Great vs good analysts, software obsessions and informing executives

Small_packed_bubble_chart1.pngCharts and graphs to inform
Are your charts and graphs actually informing people?
Can I actually make a decision from your presentation?
If not, why should I or the executives care?

Cool software, but what about the business?
It's easier to obsess over the complex but oh so cool software features
than to truly understand the business issues and create substantive analysis.

Good analysts versus great analysts
Good analysts make data stores and useful analyses.
Great analysts speak the language of their audience & make informed recommendations.

Stephen McDaniel
Chief Data Scientist at Freakalytics

Honesty, the first principle of good analytics

 

ph-design-evolution-eichler-rams
Photo from Braun.com

Meet Dieter Rams – Industrial Designer

Dieter Rams is a German industrial designer most closely associated with the minimalist designs of the consumer brand Braun. Dieter was head of design at Braun for over 30 years, where he became famous for creating an austere aesthetic while focusing on user-friendliness. His philosophy is summed up in his saying, “Weniger, aber besser.” which translates into “Less, but better.” He has won many awards through the years including the World Design Medal and the Ikea Prize.

Dieter's impact reaches beyond his retirement as he is now impacting design in the 21st century, with a company widely considered a leader in technology design, Apple, acknowledging a debt to Dieter as inspiration for many of their design decisions. The Head of Design at Apple wrote, “Rams's work is beyond improvement... Rams's ability to bring form to a product so that it clearly, concisely and immediately communicates its meaning is remarkable.”

Braun-Apple-Link
From Braun, "90 years of history"

 

 

Ten principles of good design

As a prolific designer, Dieter formulated ten principles of good design. In this series of articles, I will adapt several of these principles for guidance in creating good analytics. I have selected the sixth principle of good design as the one I consider most important for good analytics.

 

The sixth principle of good design

6. Is honest - It does not make a product appear more innovative, powerful or valuable than it really is. It does not attempt to manipulate the consumer with promises that cannot be kept.

 

Freakalytics_princples_good_analytics_1_003_smallThe first principle of good analytics
Continue reading

Build a Basic Dashboard in Tableau 8
Excerpt from Rapid Graphs with Tableau 8

Freakalytics_Dashboard_Tableau_8_BasicAppendix #2—Build a Basic Dashboard

We’ve taught entire courses on how to design and build dashboards in Tableau over multiple days, so the topic can be quite complex. In this section, you will learn the basic functionality so you can get started. For more advanced dashboards, visit www.Freakalytics.com/examples.

Dashboard Audience
VP of Sales at a cheese maker that sells to the public and to gourmet retailers

Overall Objective of Dashboard
Sales updates for monthly review by Sales Vice President (VP)

The Sales VP has four questions:
1. What are sales by state?
2. What were sales by customer contact method in 2013 compared to 2012?
3. What are the actual sales by item versus the target sales?
4. What are the actual sales by customer contact method versus the target sales?

Download the Cheese Factory sales data packaged workbook from www.Freakalytics.com/rgts8d

Included in the workbook:
Data source contains two years of sales data for 2012 and 2013.
Four worksheets with views answering the four questions above.
The final dashboard is illustrated at the end of this appendix.

Build the dashboard
Open the workbook, and in any one of the worksheets, on the main menu, select Dashboard → New dashboard. In the Dashboard pane on the left side, double-click on each of the four worksheets in the numbered order. Also in the Dashboard pane, in the Dashboard Size section at the bottom, change Desktop to Automatic so the 4 views fit within the workspace. The worksheets are arranged in the order added and the legends for Q1 and Q2 are on the right.

Adjust containers

Dashboards are comprised of containers that contain the views, legends and filters and are outlined by solid lines when you click on them. First, move the legends to the left side of the dashboard. Continue reading

PowerTrip Analytics™ 2014
Candid Quadrants™
Based on three-year visual analytics cost estimates

PowerTrip_Analytics_Candid_Quadrants_2014_by_Freakalytics
Click for full page image

Detailed cost comparisons across all seven visual analytics offerings
at PowerTrip Analytics™ pricing review tables.

Read individual product details, highlights and examples:
Microstrategy Analytics Desktop,
SAP Lumira,
Tibco Spotfire,
Microsoft Power BI,
Tableau Desktop,
QlikView, and
SAS Visual Analytics.

Download data to predict gender using first name (US data)

Download US American first names and initials to predict gender sex 1Do you have data with just first names or even just first initials but no information on the person's gender/sex? If you would like better insights on your customers, based on whether they are likely male or female, then this data download is a great way to maximize your ROI! Download it today and begin using it to tailor your messaging and improve future communications.

There are three licenses available for this data- individual, corporate and corporate for multi-company consumers. The individual version is available free (with discount code) for a limited time. Simply select the Individual license for purchase and use discount code discfreepers at the checkout page- this will deduct $3.99 from your purchase price.

The primary table in this data download is First names by Freakalytics with 5164 rows (distinct names and common misspellings). You can use this data to guess if someone is a male or female based on their first name or find the probability that they are male or female based on their first name.

