Editor's note: this post has been edited to sort the comments by time and to point to the recorded webcast.

This is the post that we updated live during Tom Davenport's Competing on Analytics webcast on October 31. Once we get this out of our system, I promise we'll close the week with a few killer Excel tips.

In the meantime, here's some background reading.

10 Ways Not to Build an Analytics Based Business: "We disagree with quite a few of these points and even where we agree, we want add real-world nuance." Includes a rather sharp discussion in the comments about the relative value of centralized vs. decentralized analytics.

Neil Raden on Competing on Analytics : Neil Raden, an experienced consultant, gives his perspective on centralization of analytics

Tom Davenport's original Competing on Analytics Paper

----

Commentary begins:

12:59 (Chris) Just got an email confirmation from the AMA 1 minute before the presentation. That's "just in time." Bad news. Meebo seems to not be working.

1:04: Apparently this is to be a book. Preorders available.

1:08: Niel Raden points out that Tom seems over his head on this topic. Feels true in the intro--"Things just seem different" Tom says. Can't describe why.

1:09: Describing lots of "new" tools: conjoint analysis, CHAID, .... This feels more like one man's personal journey into learning new analytic techniques rather than a change in industries.

1:12: (Zach) My take so far: none of this is news. we could probably find a presentation from 20 years ago making these points

1:13: Meebo's back. (Chris) this historical review is killing me. I want to get to the more interesting issues of centralizing vs. decentralized analytics.

1:15: (Zach) If he was putting the spotlight on something that wasn't well known and fully within the conventional wisdom, then I would appreciate his efforts

1:16: (Chris) "BestBuy shifts from ready-fire-aim to ready-aim-fire" on more initiatives. Of course, BestBuy didn't call their previous strategy ready-fire-aim. I'd like to see ALL business writers be more careful with their word usements.

1:20: (Chris) Reviewing his list of top analytics competitors. The absolutely brutal counterargument to this list of companies is http://www.hiredbrains.com/Davenport_Rebuttal.htm. This shows "competing on analytics" companies have generally underperformed the market recently.

1:23: (Zach) This presentation is perfectly suited to a room through of wanna-be MBAs. It would give them a bunch of superficial anecdotes and catch-phrases. [ed. note: Zach has a MBA]

1:25: (Zach) I found a sentence I like: "Find your distinctive capability, and use analytics to support it"

1:27: (Zach/Chris): There's a real problem with how the sample set of companies for Tom's study is built. He's starting with a very limited understanding of analytics, it's a non-random sample (two of the teams are local sports teams he's interested in), the sample is too small (only two companies are in "Stage 1"), self-reported data. Tom is far too willing to draw conclusions from this weak base.

1:33: (Chris): Talking about the need for hardware.

1:35: (Chris): From the presentation: Is your senior management committed? If yes, "go full steam ahead", if no, "prove the value". How do you measure commitment? Most everyone pays at least lip service to the need of using data for analytics

1:37: (Chris): "data dog", mmm

1:38: (Jules): "If your organization prefers to make decisions by the gut...find a new CEO" ?!? It's hard to get rid of problem employees at the lowest levels. How are you going to slot in a new CEO? [ed. note: Jules is our HR guy]

1:39: (Zach): well, Jules, if he/she doesn't like analytics, I think that is all the ammunition you'll need

1:40: (Chris): A ladder to heaven. Campaign management is considered "higher up" than event-based triggers. Bzzt. I think most experienced folks people would reverse those. Anyone disagree?

1:45: (Jules): This is a wonderful company where you can just tell IT to go and do the stuff you want. Do they knock it out over the weekend or are they that talented that they only work 35 hours a week? Marketing and IT have always traditionally been had a hard time seeing eye to eye.

1:46: (All): Some things he's missing: IT challenges, getting culture-wide buy in, using the right tools, getting tools into the hands of decision-makers,

1:46: (Chris): An absolutely brutal slide appears. A X-Y graph with NO LABELS whatsoever. This is madeupware. Timestamp in the presentation: 49:18.

1:46: (Zach): last two slides are a disaster, a black hole of information

1:47: (Jules): Don't ghostwrite charts, Tom.

