Infographics: Telling Stories with Data

Visual.ly is a website that showcases thousands of infographics where data is displayed in unique and fun ways.  For consultants who are in the business of telling stories with data, it is worth a few minutes of your time.

Visual.ly - ScreenshotConsulting presentations are simpler.  90%+ of presentations will be on PowerPoint and only use simple graphs with very straight-forward messages.  It’s a good thing because executives look for compelling recommendations, not cutesy graphics.

Infographics - Simple graphs in Powerpoint

As I noted in another post, McKinsey & Company printed their client presentations in black and white prior to 2009.  For them, the strength was in the analysis and clarity of argument.  If “simple” is good enough for McKinsey, it is good enough for you and me.

So why bother? A cynic might ask, “So why should a consultant even bother looking at infographics, if they are not the type of graphics we use with clients?”

Answer: It’s good for you.  It’s easy for us to get set in your ways.  We need to feed our minds new ideas and new ways to improve our craft.  It’s instructive to see how professional graphic designers represent data – even if you apply it later in a more toned-down or straight-laced way.

1. Learn to build an argument.   In Osama vs. Air Travel: The Post 9-11 Effect on Travel, see how the designer combined data to show the impact of 9/11 on US travel patterns.

Infographics - Airline Bankruptcies

2. See how others analyze data.  You can take the most mundane data and do amazing things with it.  Look at this analysis of the Songs that Metallica Played on Stage during 1982-2012.  It’s amazing to see how many ways they sliced the 30 years of data.

Infographics - Metallica Songs Play on Stage

3. See good (yet crass) examples of consulting tools.  Many of the infographics use classic consulting tools like fishbone diagrams, venn diagrams, pie charts, waterfalls.  See this Taxonomy of Sh*t which is just an affinity diagram.    Albeit, a vulgar one.

Infographics - Taxonomy of Sht4. Learn from bad examples.  There are examples where the designers goes overboard and makes the data confusing.  In Texting While Driving, you can see a muti-layered pie chart that takes about 30 seconds to fully decipher.   Do not do this.

The main point is simple – that young people text while driving – but the graphic is unnecessarily ornate and confusing.   Also, it is bad graphics hygiene because the age groups are unequal: 16-19 (4 yrs), 20-24 (4 yrs), 25-34 (10 yrs), 35-44 (10 yrs)?

Infographics - Texting While Driving

5. Enjoy the creativity.  Live a little and enjoy the creativity.  Face it, data analysis, pivot tables, MiniTab, and Powerpoint can get boring.   Check out the How Much You Can Trust a Bearded Man? and see if you agree with their analysis.

Infographics - Trustworthiness of Beards

Another savant analyzed the color palette of 10 artists over the same 10 year period.   I swear, there are some creative people out there.  More right-brained than I am.

Visual.ly - 10 artists over 10 years

Let me know when you find great infographics, I will add it below.

Olympic Medals Measured by Population, By Productivity, By GDP

As a consultant, it always seemed odd that news reporters routinely compare the medal count between the USA and China – after all, the USA has 530 athletes compared to China’s 380.  Seems like the total number of medals is only one of many potential metrics.  For example, productivity (i.e., # of athletes needed to win a medal) would be a metric that levels the playing field among countries of different sizes.  Thinking along those lines, I smashed together data of the last 40 years and came up with this:

If you’re from Iceland, there is a good chance you are an Olympic Athlete. . .

Most selective Olympic team

In 2010, there were 319K people in Iceland and yet they sent 27 people to London for the Olympics.  Simple math tells us that they sent 1 athlete for every 11K people.  In fact, it is really easy to be an Icelandic Olympic athlete compared to the Bangladeshi.

I took the country’s 2010 population here and divided it by the number of athletes that each country sent to the London Olympics.   You can see that China sent 1 athlete to the Olympics for every 3.5M citizens.  The US sends 1 athlete for every 580K citizens.  It is easier to go to the Olympics as an American than a Chinese – less selective.

Olympics by Population 1972-2008If you are from Bangladesh, your chances of going to the Olympics as an athlete are slim.  Only 1 in 30 million Bangladeshi go to compete in the Olympics.  As a thought-experiment, if all the teams sent 1 athlete for every 30M people . . .

