Fundamentals of Design for Non-Designers
This is a compressed, annotated version of a presentation I gave at Esri Australia's Ozri conference in September 2015. It's intended to be a high-level overview of a few visual design principles that are relevant to mapmaking. Being mindful of these principles while you're making and viewing maps will help you improve your visual communication skills. Check out the resources section at the end of the presentation (from slide 44) for a bunch of links to useful blog posts, websites and online tools.
Published on: Mar 3, 2016
Transcripts - Fundamentals of Design for Non-Designers
(e.g. GIS people)
This is a re-edited, annotated version of a
presentation I gave at Ozri 2015, in
Get in touch if you have any questions or
is clear thinking
I’m going to teach you a new language.
Because visual design, just like a spoken
language, is just another way of
The power of a great map is in its ability to
clearly communicate information, and open
up new insights into the data it presents.
But the problem is that very few GIS
professionals have any formal visual design
training. For those of us who live and work
in the world of data and analysis, design
David Imus - imusgeographics.com
Humans are visual creatures, and we can
intuitively recognise good design when we see
it – and not just when it comes to maps.
Nate Wessel – cincymap.org
John Nelson – UX.Blog
But as non-designers, it can be hard for us
to articulate exactly why these things are
Visual design can seem almost like a
Actually, the fundamentals of design are a
lot like the fundamentals of a spoken
There’s an artistic vocabulary,
and these elements make up
the content of our
…and there’s a visual grammar,
the principles of which make up
the building blocks of visual
composition; they give context
to the content of our designs.
The relative importance of each
of these elements and principles
changes, depending on what
exactly it is that you’re designing.
If you’re making a flag
or a logo, you’re
with generalisation and
If you’re an architect,
you’re going to be very
mindful of space in
And if you’re designing
a quilt, you will be
paying the most
attention to things like
pattern, contrast and
But what about maps?
When we’re making maps,
we’re usually dealing with
complex spatial data, and
we’re often trying to
communicate very specific
messages to the map reader.
So, given that our work as
mapmakers tends to be data-
driven, what are the
fundamental elements of
Relative visual weight
The overarching principle of cartographic
design (and actually of most fields of design),
is that of hierarchy.
Hierarchy is all about defining the relative
visual weights of the elements on your map.
If the data you’re displaying on your map is
unweighted (in a visual sense), it can be almost
impossible to decipher what the map is about.
Of course, I think most of us have an innate
understanding of the importance of visual
weighting, and even our most basic maps tend
to look a little better than this.
We naturally tend to establish a visual hierarchy
by manipulating the size of elements…
…or their shape…
…or through the use
…or by adjusting the visual
contrast of the elements.
But to take this to the next level, you
need to ask yourself: what is the point
of this map? What message am I
trying to communicate to the viewer?
When considering the visual hierarchy of your
map elements, you should try to boil them down
into one dominant theme, one sub-dominant
theme (if required), and have all of the other
elements on the map being subordinate, from a
design point of view.
So on this example…
…the dominant element
could be these points…
theme might be transport
(roads and rail)…
…and the rest of the data on the
map can be subordinate in the
Once a visual hierarchy has been
defined, the next step is to make
that hierarchy explicit with
So we could take the points on
this map and make them even
bigger, and give them shadows to
make them even more visually
We could increase the boldness
of the roads and rail lines (but
only a little.)
And we could really push the rest
of the content on the map into
So when you’re considering hierarchy, the key thing is
to really examine your data from the outset, and
decide what the key message of your map is going to
Then, group each element of the map into the
dominant, sub-dominant and subordinate categories,
and symbolise them accordingly.
Contrast without conflict
The effective use of colour on a map is all about
making sure that the elements on your map are
coloured in a way that they’re distinct where
necessary, while still ensuring that your map as a
whole is visually harmonious.
Different colours actually have different visual weights.
If people are shown an image like this one, and are asked
to move the fulcrum so that the two colours appear
…they will (on average) tend to move it towards the red
square, indicating that red is perceived as having a
heavier visual weight than blue.
Given enough iterations of this
experiment with enough
different colour combinations…
…it becomes possible to build up
a picture of the relative visual
weights of different colours.
That sounds easy enough in
principle, so what if we try to
apply that hierarchy of colour
weights to the visual hierarchy
we’ve already established?
Turns out, it looks pretty gross.
