti, she has been the first
Italian woman astronaut, and well yeah, of course I’m Italian, so it’s easy to talk
about Italian projects. She contacted us before being launched on a six-month-long expedition to the International Space Station, and she asked us if we
wanted to collaborate on some real-time data visualization while she would be in space, how awesome, like how much data there
will be to visualize, like the orbit around Earth, the number of space walk, or the speed and positioning of the ISS, or also her Twitter feed
because she could Tweet from up there, visualize on a map. I could really go on
hours to list how many information there are
available to visualize, but a technology-driven
approach would have taken as far from anything meaningful, this is exactly what we didn’t wanna do an actually like who needs that. But still we were so compelled
to use all of the numbers we have to show the
power and the complexity of our engine to display everything that we analyze all together. But ultimately what both Samantha and I, and us wanted to achieve was a way to make her presence felt through her data. A way to remind people down here on Earth that there was a human
being orbiting around them on like beyond Earth. A human being who was trying to find ways to communicate with them. We decided to frame just one question to guide our design process. Is it possible to use this data to promote a very simple and basic human connection? If you think about it, the
idea was there already. In fact the quality of Friends in Space, which we like to describe
as the first social network that extends beyond Earth, is very simple. You log in with your social provider. Here is me. You can see Samantha’s real-time
position above your head, with the trajectory of her current orbit, which is the yellow arc. As well as a map of all of
the people around the world that are online in the same
moment with you in the platform. Well, yeah, it might
take you a few seconds before you realize that
those little symbols indicate all of the people
that right now are online with you on Friends in Space, represented through their
location on a world map that you can hide and show. As soon as you realize that one
of these little dots is you, the fun begins. So you just click Hello
and a simple arc on the map connects you to other
people who just say hello from all over the world, you’re linked, you’re connected with somebody
in Japan, in New Zealand, and actually you see it,
you see it on the map, and you feel that. And the whole idea is very simple, is that you are part
of a map of the world, that is connecting different people who at least share one
interest and one emotion. If you were lucky enough, you also saw and received a feedback from Samantha when she was waving back from the ISS, because everyday she like
Tweeted “hello earth,” with an hashtag that we retrieved, and specifically from her
position we just like, got the feedback to the
people that were connected. And all of this connection
have then been recorded for you in your control room in form of like visual
souvenirs from space. So you see that data powers all of this, in fact then a very simple
human interaction like, the way of connecting people
drives all of the experience, and you can see that we really limited the interaction by design. We wanted to play with the simplicity of the gesture of waving and saying hello from where you are to other places. The response of people was incredible. Really through Friends in Space, tens of thousands of people
connected with Samantha, and between themselves and
the very positive response really taught me an very important lesson, that was what I was telling
to you in the beginning, that limitations with
data are the true way to, in a way transform the
abstract and the uncountable, into something that can be seen, felt, and reconnected to our
lives and to our behaviors. This is my last one, my last point, very connected to the previous one, which explored limitation, and I’d like now to talk to you about how we can use data to become more human, to connect with ourselves and with other at a deeper level and to advocate that this is totally possible. The last experience is
definitely more radical in terms of using limitations as an asset. And for me it was a big
data hangover relief. It’s a zero-technological, year-long, very laborious personal project, that consumed, I guess, all
of my evenings and weekends for the last year, it
was a collaboration with information designer, Stefanie Posavec. I am an Italian and I live in New York, she is an American and
she lives in London. We only met a few times in our life, but last year we decided to work together, because we found that we have many personal and work similarities. We’re both the same
age, expats in our 30s, so both only children, struggling with being so far from our family. We both work with data in
a very hand-crafted way, so trying to add a human touch to the world of computing and algorithms. And most of all we’re both obsessed with drawing with data
and with sketching data. So we decided to challenge our self. We would get to know each
other through our data and through our drawing, of
course, drawing with data. And we conceived and started
what we call Dear Data, Sort of like uncommon
kind of correspondence of hand-drawn data,
postcards across the ocean, so each week since September 1st, 2014, and for a year, we
collected our personal data around a shared topic, from our complaints, to the
interaction with our partners, from the compliments we received, to the sounds of our surrounding, from the negative thoughts for the week as they popped up, to
our obsession and habits, 52 pretexts in form of data to investigate and reveal
a particular aspect to our self to the other person, of our self and about our
days to the other person. At the end of the week we’d then take, we would take the time to
analyze our information, and to create an hand-drawn
correspondence to each other. So unfolded, data postcard
that we would send from New York to London and
from London to New York, for 52 weeks. Eventually the postcards arrived at the other person’s address
with all of the scuff marks of this journey through the ocean with the third party in the project
being the postal service. So we purposely designed
initial constraint for the postcards and for the project, to form a consistent collection, but also to allow us experimenting more with our weekly data. The front of the postcards
