V. Kupriyanov - The Power of Data-Driven Storytelling

Slide 1

Slide 1 text

The Power of Data-Driven Storytelling @v_kupriyanov

Slide 2

Slide 2 text

Hello! Volodymyr [vo-lo-di-mir] Kupriyanov Data analyst / journalist / consultant Ukraine -- UK -- Denmark

Slide 3

Slide 3 text

My work I help digital marketing teams and agencies use data in their marketing campaigns

Slide 4

Slide 4 text

How? By gathering and analysing data Digging out insights and stories Helping with visualisations

Slide 5

Slide 5 text

Why? Data = source of stories Viz is a powerful way to communicate them Interesting stories make for effective campaigns

Slide 6

Slide 6 text

What’s new? Marketing & PRs have been trading in stories for ages

Slide 7

Slide 7 text

Getting (more) data-driven

Slide 8

Slide 8 text

Three directions

Slide 9

Slide 9 text

1) Data-informed ideation Customer surveys Industry reports General news and cultural landscape Web analytics ++

Slide 10

Slide 10 text

No content

Slide 11

Slide 11 text

Core questions What does our audience care about? What topics are of interest to them? What content do they like to consume?

Slide 12

Slide 12 text

2) Grounding campaigns in data Government data and national statistics Industry reports Internal client data Surveys

Slide 13

Slide 13 text

Core questions What data source can we use? How recent and comprehensive is the data? What biases does it carry?

Slide 14

Slide 14 text

3) Data-driven visualisations Dashboards Static graphs Big interactives Motion graphics

Slide 15

Slide 15 text

Core questions How do we visualise this? What makes the most sense for ● audience ● data ● budget

Slide 16

Slide 16 text

From theory to practice

Slide 17

Slide 17 text

Example 1

Slide 18

Slide 18 text

Client Business ﬁnancing platform focused on tech and startups in the UK

Slide 19

Slide 19 text

Audience Founders and entrepreneurs People working in tech

Slide 20

Slide 20 text

What are they interested in? Tech, innovation, the startup scene Famous tech companies Famous tech founders and entrepreneurs

Slide 21

Slide 21 text

No content

Slide 22

Slide 22 text

Existing narrative

Slide 23

Slide 23 text

Existing narrative

Slide 24

Slide 24 text

How can we enrich that knowledge? What data can we use? How can we visualise it in a way that would show something new?

Slide 25

Slide 25 text

Relevant data source

Slide 26

Slide 26 text

Four things we were after For each “maﬁoso” we wanted to ﬁnd out the companies they went on to… 1. Start 2. Invest in 3. Lead 4. Advise

Slide 27

Slide 27 text

Some light-hearted scraping

Slide 28

Slide 28 text

Some light-hearted scraping

Slide 29

Slide 29 text

Following the hunch We knew … that these guys founded some big-name companies … that they often worked and invested together But what we didn’t realise was just how connected they are

Slide 30

Slide 30 text

What is an effective way to show connections?

Slide 31

Slide 31 text

Prototyping

Slide 32

Slide 32 text

No content

Slide 33

Slide 33 text

Core story points The sheer degree of collaboration between members of “the maﬁa” Well-known names, like LinkedIn and YouTube Staggering number of ventures Non-tech ventures, such as “Thank You For Smoking” ﬁlm and the lobbying group FWD.US

Slide 34

Slide 34 text

Scrollytelling

Slide 35

Slide 35 text

Interactivity

Slide 36

Slide 36 text

Mobile

Slide 37

Slide 37 text

Results Features on VentureBeat, Gizmodo +70 sites Thousands of social shares and engagements Building authority as a tech-savvy brand

Slide 38

Slide 38 text

Personal highlights

Slide 39

Slide 39 text

Why did it work? Tapped into what the audience care about Used relevant data to add insight Visualise it in a way that reveals something new Make it work for the audience and platform

Slide 40

Slide 40 text

Example 2

Slide 41

Slide 41 text

Client GPS ﬂeet management solutions provider (≈ telematics)

Slide 42

Slide 42 text

Audience insight US Drivers (basically, the whole country) Enjoy driving, road trips, travel Don’t like getting stuck in traffic

Slide 43

Slide 43 text

Federal Highway Administration (FHWA)

Slide 44

Slide 44 text

Highway Performance Monitoring System A database of all the roads in the US detailing, among other things ● Location ● Length ● Volume of traffic

Slide 45

Slide 45 text

What can this data tell us? The busiest roads in America but also… The quietest roads in America

Slide 46

Slide 46 text

Turns out it’s a thing

Slide 47

Slide 47 text

Something people care about

Slide 48

Slide 48 text

Fairly simple analysis Extract data layer from a shapeﬁle Split state-by-state Estimate avg ﬂow of vehicles Low # of vehicles = quietest roads

Slide 49

Slide 49 text

Given our audience, something simple Scannable at a glance Minimal interactivity ++ Imagery Visualisation

Slide 50

Slide 50 text

PR: People like maps. Me: But this has no regional patterns to warrant a map visualisation. PR: People *really* like maps. Me: But maps aren’t mobile-friendly! PR: The client likes maps! Me: Fine... Actual conversation at the time

Slide 51

Slide 51 text

Visualisation

Slide 52

Slide 52 text

Visualisation

Slide 53

Slide 53 text

Visualisation

Slide 54

Slide 54 text

The Atlantic, Thrillist, Lonely Planet + 175 other news articles Thousands of social shares Brand awareness ++ Results

Slide 55

Slide 55 text

Why did it work? Brought data to the conversation Revealed something previously unknown Chose the audience-appropriate visual format

Slide 56

Slide 56 text

Learning along the way

Slide 57

Slide 57 text

Do the research Research your audience Figure out what’s important to them Ask yourself, how can you add insight/value to that? What data can you use?

Slide 58

Slide 58 text

Add insight with data Bringing *good* data to conversations that aren’t traditionally data-rich Adding new data to existing data-rich narratives

Slide 59

Slide 59 text

Add insight with visualisation Using visualisation to make data more accessible Dataviz doesn’t have to be complex to be effective

Slide 60

Slide 60 text

As long as we’re: Not lying with data Communicate something meaningful and compelling to our audiences

Slide 61

Slide 61 text

Thank you!

Slide 62

Slide 62 text

Sources and references: Paypal Mafia ● http://fortune.com/2007/11/13/paypal-mafia/ ● https://www.crunchbase.com/ ● https://www.nytimes.com/interactive/2014/06/20/sports/w orldcup/how-world-cup-players-are-connected.html ● https://gephi.org/ Quietest Roads ● https://www.fhwa.dot.gov/policyinformation/statistics.cfm ● https://www.thrillist.com/travel/nation/most-fun-stretche s-of-road-in-america ● https://www.outsideonline.com/1926186/americas-best-car- touring-roads ● https://www.r-project.org/ Sources and image credits Image attribution (where required): ● https://xkcd.com/1732/ ● marcinignac on Visualhunt.com / CC BY-NC-ND ● https://commons.wikimedia.org/wiki/File:Elon_Musk_Roy al_Society.jpg ● https://jamesaltucher.com/wp-content/uploads/2015/09/p aypal-mafia.jpg ● https://www.instagram.com/p/BUkUZh7DxOl/ ● https://www.informationisbeautifulawards.com/awards/20 16 ● http://datadrivenjournalism.net/ ● https://www.linsonbusinessconsulting.com/wp-content/u ploads/2017/09/Non-disclosure-Agreement-1024x675.jpg