1:50: (Chris): Skeleton mouse?

1:53: (Chris): Q&A time. Where are these questions coming from? Wait-a-sec: I thought this was realtime? ;-)

1:56: (Zach): "no offense to any PhDs out there, but if you were cultivating your analytical skills you probably weren't developing your social skills" (paraphrase of Tom Davenport) Ohhhhhhh, snap!

1:58: (Jules): Do you think Anna is a bot? [ed. note: Double snap!]

1:59: (Chris): And we're clear. Thanks guys.

1:59: (Jules): Interesting. I downloaded the wrf file and it's timestamped Oct 25 at 5:06pm.

2:10: (Zach): Why do we have to be live when he isn't?

We've been critical of Tom Davenport's effort to define what it means to "Compete on Analytics". If you want to hear the argument from Tom directly, he has a webcast tomorrow (at 1pm Eastern, 10pm Pacific) where he will lay out his vision and research. To attend, click below.

Competing On Analytics: Move Faster, Accomplish More, and Avoid Mistakes by Learning From The Best

We're going to attend and we'll maintain a live post throughout his presentation giving a realtime commentary. We'll be available at that time on Meebo (the little chat box on our site) if you want to chat. So, keep one window on Tom and one window on Juice while you eat your lunch.

I've been trying to become better informed about the state of the business analytics industry. This has meant wading into the dreary swamp of sales deck webcasts. Here's are a few thoughts on how to drain Sales Deck Swamp:

Respect your audience

Too many presentations ask that you attend a LIVE webcast at a specific time on a specific day. What's more, they require pre-registration. Not all of us are willing to stick our heads that close to the maw of the sales-lion.

You can respect your audience by letting people watch on their own time on their own machine. The archive of the presentation IS the presentation. Make the archives of your webcasts prominently available, searchable, and historically accessible. Give people an e-mail address, or better yet, a open forum or comment system where they can ask questions.

Many webcasts are glorified sales presentations. If you accept that a webcast is a cheaper way of doing an initial sales meeting then the wisdom of "respect your audience" is clear. You wouldn't tell a prospective client that if they want to see your sales presentation, they need to clear time on Monday at 1:00 pm. No, you'd let them find a time that worked for them. And you wouldn't require that everyone who attended the meeting "register" and give you their name, address, phone number, and role at their company. You'd be happy if a few extra people dropped by. If the SVP of Marketing happens to drop in, so much the better.

Show a tool, not a sales presentation

If you can, show the product in use solving a real problem, rather than a canned series of screenshots. Jon Udell's Screening Room is a great example of how to this can work.

Showing your product being used by a skilled user in a real situation helps me imagine how it could solve my problems. It also allows you to convey tacit knowledge—ways of working with your product that advanced users know but they can't really transmit in words.

Show your face

Google has made available a series of in-house lectures on various technologies at Google University. These lectures aren't presented in a sophisticated way, but to an analytics geek they're pure gold. Do I want to hear Guido Van Rossum talk about the next generation of Python or Richard Hipp break down why he built SQLite? I do.

The videos are simple, presenter at a lectern, slides on the wall behind them. Yet they've been viewed thousands of times.

Would I like to hear a down to earth discussion about Cognos' report builder or Business Objects infrastructure plans or Crystal XCelsius talking about integration with Excel 2007? It would be a pleasure next to some of the webinars I've seen.

The Washington Post had an article last week entitled The Top Pickers vs. The Pack describing web sites that attempt to discover "experts" through investment or sports picking...

A small number of Web sites seeking to turn the wisdom of the Internet on its head by sifting through its vast number of users to identify a handful of experts...the wisdom of the crowd could be outsmarted by what Michael Arrington, editor of the TechCrunch blog, recently dubbed the "wisdom of the few." Sites like PicksPal rely on input from the masses chiefly as a venue for auditioning prospective experts, on the theory that these virtuosos could provide even more accurate information and predictions than the crowd.