  • The US Olympic team would only send 10 people (not 580)
  • The total Olympics would only be 233 Athletes (not the 10,700+ it currently is)
  • Only 40 countries (over 30M population) would be competing

The most productive Olympic athletes are from . . . . Panama

Athlete productivity

Why does data surprise us?  Looking at all the medals won in 2008, and then dividing it by the number of athletes sent to Beijing by country, Panama wins the prize for the most productive athletes.   Panama will win 1 medal (over average) for every 3 athletes they send to the Olympics.  It is correlation, not causality – but it is still impressive.

Olympics Medals

  • South Africa won 1 medal in 2008 with 136 athletes
  • Burkina Faso also won 1 medal in 2008, with only 6 athletes

Of all the medals that China wins, more than 50% are gold

Gold as a percentage of medals

In a previous post, I discovered that only 32% of medals are gold (more than 36% are bronze).  When analyzing the medals over the last 40 years, the Chinese have stood out as a strong bunch.  51% of their medals are gold vs. 32% of US medals being gold.

Olympic Gold %

In terms of the GDP per athlete, the Indonesia was the highest

GDP per Athlete flag

Looking at the country-based GDP numbers here, Indonesia has $32B in GDP (2010) for every athlete sent to London 2012.  By comparison, every Tuvalu athlete symbolizes only $10M in GDP output.  Likewise, the US Olympic athlete symbolizes $27B in output compared to $4B represented by the British athlete.

GDP per Athlete

China keeps winning. . .

Looking at the current medal leaders in London, you can see that the United States (in white) has been consistently winning ~ 100 medals every summer Olympics except in the 1980 Olympics where there was a boycott.   The British  (in blue) have done well since 2000, but the Chinese are the stand outs.  They had no Olympic team in 1980, yet they had 100 medals in 2008.

Summer Olympic Medals US China UK

I am a patriotic American.  Huge believer in US constitution, freedom of speech, Chik-Fil-A sandwiches, our top-shelf graduate education system, Pixar movies, micro-brews, Teddy Roosevelt, National parks and NPR.  (Yes, I do wish the Economist was a US paper).  Love watching the women’s soccer (not football) team beat the Japanese, and the women’s volleyball finals US vs. US.  Michael Phelps is impressive too.

That said, we also need to give credit where credit is due.  The Chinese have been rocking the Olympics.  A balanced scored card of the Olympics might look like this.

Scorecard US China UK

China won 100 medals in 2008 summer Olympics which was a 59% increase over 2004.  Each Chinese athlete has a 1 in 3.5 million chance of making the team, but when they get to the Olympics ~ it takes 6.4 athletes to win a medal, and 51% of the time it is gold.

The Olympics are awesome to watch and one of the most democratic sporting events, if you are willing to measure medal counts a little bit differently.  Enjoy the games.  Go USA.

Note: The Huffington Post has also done an excellent job with this line of thinking – alternate ways to count medals.  Look here.

Data Analysis: the meaning behind the US state names

Happy Birthday!  The United States is celebrating its 236th birthday today.

I saw an article from Mental Floss Magazine by Matt Soniak called How All 50 States Got Their Names.  It was well-timed for July 4th, but a bit long at 5,000 words.  After reading through the alphabetical list A, B, C , I started getting frustrated by all the words.

Call me lazy, but I wanted to see the big picture.  Understand the narrative.  Dig out the insights.  So, I did what all consultants do when confronted with data. . .

1) Think about the logical groupings

  • I noticed that the state names really boiled down to either a) Native American Indian words or b) European (read: Spanish, French, English) words

2) Put the data into excel for analysis

  • I started the un-fun task of reading the words, and boiling it down into its parts
  • Here are the first 10 states (A-G) with the name origin and meaning

State Names and Meaning A-G

3) Look for anything interesting: trends, outliers, common threads

  • Consultants love maps.  I am no exception.  Tableau is my favorite for visually mapping data, but I found a free alternative at www.diymaps.net
  • 60% of states have names with Native American Indian word origins.  All 28 states shown below in blue trace their names back this way

States with Native American Indian Names4) Make it graphical if it helps to convey the point

  • I ran the map again for the states which names that have multiple name origins
  • Surprisingly, eleven states (shown in green below) have unknown name origins or multiple conflicting stories.  For example, New Mexico obviously comes from “Mexico”, but even that name has an unclear origin as described here.