is always the data drawing. It contains no explanation at all, and it’s hopefully beautiful drawing that one could only
take as an illustration if you didn’t know that
there is data behind. The back of the postcards contains of course the address of the other person, the title of the project and of the week, and the legend, so how to
read our data drawings. It’s interesting that we
didn’t send each other any digital scans of our postcards, so we have both been eagerly waiting to get the data weekly
portrait of the other person in the mailbox for a
year, also rediscovering the pleasure of checking
the postbox as you get home. It has really been a type of slow data, small data and analog data transmission. So, yeah, during a time when everybody’s talking about big data, virtual reality, we of course do small data and
physical postcards, you know? It doesn’t sound revolutionary, but, by removing technology from the equation we were really forced to
extend our self as designers. Because from the one hand we have each been forced to invent 52
different visual languages, because hand drawing
with data leads to design that are of course incredibly customized to the data that you’re
working with, but also, removing technology from the equation, triggered us to find
different ways to look at data as excuses to tell
something about our self. In fact, as we gathered our weekly data, the process was definitely
more labor intensive than just arriving standard method from our technological devices. We like to call Dear Data
more a personal documentary then a quantified self project, because we didn’t here
only quantify numbers, but we have been adding qualitative detail to our data collection and that was really the most important part for us. For example, the very
first week of Dear Data, we chose a pretty cold
and impersonal topic, how many time do we
check the time in a week. So here is the front of my
postcard and you can see that every little symbol
represents all the time I check the time, all
the per days, per hour, chronologically, nothing complicated, but you also see how added anecdotal details about those moments. In fact, the very different
instances of my symbols indicate why I was checking the time. What was I doing, was I bored,
was I hungry, was I late? Did I just glance the clock without really wanting to do that, and
this is the key part, so giving Stefanie or
the other person an idea of my days to the pretext
of my data collection. Something that is not really possible if don’t actively add
meaning to your tracking. Of course a lot of us
work with a personal data, we all have many apps that
are supposed to unlock key revelation about our self
and about patterns in our life but I think we shouldn’t really
expect a digital app to tell something about our self
without any active effort by us. We really have to engage in sense making of our own data. Most of the weeks of
our Dear Data project, the topics were something
that a digital app cannot track and that you
really have to engage in, because of course we can
also find data in our minds, and in the words we use and
not only in our activities. Which is even more compelling if the goal is telling something about
our self through our data. On week seven we tracked our complaints, and I composed this
musical complaints card, borrowing a very literal
visual inspiration from the music notation system to show the repetitiveness
of my complaints, of different type over
time and their pitch, their loudness through
the positioning of my complaints’ notes over
the lines of each score. Did I truly need to complain, and explain Stefanie how
to interpret my protests and being very honest
about how grumpy I’ve been in the true spirit of sharing. Then also using my data
drawing as a further data set, sort of like realizing
that the real data drawing could have been pretty simple
but you see what I mean, anyway, like admissible
complaint, not really so many. I really invite you to count
every time that you complain, because you realize that most of the time it makes no sense to do that. But then because we wanted to experiment we figure that we can find it even beyond the daily tracking and make a survey of what we own for example. Going into our wardrobes with
the eyes of the data collector looking for data in the way we categorize or classify our garments,
how many do we own, what color is, how often do we wear them, do we really need all of
the clothes that we have, and again for me, discovering
what should’ve been pretty clear from the beginning, (laughs) from the overall story of my closet. But when you see that visualized, you really see, oh my god,
really, really that is true, and you see that and
it really jumps at you. Or at week 46 we categorized
the books we own, or we explored our emotions, tracked the darkest feelings
of our negative thoughts for a week, which was
quite difficult to do, and made us realize how complicated it is to really discern what makes you feel bad, and what darkens your mind. It’s really hard to make
these things comfortable, but we tried. We also tracked our laughter for a week. Like what we did laugh
about and with whom. This is one of those data
set that in a way intrudes in our life because you can’t
fully enjoy your laughter, because you know that
you have to track it so, it’s kind of like, it
was really a tricky year. Or lastly we can also
try to use data to become better human beings, at least for a week, and to be able to perform acts, like to perform acts to
then be able to track them, like this week where we
purposely smiled to strangers and tracked their reaction,
if there was any reaction. I’m not going to show
you all of our postcards, you can find them online, but what I wanted to say with that, is that over a year, Stefanie
and I shared everything about ourselves through
the excuses of our data. We truly became friends through our data, and most importantly we
also started to look at data in our profession
through different lenses. I’m not suggesting you to start
drawing your personal data, or to find a pen pal across the ocean, even though actually,
if you want to do it, now we have an open
section of our website, where you can find a data pen pal, and it’s already full of people who are participating from all around the world. But anyway, it’s of course not imaginable that we all like hand-drawn