The author Alan Sipress appears excited by the potential of this approach, though he does allow for the possibility that this "phenomena" may be an illusion ("Some business professors remain skeptical, warning that luck can often be mistaken for expertise.") Meanwhile, Tom Jessiman, the founder of PicksPal, breathlessly remarks: "A very small set of people keep winning," Jessiman said. "It kind of blew me away."

Before we crown a gaggle of new geniuses, we thought it would be interested to simulate the results of a site like PicksPal if only luck was at play. PicksPal has a user base of 100,000 and identifies the 30 most successful players over the previous five weeks. ("Jessiman said he tries to rule out flukes at PicksPal by requiring that his experts play actively in three of the most recent five weeks."). What those top 30 experts would look like if they were entirely based on luck? To find out, we ran a simple Monte Carlo simulation for 13 games picked a week for three weeks (39 total games) with a 50% chance of winning each game.

The results: The top 30 winners predicted almost 80% of the games correctly, boasting records of 30-9 or better. You can see why it would be tempting to wrap these leading pickers in plastic and take them to Vegas. But if you did, you'd find their streak would end in a hurry.

There's something rhythmically delicious about the haiku. It's a highbrow equivalent of a 3 frame cartoon as your thought needs to be a tight, right, and an exact fit. Purists like Chris (hokku anybody?) want to work in a reference to a kigo but in the great American spirit of borrowing and localizing ideas, I'll just march off in my own co-opted, local direction.

This morning at Juice HQ we stumbled on Wired's Very Short Stories and immediately fell in love. A lovely change of direction from the Haiku? Yes, the beloved Haiku makes you compose phonetically and maybe drum your fingers at the same time. The Very Short Story pulls your brain in a slightly different direction. How can the geeky amongst us not dig that?

So, which is it to be?

Economy: Controlled by giant balloon men?

or

Giant baloon men

control all the world's money?

Look behind, Japan!

(Original story at Spiegel and inspiration by this infographic.)

prairieFyre Software, a provider of contact center solutions, has created a reporting tool that takes a table of data and encrypts it in chart form. The original numbers and trends are virtually unrecoverable. Congratulations, prairieFyre, for this exciting new approach. This may be patentable, but I'm afraid there is prior art.

Beat that, Junk Charts.

Luc Girardin of Macrofocus contacted us in response to our post "When Will Survey Analysis Grow Up?" to point us to their SurveyVisualizer analysis tool. I had a little time this weekend to download and play with this application. There is a lot to like.

SurveyVisualizer is designed for surveys that have a hierachical or tree structure. Luc describes the relevant data structure in a background paper about the product:

The questions—also called quality criteria—are then aggregated into 23 quality dimensions (e.g. network quality, ticketing, cleanliness, security, reliability). They represent the level of satisfaction with a whole group of questions pertaining to a particular issue. The quality dimensions themselves are further aggregated into three different customer satisfaction indices, reflecting the different areas of responsibility.

The free download has multiple satisfaction "criteria" (e.g. friendliness of crew) roll up to "dimensions" (e.g. cabin crew) which fall under "indices" (e.g. index of flight services). This may be an appropriate structure for a satisfaction survey—but it isn't one I've encountered before.

Despite this limitation, the analysis capabilities delivered by SurveyVisualizer are intuitive and innovative. For example, all your survey data is displayed at once in a kind of relational map. This lets users visually identify patterns in the full set of results. Each of the vertical hashes represents a question or roll-up of questions. Clicking on any one of these hashes highlights the hierarchical relationships. The "ghost" lines represent the results across questions for a multitude of dimensions or respondent types.

Users have the ability to select specific dimensions to identify patterns in the corresponding results. An easy-to-use interface lets you choose a dimension then apply a color to the line within the relational map.

Also, users can click on individual display lines to investigate the results (e.g. I wonder who had that particularly crappy score for flight delays?)

If your analysis requirements don't fit this particular structure, Macrofocus has a more general-purpose tool called InfoScope.

PivotTables seem like a great tool to use to build reports—except once you start building the report you get a huge stack of metrics like the one shown below.

This doesn't help people pick out natural groups of metrics and doesn't match the formatting people expect. We're going to show a simple PivotTable hack today that lets you create blank lines between metrics. This allows a report using PivotTables to look a lot more professional.