States with Unclear Names

For those in the US . . . Do you know the meaning of your state’s name?  I didn’t

  • Here are the other 40 states, letter H – W.  This is good Jeopardy stuff
  • If there are “quote”, it is a Proper noun.  For example, Louisiana was named after King Louis.  Also, Virginia was named after Queen Elizabeth I (the “Virgin Queen”)
  • Some names have both Native American Indian & European origins; for example, Illinois and Michigan were Native American Indian words that were anglicized

State Names and Meaning H-W

How consultants do industry research

Industry Research - Trucks - Consulting blogManagement consultants need to be quick learners.  Junior analysts are routinely asked to support proposals and projects across different industries.  The good ones are fast, and proficient with Excel and PowerPoint.  The great ones get up-to-speed quickly on the industry dynamics and can add in industry specifics to the pitch.

So what happens if a partner asks you to help on a proposal for something you know  nothing about?  For example, how do you get smart on commercials trucks?

1) Start with Standard and Poor’s (S&P) industry surveys -  If you have access to these analyst reports through your work or university, start here.  It is loaded up with charts, tables and graphs to get you up-to-speed on the major trends, value chain, competitors, key financial drivers and earnings.  See a S&P survey on Trucks here.

Industry Research - S&P Survey on Heavy Trucks - Consulting blog
2) Download SEC filings and investor presentations - After finding out who the main industry players are at www.finviz.com, go to to the investor relations section of the website.  Download the 10-K and annual report.  Also, download the presentations made to investors and Wall Street analysts; it is a treasure trove of information on the company’s strategy, and marketing.
Example #1: PACCAR:  From this example, you can see that one manufacturer segments the customer base into six groups: 3 of them by size (large, medium, owner operator), 2 by ownership type (leasing vs. private fleet), and also the vocational segment.

Industry Research - PACCAR customer Mix - Consulting blogExample #2: Navistar: In this Q1 2012 earnings presentation, the management team shared their financial forecast, manufacturing strategy and market share by product type.  In the slide below, it shows a worsening market share trend, so you probably want to also listen to the webcast, where management answers tough questions from analysts.

Industry Research - Q1 2012 Navistar earnings - Consulting blog3) Search for industry trade organizations websites - Look at industry trade group websites to get a summary of trends, and regulatory issues.  These groups do an excellent job of simplifying the story for non-experts.  In the Google search box, simply add “site:org” to whatever search terms you use.

Google search trucking site org - Consulting blog
Here are some of the websites that came up from the search for “trucking” and “site:org”
4) Compare financial metrics – It always helps to compare the financials of different companies to see the 80/20 breakdown of revenue and profits for the industry.  Finviz is a great tool, but so is SmartMoney.  As you can see in the graphic below in gray color, Smart Money ranks the different companies by market capitalization, revenues, earnings, profit margin, dividends, ROA, ROE etc. See the example for Navistar (NAV) here.
SmartMoney.com
5) Don’t forget about blogs and forums - If you are looking for more qualitative information, it is useful to spend 15-20 minutes looking at industry blogs.  Just type in “blog” or “forum” with your targeted search term.  Then search the specific topic on the blog.  If is much better for you to learn the basics from an online blog community, rather than asking your client a question that shows your ignorance of the industry.

Linkedin - Consulting blog

6) Look at LinkedIn.com - Go to LinkedIn and see what current or previous employees are putting on their profiles and descriptions of projects.  In the past, I did a market sizing from the tidbits that marketing managers mistakenly left on their LinkedIn profiles.

7) Prepare for primary research
There is only so much that you can get online. Think through your questions,  then start reaching out to people you know via LinkedIn contacts or undergrad / grad school alumni.  Sometimes a 30 min conversation with a subject matter expert is worth 3 days of online research.   Potential interviews might be:
  • Suppliers or customers (up and down the value chain)
  • Distributors or dealership owners
  • Former employees
  • Industry researchers or professors

8-100) There are so many other ways to do industry research, but you gotta start somewhere.  Next steps might be: trade magazines, conferences, government and industry data sets, product reviews, surveys, focus groups, interviews.  As a consultant, I have bought my fair share of  – - – - for Dummies books.  You will too.