data in our jobs, and thank god we have our computers and our softwares, but
experimentation of this kind, where we radically limit our self, and where we drastically limit our tools and our possibilities,
can really teach us a lot about the perspective that we look at all kind of data from. Because by shifting away the
focus from the technology, we get closer to the real meaning, and then we can definitely bring technology back in the process. And yeah, it’s painful, it was painful, laborious, frustrating and it make you realize
that the undo comment is really amazing because
here you can’t undo, it’s painful when the postcard
is coming out very nicely and then you have to start from scratch because you make a mistake. But there is a value in
spending time with your data, and in this specific project,
since it was a process of discovery about our
self by hand counting, and hand analyzing,
analyzing our data manually, we really got to know our
self and the other person, at a deeper level, and
maybe what I’m saying is also that we have to spend more time with our data in general. Because if you think about it, even when we work with big data, like the whole point is making
it meaningful, contextual. It’s about making it
smarter, understandable, and actually smaller. So for sure, data is not
only a matter of technology, as I was introducing before, it’s most a matter of how we collect, process and relate with information. It’s a matter of how we design
the ways to look at data, and this is why I like to
say that I believe that data is more a state of mind and
that data can be an attitude, more than a matter of skills and tools, and that we can really
find it all around us, and become collector in our days if we just put on the right glasses. On a more personal
level, I also discovered how putting myself into a project, and making very personal
information public, not only to Stefanie
but to all of the people that saw the website, well it really kinda like, not only deeply, deeper, and deeply connected with Stefanie, but also to hundreds of people
that are taking the time to write us and say how
much they can relate to a specific topic, to
a specific revelation, because if you think about it
in each postcard for 52 weeks, we manifested our flaws, our geeky habits, geeky sides and habits and
we definitely didn’t share our best selves as we sometimes
would do on social media. We shared who we truly
are and doing it in form of tiny, quantitative bits,
honestly also helps you not being afraid of doing
that, they’re data after all. Ultimately I argue that
data can make us more human, and connect with our self and other, at a deeper level if we design
the right ways to do that. And almost done, more in general, one year of this very laborious,
personal project with data, one year of what I call,
unnecessary creating, really lead me to think about my work and profession different way. It also led to amazing
stuff like the fore coming publication of a book, various exhibition, and it is been really a
great learning experience, even almost like seven
years of professional work around the field. And I’m also happy to say that it made me reflect upon collaboration and it has not been easy at all, we definitely got on each
other nerves many times, we are two strong-minded designers with different design sensibilities, and there have been and there are, so many design battles around how to communicate the project,
how to shape the website, how to compose our talks, but also like let alone how to
design a book together that we are doing it right now. But with two of us, we’ve
also been able to stick to this project through holding
each other accountable, pushing each other forward every week. We have been able to
progress in our practice in a way that we could never
have ever done on our own. And think about it, if in your collaboration,
you agree on everything, the other person is not necessary, and like you know, I think it’s important
to always keep it in mind when in any kind of
professional relationship, you just desperately
want your ideas to win, but you don’t learn this
way, you don’t grow this way, and you don’t make big
steps this way, I believe. I guess I’m done? Thank you. (applause) – [Voiceover] Thank you so
much, that was fantastic. It was just so much to
absorb and to process. I’m wondering if you can talk a little bit about something you touched
on earlier in your talk, where you talked about data
strategy before design, and creating data processes before designing the data itself. Can you speak a little bit more to that? – Yeah, it’s something that
we started to do lately, because the more we work
with big organization and with organization that
have many different kind of department within their
business that work with data, we really understand that sometimes it’s, the reason why they’re not
optimizing their work with data, is also really a lack of
communication workflow, a lack of really a workflow for the data. Sometimes, often, very often, you have the data analyst that
just like analyze their data, and then you have the developers
that build the back end, and then you have a visual
designer that ultimately, sometimes they kind of
like call in the process, but the most of the data analysis and the data structure
has already been done. So what we are trying to
do at some organization that we are working with,
is really to talk to them from the beginning to
the end of the process, to try to start collaborating
on analyzing data before really
compartmentalizing everything. So what we are doing with
different kind of organization, specifically financial institutions, is really trying to bake the design into the beginning of the process. You need to be lucky to find a client that is willing to change their workflow, and the biggest the organization is, the hardest is to convince them. But I think there is a value
to that because otherwise, really having this very, very, compartmentalized, can you say that? Compartmentalized ways
of working with data. It brings you I think far
away from meaningful solution that last in the long term. It really, you do your
work with data but then you don’t really get into
any kind of innovation part. I’m not a data strategist,
we have a data strategist in our company because my
partner is a sociologist and have been studying
a lot of data strategy, so it’s definitely him
responsible to do that, but I enjoy being part
of that as a designer. Sure. – [Voiceover] Hi.