So there’s clearly more to colour
than just the weight of the hue
The apparent weight of a colour
is also strongly affected by its
saturation and value.
Saturation (or chroma) referes to the
quantity of a given hue.
A more saturated, purer colour has a heavier
visual weight than a less saturated, grey
And value (or brightness) describes the
extent to which a hue is tinted towards
white, or shaded towards black.
A colour with a darker, lower value has a
heavier visual weight than a brighter, whiter
version of that colour.
So if we take the saturation and value part of the equation
into consideration, we can come up with a much nicer
colour scheme for our map.
But this brings up the issue of colour schemes in general,
and how we as mapmakers tend to select and use them.
Even on this improved map, all of the colours come
from the default palette. This palette does the best
it can with just 120 colours.
But there are millions of colours available to us;
understanding how to use them can really enhance
A nice example of this is the Esri
It uses a very carefully chosen
selection of blues.
The way that these blues communicate
depth and shadow is much more effective
than would be the case if they had just used
the blues in the default palette.
So try to break away from the default
palette as much as possible, and really think
about your colour choices.
There are some great online resources that
can help you select colour schemes and
palettes; go to the end of this presentation
for some links!
If we don’t limit ourselves to the
default palette, a much subtler
colour scheme is possible for our
Maps are representations of reality, but reality is
So generalisation in cartographic design is all about
creating clarity and insight by abstracting data.
You need to simplify your data where necessary,
and direct visual emphasis to the most important
elements of your map.
Raw SRTM Data Manual Shaded Relief
A good example of this is using shaded relief
techniques to illustrate the terrain of a region.
You could take raw elevation data and
apply a simple hillshade to it, but at
smaller scales this doesn’t really describe
the true nature of the landscape.
Instead, use the raw data as a starting
point, and then generalise the data by
removing unnecessary detail. This draws
attention to the overall shapes of the
mountain ranges and valleys.
These manual shaded relief maps by Tom
Patterson are a really nice example of
They use generalisation techniques such
as subtraction, combination, smoothing
and selective enhancement to make a
much more effective map.
Tom Patterson is also behind the Natural Earth Project, which is a set
of spatial data – borders, roads, urban areas, terrain etc. – that have
all been manually generalised to look their best at fairly small scales.
The key message with generalisation is that it’s totally ok to
alter the geometry of your data, in order to make your map
This might feel a bit wrong to those of us who are data-oriented, and
who generally think that more spatial accuracy automatically makes a
better map. But all maps are generalised, to some degree.
So how might we apply
generalisation to this
There’s a lot of unnecessary detail
in the rail network, such as rail
yards and duplicate lines. It’s just
visual clutter, so let’s get clean
No one’s going to use this map for
navigation either, so we can do
the same thing to the roads.
And we can generalise the main data (the points)
on the map as well, perhaps by abstracting their
spatial locations and dispersing them…
…or by summing them up at generalised
…or by binning them into an areal unit and
displaying the results as a choropleth map…
…or even by highlighting concentrations of points
by creating an interpolated density surface.
Obviously there are tons of options when it
comes to generalisation, and which techniques
you select will depend on the specific nature of
Remember, though, that “good design must be
honest” (Dieter Rams).
Don’t generalise your data to the point where
(intentionally or not), it starts to tell lies!
The best maps out there tend to have a simplicity
to them which is the result of a lot of deliberate,
thoughtful decisions by the mapmaker about
what can be safely abstracted for the sake of
Drive your point home
If the previous design principles were about
taking a high-level view of your data, and
identifying its key points to display it in the most
effective way possible…
…then composition is all about taking a high-
level view of your map as a whole, and making
design decisions that tie everything together.
I realise this might be hard for GIS people to
relate to, since the arrangement and styling of
elements like legends, titles, charts and scale
bars is pretty far removed from the underlying
And defaults and templates are both a
blessing and a curse here.
It’s so quick and easy to create these
elements in ArcMap, but this in turn makes
their appearance sometimes seem generic,
and their arrangement can seem almost
But by applying the design principles I’ve already
shown you to your map surrounds rather than
your map data, you’ll find it’s actually pretty
straightforward to create more effective map
Here you can see the principles of
hierarchy and colour at work, giving
consideration to the relative visual
weight of each layout element.
And if another form of data
visualisation, such as a chart or a
graph, can help you get your point
across, then it should absolutely be
incorporated into the map layout
itself, rather than have it sit on a
separate page or in an appendix.