To get started, find the PivotTable menu, and add a calculated field.

Call the Calculated Field "Blank Line" and make the formula 1/0.

Now drop the "Blank Line" formula in to the data area of your PivotTable. You can drop it in multiple times to create several blank lines.

Go to PivotTable Options (right click on the PivotTable and choose Table Options..."). Turn on the "For error values, show:" check box. This will display an empty cell instead of all the alarming #DIV/0! messages.

After a little more prettification, the table looks like this.

A total of 5 blank lines were added. Three were given titles and bolded: "Revenue and Cost", "Warranty Details", and "Profit." Two are blank. To make a completely empty row, change the title from "Sum of Blank Line" to a single space key " ". Every title in the PivotTable must be unique, so the title of the other empty row is two spacebar keys.

Welcome, Jules! Here are a few more sources for tech haiku. Error messages: pain rendered as poetry.

A file that big?

It might be very useful.

But now it is gone.

Dratz offers the occasional Thursday haiku.

Roads covered with ice,

Winds blowing danger my way.

Telecommute day.

Lastly, here's another Juice haiku.

Wet October sky,

Alt E D R cleaning sees

Scrubbing fingers dance.

Human kind is doomed.

The earth looks like a grapefruit.

Segmented downfall?

(The full article is here.)

At the 2004TED conference, Malcolm Gladwell tells the story of Dr. Howard Moskowitz, a man who revolutionized the prepared food industry through a new kind of analytical thinking. Long story short: Dr. Moskowitz was one of the first people to argue that companies should pursue multiple products targeted at customer subsegments rather than try to create the perfect product for all customers. He realized that an attempt to create a "platonic ideal" —whether it was pickles, mustard, or pasta sauce—would be a suboptimal result for most consumers. Consumers are individuals with preferences that are better clustered than averaged. Mr. Gladwell states that this change in business thinking (spurred by Moskowitz's study of pasta sauce) mirrors a more general scientific shift from a focus on universal truths to the study of variation.

The prepared food industry gets it—as evidenced by nine variations of Ragu sauce on the grocery shelves—but I'm not convinced that these lessons have permeated the rest of the business analytics landscape. In particular, I am struck by the inability of most survey analyses to reveal insights about respondents.

The tools may be part of the problem. Here's an example of what WebSurveyor provides its users to help them analyze online surveys:

Their site tells us:

"Each question is graphed independently allowing you greater flexibility in customizing the layout of reporting for each question...Filter results based on specific responses or cross-tabulate results from two different questions, giving you powerful tools for detailed analysis."

Powerful? Flexible? More like barebones. WebSurveyor is putting the analyst in a very constrained box that won't help deliver an better understanding of respondents. WebSurveyor's tool demands "question-centric" not "customer-centric" analysis.

Consider how this typical survey approach would serve you in an effort to understand the passengers of Noah's Ark. A surveyor would ask each animal to fill out basic information about their height, weight, number of legs, food preference, etc. The results would then let us know that the average animal weights 23 pounds, has a height of 1.2 feet, 5.6 legs, 30% omnivore and so on. All of which would miss the essential insight about the animals on board: there are two of each.

Unfortunately, the kind of analysis needed to reveal personality / needs / behavior clusters in your respondent population isn't well supported by out-of-the-box analytical tools. One approach is factor analysis—a statistical technique that is used in marketing to "identify the salient attributes consumers use to evaluate products in a category" (Wikipedia). Another approach is to examine individual visual representations of individual respondents—a technique that we term (rather clumsily): customer flashcards.

A new dramality series features NetFlix super-model Cinematch that gives real people the opportunity to prove they can make it in the high-stress, high-stakes world of super-predictive modeling. The contest follows a group of data nerds of various backgrounds, shapes and sizes, who vie for a $1 million prize. (Thanks America's Top Model)

NetFlix, the online DVD rental service, has put together a contest that is the analytics equivalent of the X-prize. Finally a chance for data junkies to step out of their windowless offices onto the proverbial catwalk.