– Hi. – [Voiceover] Thank you so
much for giving us your time. I really enjoyed the presentation. – Thank you. – [Voiceover] I’ve heard
that Jer Thorp talked about how data, people working on data can be categorized in three different sections. Designers, artists and scientists. My thesis was on data visualization and when I was interviewing people, I realized that there’s
always this battle between scientists and designers or artists, that they criticize them
for the lack of accuracy with their visualizations. How do you handle these
kind of situations? How do you… – I think that like it’s
also mostly a matter of definitions and goals, there are so many infinite posts online between the differences between data art, and data visualization and
data design and something. Ultimately I believe that
it depends on the goal that you have in hand, so what is your communication
goal that you have with data? What I personally find interesting and this is also why I
don’t work by myself, but I work with a team, is
having a group of people that comes from different background, and that can really tackle
data from like 360 degrees, so in my company we have data scientists, data analysts, we have this
sociologist partner of mine that works with data strategies, we have interaction designers, and data visualization designers, and so I think that
depending on the goals, the best thing possible is
to have different voices in the process but I’m not sure
if I answer to your question so what is specifically
your question about that? – [Voiceover] What if I told
you that your visualization is a piece of art but
it’s not informative, like I’m not, it’s not accurate, like your numbers aren’t actually, your visualization
doesn’t actually represent the true numbers? – Well this is a matter
of going back and check– – [Voiceover] So it’s
misleading, it’s misleading. – Well if you do make mistake with data, and misrepresent the information, this is something that
you’re not supposed to do, and this is something that is
wrong in all kind of practice. As if you were a journalist and you tell a story that is not true, so I think that what I say,
what I think in my work, is that accuracy with data is granted, it’s the base level, we shouldn’t
even talk about accuracy, this is really something that
we should give for granted. If you make a mistake and
you misrepresent information, you’re not doing your job correctly. But I think that on top of
that, you have some goals. If you’re working with data
for decision-making purposes, for sure you’re not building
these rich and dense visual narratives because
you need to have people focused on the right
information immediately. Maybe you’re building a
very simple dashboard, that is something that
people can really understand at a glance but if you’re
telling for example, a journalistic story and if
your purposely experimenting, I think there is more room to say, okay, we combined this kind of information that maybe are not the
ultimate point of view on, the Nobel Prize’s history, or
the Brain Drain researchers, but this is a piece of journalism, and every kind of journalist, you have an interpretation of that. – [Voiceover] That’s great. – Yeah. (laughs) I think we can talk hours on that, so but Jer, I know Jer very well, I didn’t read his piece–
– [Voiceover] I saw his tweet. – I’m sorry? – [Voiceover] Oh, I saw his
tweet to the lady astronaut. – Okay, yeah, (laughs) yeah but, I didn’t know
that he was categorizing designer, artist and data scientist. Probably is because of the scope that they have. I think that data art, to me data art is a subset of data design, data-visualization design. But it’s really– – [Voiceover] I think
that layer of abstraction kinda takes away from what
scientists call accuracy. And that creates that tension. – To me accuracy is just not
misrepresent the information, and not really misrepresent the numbers because everything that you
saw before is really accurate, the numbers are placed in
the exact same position according to the years or
the numbers or the values. We can say that it’s
complicated and it’s rich, that it’s too rich, probably, but we can’t say that it’s not accurate. And if it’s not accurate it’s
because we made a mistake, but it’s not that, I
mean what I am saying, accuracy should always be there. And then we can talk