When it comes to map composition,
it’s worth going the extra mile for the
boost it can give to the overall
accessibility of the map as a whole.
So these are the fundamentals of
visual design that are most relevant
to mapmakers. Pretty simple to
understand, aren’t they?
Now that you’re a budding
cartographic designer, where do you
go from here?
Well, now that you have the knowledge required to break a map design down into
its individual components, you should try to do exactly that, as often as possible.
Look at your own maps, look at maps made by others, and even look at non-map
examples of visual design.
Use your newly-trained eye and deconstruct everything you see into these
component design elements; try to describe how each one has been used.
David Imus - imusgeographics.com
For example, on this map, the saturated
yellow of the urban areas (with a higher
visual weight) stand out as the
The sub-dominant roads are darker in
value and less saturated.
And the rest of the topographic data is
visually subordinate, fading into the
background with lighter hues and even
There is a clear grouping of labels into
dominant, sub-dominant and
subordinate hierarchical levels, based
on size and weight.
And in terms of the roads and natural
landscape, a lot of thought has gone
into removing unnecessary features for
the sake of overall clarity.
On this map, obviously the spatial
locations of the cables themselves have
been heavily abstracted.
And that’s because the point of this
map is not to show the exact path that
the cables take across the seafloor, but
rather it’s about the magnitude and
type of linkages between continents.
The unified design theme in the
composition of this map is also pretty
remarkable; the theme has heavily
influenced the colours and fonts
chosen, as well as the overall layout.
The vignettes in the corners show cable
capacity; and the apparent scale bars
actually provide data on signal latency.
So the composition of this map actively
encourages the reader to engage with
the data it presents.
To emphasise the importance of
composition, here’s another cable map,
also by TeleGeography, and showing
much the same data as the previous
But the design decisions made in the
composition of this map mean that it
has a completely different, more
Notice in particular how the orange and
blue colour scheme, chosen to
compliment the underlying satellite
imagery, flows through to all of the
hierarchical levels of design on the map
- and these colours can even be seen in
the charts in the map layout.
This one is just a nice example of how
less can be more in map design.
The solar radiation data has been
generalised into just eleven categories.
And there’s no other spatial data on the
maps besides light state borders.
But the unambiguous colour scheme
and the use of small multiples in the
composition makes the temporal
nature of the overall story crystal clear.
Nate Wessel – cincymap.org
The data shown on this map is only that
which is relevant to cyclists.
The map as a whole has a very
harmonious pastel colour scheme; but
the subtle differences in hue, saturation
and value are enough to differentiate
By making the bike lanes and paths
thick, black, visually dominant lines, a
clear visual hierarchy is established in
A lot of thought has gone into the
design of this legend. It’s
informationally dense, but still
aesthetically pleasing, and it invites the
reader to explore the data.
TRAIN YOUR EYE
The other thing you should do, besides
training your eye for visual design as
much as possible…
…is to lay off the templates (and pre-
symbolised layer files, and default
symbology), for a while at least.
Making a great map requires conscious
and intentional design decisions, and
templates tend to get in the way of
Build up your maps element by
element, being mindful of each of the
key design principles as you go.
Finally, I think the most important thing
you can do to improve your design skills
is to work with other mapmakers.
Share your maps as much as possible,
and collaborate on design challenges
To return to my language analogy,
collaboration is so important because,
just like learning a spoken language,
you will only really improve by having
conversations with other people.
So I hope I’ve been able to at least give
you a phrasebook that you can use to
start speaking the language of visual
And I’m looking forward to seeing some
of the beautiful maps you’ll go on to
Get in touch if you have any questions or
Now check out the collection of links
and resources at the end of this
presentation, and get inspired!
MAPS USED IN THIS PRESENTATION
The Essential Geography of the
United States of America
This map is the result of nearly 6,000 hours
of thoughtful design and attention to detail.
Middle East Telecommunications Map 2015
All of the maps in the TeleGeography map gallery
are great examples of how a well-designed map
facilitates curiosity and understanding.
Big thanks to all of these cartographersfor lettingme use their work
to illustratemy points. Check out their websites for more inspiration!
Submarine Cable Map 2015
The Dispersion of Life and Gender in New York
Simple, almost minimalist design can
sometimes be the best way to illustrate a
story told by complex data.
