Web Analytics Demystified

How To Tell A Story with Your Data

Published by John Lovett on December 29, 2009.

« Back to all posts Share, Save or Email

A few weeks ago, my business partner Eric and I attended a basketball game in Minnesota. Eric purchased the tickets a few days ahead of time and I really didn’t have any expectations going into the game except to have a great time. Much to my surprise, our seats were incredible! We were sitting immediately behind the announcer’s table in the first row. Now, keep in mind, I’m a Boston sports guy and even when the Celtics were struggling through the 90’s and the early part of this decade, you still couldn’t get a seat behind the announcer’s table or anywhere near the first row without taking out a second mortgage on your house. But, this was Minnesota and the Timberwolves are not necessarily a big market team.

Anyway, as we enjoyed the game we struck up a conversation with the woman sitting immediately in front of us who was a coordinator for the announcers. Sitting on either side of her were two official NBA scorers recording all the action into their computers and generating reports at nearly ten-minute intervals. These reports were printed and handed to the announcers, which ended up in a big pile on their desks in front of them. After a while our friendly coordinator began handing Eric and I her extra copy of these Official Scorer’s Reports. So, like any good Web Analysts would do we took a look and gave the report a critical review (see the image below).

Scoresheet2

We were astounded by how poorly constructed the reports were. Sure, they contained all the critical information on each player like minutes played, field goals, field goal attempts and total points. Yet, there were no indicators of which metrics were moving, who was playing exceptionally well, or even shooting percentages for individual players. The announcers were undoubtedly skilled at their jobs, because these reports did nothing (or at least very little) to inform them of what to say to their television audiences. Clearly the NBA could benefit from some help from @pimpmyreports.

So, here is where I get to the point about telling a story with your data. Sometime during the middle of the fourth quarter a young aspiring sportscaster came running down to the announcer’s row and handed off a stack of paper that offered some new information. Finally! His 4th-Quarternotes recap was the first written analysis we’d seen that actually placed the statistics and metrics recorded during the game into meaningful context (see image below). The 4th-Quarternotes showed that:

  • A win could bring the T’wolves to 3-3 in their last six games.

  • Al Jefferson was having a good night – approaching a career milestone for rebounds – and posting his 9th double-double of the season.

  • Rookie, Jonny Flynn was about to post his first double-double (which only five rookie players have accomplished), needing only one more assist.

  • Ryan Gomes was once again nearing a 20 point game with a 58.6% field goal percentage in the past five games.

FourthQnotes

This method of reporting used all of the same data that was contained within the Official Scorer’s Report but added historical context, which really brought the data to life. This was interesting stuff! Now T’wolves fans and casual observers alike could understand the significance of Jefferson’s 16 points and 28:27 minutes on the floor – or that Jonny Flynn needed just one more assist to achieve a significant feat. After reading this, (even as a Boston sports fan) I was invested in the game and had something to root for – Go Flynn!

So here’s the moral of the story:

  • If you’re going to produce generic reports with no visual cues – do not show them to anyone because they won’t use them – and make sure you hire some damn good analysts that can interpret these reports and give a play-by-play.

  • If you do want to distribute your reports widely – take the time to format them in a way that highlights important metrics and calls attention to what’s meaningful so that recipients can interpret them on their own.

  • And most importantly – place your data and metrics in context given historical knowledge; significant accomplishments; or some other method to bring the data to life. Give your executives and business stakeholders something to cheer about!

Finally, if you ever have an opportunity to sit behind the announcer’s table, make sure you befriend the coordinator so you can get a copy of the reports for yourself.

About John Lovett

John Lovett is a Senior Partner at Web Analytics Demystified, Inc. and the author of Social Media Metrics Secrets (Wiley, 2011). A former Forrester Research Analyst and current President of the Digital Analytics Association, John blogs about web analytics industry trends, strategy, business culture, and social analytics.

Want to speak with John? Contact Web Analytics Demystified

John Lovett's Blog at Web Analytics Demystified

Posted Tuesday, December 29th, 2009 | 14 responses | Share, Save or Email


  • Pingback: associationjam.org

  • http://www.mattlillig.blogspot.com Matt Lillig

    Great post John! I love using sports analogies for explaining analytics! The light bulb seems to come on for a lot of people when explained in a sports manner.