about what is the goal for your visualization. Yeah, thank you for your question. – [Voiceover] Thank you so much. – Thank you. Oh, I should speak here maybe. – [Voiceover] You mentioned
that data can make us more human, so what’s
the definition of human? I’m thinking data makes
us think more rationally, but I think human has some emotional thinking, so I’m sure the definition of human here. – Yeah, well, definition of human to me, is what I was saying when, saying that data can
make us more human is, what I then wrote as a tagline. To connect with our
self at a deeper level. Of course it’s a sort of provocation, because something that you
say, well, data and human, but what the Dear Data project
proved to me for example, I’m not a person that have ever been able to do meditation or to go to any therapy, because of how I am, but really focusing and
acknowledging a topic, in form of data, help me
understand a lot of thing about myself, over a year. So I’m saying that sometimes quantifying or giving yourself the excuse, the pretext and the goal, of starting by quantifying
something around yourself, and your inner self,
give you the real pretext to do stuff that maybe
you wouldn’t do before. So for example, there has
been in this 52 weeks, a lot of topics I wouldn’t
have been able to address, like how much for example, how undecided am I in
my general situation. I think of myself as a
very decisive person, but a week of indecision
like really made me realize that I’m not for the most trivial parts. Like negative thoughts, when you’re really trying to discern what’s going on in your mind, when you feel that there
is something wrong, but then you have to do that
because it’s an exercise and you have to put it in form of data. What I’m saying is that data
can be a tool to help us, definitely not the only tool, it can be one of the many,
many tools that we can have to connect without our
self at a deeper level. That is why. Maybe human is not the
best word that I can use, but I think that I’d like to talk about a possibility for the
future to think about it as a humanism rather than a data, I don’t know, technological-only,
driven approach. It was of course a provocation, so thank you for your question. – [Voiceover] Thank you. – Thank you. – [Voiceover] Hi.
– Hi. – [Voiceover] I guess I
was wondering how diverse a range of industries do you
work in through your clients, and then if you’re working with a client who’s from an industry that
you don’t have that much personal experience in, how do you, how do you develop the insights into like, what is the right question to ask? When you’re dealing with
something that’s very unfamiliar. – Yeah, definitely interesting question. We work with a lot of
different industries. We work a lot with banks
so financial institutions. We’ve been working with
the healthcare system. We work with foundations. We’ve been working with the UN, so the United Nation,
building their reports. So really with a lot of startups, startups that are definitely
focused on different aspects, and what we do because we are not expert in the domain field, we always
set up with our clients, a very first phase where
we dig deep into the topic, and we always, always, make sure, that throughout all of the process, we work with one of our clients, like somebody within
the team of our clients, that is expert in their domain. Because we can’t like fake or pretend, that we are most expert in healthcare. We are expert in designing the ways information would be perceived. So what we do is we
always, always, always, collaborate very strictly,
like in a strict relationship, with the core, knowledge
domain of our clients, and for that we, every time that we have a project, we start with two days
of kick off meeting, like two-day workshop with the client, and then we also have very, often moments of feedback, so we really kind of
talk not only everyday, not necessarily every day
but every couple of days, and we send draft and then, all of the process is very imperative, but what I think that the
answer to your question is, we don’t even pretend to become more expert than our clients, we expert in our job which is
how to visualize information, how to make this information accessible, but we do really work with expert. This is actually also why I love my job, is because every time
and in every project, I learn a lot of some, I mean, I really learn something new, and I can dig deep into a different aspect or field. – [Voiceover] Hi.