Cincinnati Bike Map
For a map that communicates so much information,
good design choices make this one impressively clear
and readable. Great legend, too!
Where America's sunniest and
least-sunny places are
Unambiguous colours and effective
composition give this simple design a high
degree of clarity.
Esri Map Book
Past volumes of the Esri Map Book are
available to browse online, highlighting
some of the best examples of map design
using the ArcGIS suite.
Atlas of Design
Two volumes of diverse and very well designed
maps. The blog on the website also has some
good detailed discussion on the design of
Semiology of Graphics
Esri has reprinted one of the foundational
books about cartographic design, Jaques
Bertin’s Semiology of Graphics from 1967.
Non-Designer's Design Book
An accessible, in-depth introduction to the
key principles of visual design.
Creative Workshop: 80 Challenges to
Sharpen Your Design Skills
While not specifically related to map design, most of the exercises in
this book can be modified slightly for cartographers. Giving yourself
constraints can be one of the best ways to unleash creativity.
Steal Like an Artist: 10 Things Nobody
Told You About Being Creative
A short, humorous overview of the professional challenges
faced by visual designers. I found that almost everything in
this book applies equally to map makers.
A collection of visual design-related resources
that I’ve found to be useful and engaging
Why city flags may be the worst-
designed thing you've never noticed
Roman Mars on flag designs; specifically, why
municipal flags are often so terrible. An engaging
perspective on the perils of “design by committee.”
The Fundamental Elements of Design
A quick overview of the basic building blocks of
John Maeda on the simple life
Designer John Maeda talks about the merits and
importance of simplicity.
These are all online palette creation apps; very useful
for creating colour schemes that work.
This is a simple but very effective tool for creating colour
gradients. Playing with this will give you a clearer understanding
of how colour spaces, hue and saturation all work together.
The New Defaults
These sites are both attempts by designers to create basic palettes that
are much nicer than the default Windows palette. Grab the RGB
values and use them as a base for your own ArcMap colour styles.
Adobe Color CC
A nice selection of user-create palettes and
Better default color palette
MORE ONLINE TOOLS
These are both examples of online map styling
tools. These are a good way to get a quick feel for how
different colour schemes work on a map, even if
you’re not creating an online map.
A good site for general design inspiration, but the
real value is in the ability to search by colour.
An excellent repository of fonts and symbols,
free for commercial use.
Mapsense CSS Machine
Vector Tiles Styler
Daniel Huffman’s Project Linework - A set of
generalised administrative boundaries for small-
BLOGS AND ONLINE
Maps We Love
Esri’s “Maps We Love” blog. A good showcase of
what’s possible using the ArcGIS platform.
Kenneth Field blogs about a different map each day for a
whole year, highlighting what makes each one great (or
not so great). A really diverse selection of maps here,
great for honing your visual design critiquing skills.
Another blog by Kenneth Field, which largely
focuses on poorly-designed maps, and what could
be done to fix them.
Make maps people want to look at
An introductory article from Esri on the basic
elements of good map design.
Map Design (wiki.gis.com)
This is the root page for Map Design on Esri’s GIS
wiki. Check out some of the linked articles and see if you
can contribute some improvements or additions.
MORE BLOGS AND
UX.Blog (John Nelson)
Insightful posts on data visualisation, with a strong
cartographic bent, and a good eye for design.
The Cartographic Aesthetic
The rest of the blog is worth a look, but this post in
particular could make you reflect on why your maps
look the way they do.
Labeling and text hierarchy in cartography
A good, detailed article about labelling maps, covering
the importance of hierarchy in label placement, as well
as font selection.
NACIS’ CartoTalk forum - a great place to get feedback
on the design of your maps.
Material Design Specification
This is Google’s design specification for their
“Material Design” standards. It’s worth having a look
at, not necessarily for the specific details, but just to
get an idea of what a detailed design guide looks like.
Cartographic Design Principles
A very good (and recent) series of brief blog posts by the
UK Ordnance Survey on cartographic design.
MapLift is an initiative by NACIS to improve the maps
that appear on Wikipedia articles. This is a good chance
to flex your cartographic muscles in a collaborative
Color Theory for Designers
Part 1 of a three part series, which covers colour
theory, composition, and creating your own
How we created color scales
An interesting account of the process used to
design a colour scheme from scratch.
Does red weigh more than blue?
A short an interesting article about the relative
“weights” of different colours.