The contest asks participants to try to improve the accuracy of the company's existing Cinematch recommendation engine by 10%. According to the NetFlix prize web site, the Cinematch recommendation engine is designed to:

"...predict whether someone will enjoy a movie based on how much they liked or disliked other movies. We use those predictions to make presonal movie recommendations based on each customer's unique tastes."

Entering is simple: sign up your "team" and download the anonymized data set of NetFlix users and their ratings by movie. From what I read on the forums, the data set is actually quite shallow, so the initial thinking is that an improved system will require linking to additional data sources like IMDB.com to add richness to the modeling.

If you work with data for a living, the following sites are worth a visit (or a subscription) to learn from some of the best, most passionate practitioners. While we're at it, we are handing out our first annual Juicy Awards celebrating contributions to the analytics community. The categories are:

1. Charts and graphs

2. Information visualization

3. Presentations

4. Excel

5. Data analysis

6. Business intelligence

(4 through 6 are covered in part 2)

Here is an OPML file containing the finalists if you want to subscribe to these feeds with a single click.

Juice Resources for Analysts.opml

***

Charts and Graphs

These sites are dedicated to teaching the best (and pointing out the worst) approaches for displaying data in business environments. And the Juicy goes to (darn... this... envelope) Junk Charts for "recycling chart junk as junk art."

Best of the rest:

• Stephen Few: "Thoughts about how visual representations of data and visual interaction techniques can be used in practical ways to analyze and communicate business information." We like Steve for his tireless campaign to improve data visualization in a business environment as well as hard-hitting critiques of the established business intelligence vendors.
• Edward Tufte's Ask E.T. Forum: The Don of Data Display. The Patriarch of Presentation. The Vicar of Visualization. The forum has a lot of great discussion about data presentation and charting.
• Communication Nation: "I believe we will be happier and more productive if we can improve our ability to communicate. This blog is dedicated to that effort"
• Stat. Graphics and Data Vis.: Straddling the fence between our charting and info visualization categories, this blog appears to have fallen fallow. On the plus side, it offers one of my favorite examples of a terrible 3D chart.
• Indexed: Sometimes a hand-drawn chart is just what is needed to tell the story.

Information visualization

Thse sites show off the best of the world of "info vis." This is where art and design take starring roles, pushing aside the straightforward data display of traditional charts and graphs. Winner of the Juicy: Information Aesthetics for Form Follows Data. A daily dose of information visualization to spark your creativity.

Best of the rest:

• Info Vis Wiki: Community platform with a plethora of links and resources. As is typical of this category: a ton of information presented beautifully is still a ton of information.
• Visual Complexity: A gallery of visualization examples with particular focus on ways to display complex, multi-dimensional networks.
• NiXLOG Infographics: "A running collection of links to infographics found on the web."
• Datapacifica: Tuned in to the latest examples of data visualization. [Edit: Jay Jakosky, author of Who Was Where, writes in to tell us that the blog has changed names and locations to Datapacifica from Who Was Where. The link and OMPL have been changed.]
• Visuale: Pointing readers to examples of cool infovis. Bonus points for using the famous Napoleon goes to Moscow graphic in the blog logo.

Presentation

Now you've got to put all your data together to tell a full story. The sites below give advice, tips, and guidelines for constructing and delivering your presentation. (We left out the bevy of PowerPoint tips blogs in part because of the ugliness of their presentation...the irony does not escape us).

The Juicy goes to Presentation Zen in a runaway victory. Garr Reynolds offers tips, tricks, and examples for making great presentations. An excerpt: "Edward Tufte says: "PowerPoint is Evil." This got me thinking... What if Darth Vader — my favorite fictional bad guy — gave a formal presentation? How would it look? How would it compare to the presentation style of Yoda, the wise Jedi master?"

Best of the rest:

• MasterViews: Lively blog focused on PowerPoint.
• Signum sine tinnitu: Guy Kawaski, author of The Art of the Start, frequently writes about presenting and presentations.
• Maniactive: A fun blog that touches on both PowerPoint and approaches to presenting.
• Tony Ramos: A rambling list of links and thoughts that will keep you up on all that is new in PowerPoint, presentations, and presentation design.