– Hi. – [Voiceover] I’m Layla,
thanks for coming. – Thank you. – [Voiceover] I was wondering if, you have any example or
story of some of the, of how some of your work has influenced or provoked changes in the industries that you were working for? And the reason why I am asking this, is I sometimes help, my boyfriend’s in a urban design program
that has doing a lot of trying to big data and data visualization, and a lot of the feedback
that he gets is like, okay, data and then what, and I think I understand why, but that he gets that feedback so often, is something that I find interesting. – Yeah. The first story that comes to mind, is we were working a lot
with an Italian bank, Unicredit Group is a European, kind of like pretty big bank and for them, we’ve been building an HR monitoring tool, and so a visual, a sort of like interactive tool, when they can monitor the state of the, the state of the art of the people, like their managers, their agents, then how much they do deliver what they are expected to deliver, how much time they do spend in meetings, so sort of like a huge
grant that put together not only a timeline but
also like singular people, and like really they, and it’s also like connected to LinkedIn so it sort of like also
helps them understand what are the skills of the people, and I remember that I was sitting there with one of my clients when
he was looking at a prototype, and he realized that they were completely not covered into one of the core areas, of what was one of their
goal for the previous year, and so then he was just like,
really, are we doing that, and he like called the
people and from that, they started a conversation
on how to optimize a specific part of their big company, that they just weren’t able to monitor at a higher level. So I think that the most insightful, how to put it, the times that I found our tools being more useful, is when we give these
tools to the top managers, that normally they don’t
have access to data, because they can’t really read the data, and they don’t really have a lot of time to just like dig deep. They rely on the people that
are below there in hierarchy to say that everything is going okay, and it happens a couple of other times, that because of these kind of tools, they changed their process of optimization of resources. Just to make it a little more abstract, maybe when you give
access to data to people that are able and in
power of taking decision, that they didn’t really
have access to data before, I think it’s there that you can advocate that there will be a sort of change, and of course you can have the person that is below you that gives you PowerPoint
with some numbers, but when you’re really able to
just play with a visual tool, that’s something that you’re
not required to be a programmer and just look at the data yourself. I think that there some
discoveries might happen. It might not happen but, if there’s something wrong
hopefully it will pop up. So many questions, thank you. – [Voiceover] Thank you so much. First of all, thank you
so much for coming here, and sharing your amazing
work with all of us, is was very inspirational. – Thanks. – [Voiceover] I have two questions. The first one is, what was a big step that you took to actually cofound your own company? Awesome, and then second one is, have you ever, do you have like, really scientific data, that has to be translated
into informational graphics, or have you ever have any abstract data that has to be translated
into informational graphics? – What do you mean by abstract data? And then I will answer
to your two questions. – [Voiceover] So in very unusual topics, that are not so scientific
and straight to the point, does it make sense? – I’m not sure if I got the question. – [Voiceover] Your data
like all the topics, (speaker away from microphone) – Okay so something that
you have to find your data, okay, so starting from a topic, yeah I guess so like
Dear Data is the perfect, I’m sorry if I didn’t get the question. I think that Dear Data
is the perfect example to say okay, you have to
address your complaints. You have multiple ways to do that, just for example, just tracking down every
time that you complain, and just that. But then you can really be
creative and find ways to say, okay but what is that I
was complaining about, who was complaining it with, or maybe goals with
some contextual details, for example, normally I am a
person that I’m always cold, and so maybe if I say that okay, it was cold in this
place and I complained, I have a contextual detail so for example, if I’m tracking my emotions
I can just say, okay, I felt an emotion, one, two, three. Or then I can say what
kind of an emotion it was, what was it trigger by, or can I add contextual
details for example, about the weather, oh I don’t know, something that helps you discover. So I guess that when you
have abstract topics, you can either, when deals
with your personal data, of course you just have to be creative and find the right question, right, to ask to the topic and
the right way to put a abstract topic into quantities, but in our job in my company, when they say, okay, for example, in these pieces for the Nobel Laureates, “oh yeah, it’s the anniversary
for the Nobel Laureates, “can you guys do something around that?” And then we start to say, that’s okay, what are the informations
that are available? And then we go and
define the main data set, which was just like when the
Nobel Prizes were delivered. And then starting by that, okay, well what can we add? And then you go and dig
deep and find online, and in different kind of like
online databases, your data. So that’s the second question, but like, for the company, I started architecture but like
within all of my five years, I was more interested
in the representation of buildings, in the design of buildings, and my, yeah, I’m trying to make it short, and my last, my master thesis
was an urban mapping project, and so it was like, of course