    I also used basketball in one of my posts…

    http://mattlillig.blogspot.com/2008/12/yahoo-assistsput-money-back-into-your.html

    Matt

  • http://john.webanalyticsdemystified.com John

    Hi Matt – I do find that sports analogies often help to drive home analytics examples (excuse the terrible pun ;-) Thanks for your comment.

    @All – BTW – I forgot to add the postscript that @AlexBrazil turned me on to a great site for basketball scoresheets. Check out http://www.popcornmachine.net for a great use of combined metrics to form a “help value” score that provides depth to a player’s stats much in the way Matt describes in his attribution example.

  • http://www.waomarketing.com/blog Jacques Warren

    Hi John

    This reminds us all that an analysis is a *narrative*, a story; analysts need to exploit this deeply wired trait of human nature, that stories fascinate us.

    BTW, the Canadiens will kick the Bruins ass this year!

  • http://john.webanalyticsdemystified.com John

    @Jacques – you’re correct indeed. Everyone loves a good story and a big part of the Web Analysts job is to tell that story to their internal teams, executives and stakeholders – so that they “get it”. Although it’s no easy task.

    Regarding the hockey, you Canadians may dominate on the ice, but I’ll match my towns team against yours on the baseball diamond anytime ;-)

    Cheers,
    John

  • Joy

    Great post, John! It’s great to see how data and clear analysis can be applied to all sorts of online, offline, no-line data :)

    Seeing the bullet list is a great improvement over just the numbers, but I want to challenge all of us to look one step further. Let’s ask, “How does the announcer find the appropriate bullet point at the time it’s needed?” If the announcer is talking about Flynn, is it easier to look at a chart with the players’ names listed and then the stats number, or to look at each bullet point to find Flynn’s name? How can we merge the data to be extremely easy to find AND in a friendly format?

    If we merge the two types of reports, we can come up with some creative ways to display the information. For example, we could group each bullet point by category (player, points scored, rookies, etc.). We could have automatic formatting to create sentences, based on historical trending. We could even have alerts when a player needs 10 or less points to break a record.

    It’s great to see data brought to people in “people terms” (as opposed to data terms), and I think we can take your 3 bullets for the “moral of the story” and apply them across many industries, media, etc.
    Thanks again for the wonderful post, and keep em coming!

  • Jing Suk

    Hi John, what a great sports analogy! I have much more respect for the sports announcers’ job now. Personally I think this is a key challenge for analysts — the ability to find and tell a good story that’s backed by data. Have you come across any good blogs or books that touch more on this area?

    Thank you,
    Jing

    • http://john.webanalyticsdemystified.com John

      Hi Jing, these stories are indeed tough to come by. But you are in luck because I’m working with Webtrends on a white paper that will set out to tell these very stories. Stay tuned!

  • Pingback: Social, mobile and other bright, shiny objects | Retail: Shaken Not Stirred

  • http://www.goecshop.com affliction

    Even if you have less ambitious aims, the lesson applies. Stories are effective in communicating meaning locked in data. So show me the numbers, but do it in a way that will bridge the insight gap. Use an age-old approach: Tell me a story.

  • http://www.purchasembt.com mbt

    Good article, I was moved

  • jon

    You guys should read moneyball. A great read even for someone who knows nothing about baseball. The definitive sport and analytics book

  • http://john.webanalyticsdemystified.com John

    @jon ~ Moneyball is one of my favorite books! As a die-hard Red Sox fan it’s hard not to love that book.

    But have you read Michael Lewis’ newest…the Big Short? It’s about the subprime mortgage debacle and all the players that were prescient enough to see it coming. There are some fantastic parallels in there to web analytics and predictive models with the jist being that models are created by people who can visualize them and therefore are inherently flawed. That was my takeaway at least.

    Thanks for your comment.

  • Pingback: Monocle Smile: Old Spice Case Study « fang found this

 
COPYRIGHT © 2011 WEB ANALYTICS DEMYSTIFIED, INC. ALL RIGHTS RESERVED. PRIVACY POLICY