in the realm of architecture, because I was dealing with the city, but I was already mapping information, and after that I went to work for an interaction design firm, and there I met one of my two partners, the one that is the sociologist, and within this interaction design firm, it was a big firm in
Italy, like 40 people, it’s called Interaction Design Lab, my partner, Simone and I
decided at a certain point that we wanted to focus
on information design, so it’s not only interaction design, all around we’re really
working with information. And we sort of like built our own small company within the company, and so we tried to start
to find our own clients, and that was something that was agreed with the partners of course, because the information design part was like a new tiny branch, and then we figured that we were, pretty good at doing that, so also to find clients
and to manage a small team, and we’re just like four people, and after that we sort
of said that naturally, we wanted to build our own company. Then for another occasion, like another person that we knew that is now our third partner. He was working on one of
his own previous company, he was doing motion graphics and videos, and he was good at
making the business part, but he wasn’t interested
anymore in the video output, and then we just started to try to see if we wanted to do something together. The specific occasion
for building the company was one of those things
that always happens, so there was an RFP, a big RFP, that somebody sent our ways, and we were like, why we just don’t try? We tried, we won the RFP and
then we had to build a company because otherwise they couldn’t
hire us as professionals, and single professional. We started in three with
two other designers, and now we are 22. We didn’t really plan in the
beginning to become that big, I mean, I know it’s small but
for being a design company, I think, to me now it feels big, but I guess that data-visualization market is very full of opportunities now. So we’ve gotten so many requests from very different kind of clients, and we also tried to build our way to be in a way a little
recognized for what we do, because we published, we
did projects for visibility, and then I went on
speaking at conferences, and so it’s not only
that they are calling us, you also actively have
to promote your work, promote yourself, even
if this is something that some designers don’t really like to do. I think it was both a
coincidence of people that were interested in the
same topics in the same moment, and also the fact that then,
when we started the company, we realized that there was an opportunity to build a company and not
only a tiny design studio. I don’t know if you have more
specific question on that. – [Voiceover] No, that
answers my question. – Okay, thanks. – [Voiceover] Do you think you
have time for just one more? – I do, yeah. – [Voiceover] Thanks for
coming, very fascinating talk. – Thanks. – I guess have two questions too. One is we had a studio visit last week, and their big mantra seemed to be, breaking free of the screens
and the constraint of screens and the two dimensions that
data visualization traffics in. I was wondering if you
felt constrained by that, and then not to put the
tool ahead of the context, but are you interested in other ways of exhibiting this data, either sonification or virtual reality? Then my second question was what do you think are the most valuable tools for a designer to have, in her tool set? – Okay. I think the first question, it always depends on the goals. When we work with data with our clients, we 99% of the time, we are required to build something that is specifically a
web tool, a mobile tool, so something they already dictate the fact that they will be
using that tool in this way. So we are somehow constrained
by the dimensionality. I think that there are a lot
of opportunities for sure, to work on other kind of
ways to represent data, maybe as you said, in a sonic way, or with VR also. But what I’m feeling is
that most of the market is not ready yet, so maybe there are, definitely possibility
to artistic exploration, and I think there have
been some very interest in artistic exploration
and how you can yeah, merge VR with data or also you can use sounds for representing data. It’s just that as always, all of this kind of avant garde has to be digested
before being really used. What I’m saying is I’ve
never had any clients, or any request for doing
something different than be dimensional data
visualization with data. The second question’s about the tools. Again, it depends on how
and what you wanna do. All of the visualization that you saw in the first part of the presentation, so the data visualization for La Lettura, we’ve been doing that with
Excel and Illustrator. Really nothing more. It was for sure more labor intensive and slower than if we would have had Web Developer back then, but it also set up our method as, I myself, I don’t code, I
will never learn how to code, because it just doesn’t make sense because I have developers, but I still can work with data. I guess what I’m saying is from the basic pencil and paper, to Excel and Illustrator, and to then maybe having
something that helps you draft visuals before you
refine them, for example, Tableau is an interesting tool because more and more it allows you to also have multiple and interactive views on a dashboard and it’s not 100% intuitive as it can be, I don’t know, a Word
document, a word program, but it’s still something that
we found some of our clients being able to learn how to use, and then of course it depends on how much you wanna go dip into coding. So I don’t know, what do
you mean exactly by tools? Do you mean like technological
tools or maybe… – [Voiceover] Yeah you mentioned D3 or whatever. – Yeah, our developers were
doing everything custom. They use D3 libraries sometimes but they just really work on, I mean, they are just amazing,
I don’t know how they do it. – [Voiceover] Well Giorgia,
I’m sure some people wanna speak to you one on one. How proud we are to have you here today. Thank you so much. – Thank you for having me. (applause)