Here is the column information and simple summaries for this table:

Data Column Max Min Average Median Mode
Name mixed case Zulma Aaron N/A N/A James
Most likely gender Male Female N/A N/A Female
Rank Overall 4,019 1 2354 2397 4019
Male Probability 100% 0% 22% 0% 0%
Female Probability 100% 0% 78% 100% 100%
Count Either Gender 99,989 32 1,079 127 32
Male Count 99,671 0 524 0 0
Female Count 83,718 0 555 64 32
Male Probability Within 3.68% 0.00% 0.08% 0.01% 0.00%
Female Probability Within 2.92% 0.00% 0.02% 0.00% 0.00%
Male Rank 1,054 1 584 608 1,054
Female Rank 3,052 1 1,825 2,131 3,052
Name first initial Z A N/A N/A J
Name upper case ZULMA AARON N/A N/A JAMES

The top few rows from this table (as a snapshot of the data in Excel 2003 format and in text):

Download US American first names and initials to predict gender sex 1

Access this valuable data download here.

Accidental Analysts®: What are they doing with my data?

Eileen McDaniel, Ph.D. and Stephen McDaniel
This article was originally published in late 2013 in
The Data Warehouse Institute FlashPoint Newsletter

Earlier this year, we presented this topic in a talk to an independent group of data professionals. When we noticed it was mistakenly promoted as “Accidental Analysts: What are they doing TO my data?”, we had to laugh! Unintentional typo or not, we’ve found that data warehousing specialists often wonder what businesspeople are doing on their end. Who are accidental analysts, how do they analyze data, and what aspects of the data warehouse can data professionals evaluate and improve upon so that they are set up for success instead of frustration?

Who are accidental analysts? 

Accidental_Analyst_in_hurry_Many business analysts either lack formal education in data analysis or took courses that didn’t fully prepare them for the challenges of real-world data analysis. They are asked to quickly answer business questions so that managers, colleagues and clients are able to identify and implement a plan of action. After teaching analysts in many organizations from all skill levels and backgrounds, we discovered that a major obstacle to obtaining good results is that many are uncertain of the steps to take when analyzing their data. They need a plan of attack, regardless of the analysis software that they are using!

 

The Seven C’s of Data Analysis

The_7Cs_of_Data_Analysis_Copyright_Freakalytics_LLC_605_175

The scientific method has been used by scientists for hundreds of years to design and analyze experiments. In our training and books, we adapted this method to fit business analysis,

Read the rest of this article

Top 5 Stories Worth Reading in Data, Data Warehousing, Analytics & BI

February 23rd, 2014
Stephen McDaniel
Chief Data Officer Advisor at Freakalytics, LLC

Finding it hard to make time to keep up with the rapidly changing world of data, data warehousing, analytics, data science, business intelligence and visual analytics? We understand! Each week, I read through hundreds of stories in this space and share the five (okay, this week I couldn't stop and chose six) most worthwhile articles with you. Each article includes a summary and link to the full story. Please note that inclusion of an article does not indicate that I agree with every point in the article, but I at least find it thought-provoking and useful for informed debate.

 

 

Recruit Better Data Analysts

In the big data talent wars, most companies feel they’re losing. Marketing leaders are finding it difficult to acquire the right analytical talent. In the latest CMO Survey, only 3.4% senior marketers believe they have the right talent. Business-to-business companies have a bigger gap than business-to-consumer companies, as do companies with a lower percentage of their sales coming from the internet.  And yet analytic skill is a must for effective marketing.

Results indicate that companies with above-average marketing analytics talent experienced significantly greater rates of marketing return on investment (MROI) than companies with below average analytics talent…

 

Big Data doesn’t lie. Or does it?

If you’ve seen any indication that humans are getting smarter and more sophisticated, please inform me, and you don’t have to read what follows. For everyone else, who sees no lack of stupidity and misinformation in both business and public life, read on about the joys big data will bring.

The insights offered by analysis of big data are only as good as the human beings that create the data, gather and assemble it, decide what questions should be asked and how the data is presented. And interpreted, especially that.

I am very concerned that big data, misapplied and misunderstood, will create big lies. Or more likely a combination of lies and truths that prove very difficult to sort out.

 

What is Data Mining?

One way businesses can turn the information into something useful is through data mining. Data mining is a process used to analyze raw information to try and find useful patterns and trends in it.

"Data mining applications help users discover correlations and connections within large data sets," Software Advice writes on its website. "These might have gone unnoticed without these algorithms."

 

Scientific dashboard with periodic table of elements
Includes history, photos, great filters, discoverers and more

There is nothing we love more than sharing great examples of what is possible with visual analytics and dashboards. It’s one of the best ways to inspire new analysts and expand people’s horizons of the nearly infinite number of ways that visual analytics tools like Tibco Spotfire can be used.

Continue reading

Top News – Analysis & Commentary
Data, Data Warehousing, Analytics & BI

Five Business Intelligence Predictions
from Paxata for 2014

January 26th, 2014
Stephen McDaniel
Chief Data Officer Advisor at Freakalytics, LLC

Paxata logo blue 201401Finding it hard to make time to keep up with the rapidly changing world of data, data warehousing, analytics, data science, business intelligence and visual analytics? We understand! Here’s a top new story worth reading and that we considered noteworthy enough to add commentary and analysis by Freakalytics (in purple).  A summary of the article and excerpts that I comment on are in black.

In this commentary and analysis, we cover the growth of Tableau and QlikView, the opportunities that exist for Microsoft to disrupt the second-generation business intelligence market and how self-service data integration will likely make data scientists & data enthusiasts much more productive- enabling wide swathes of Accidental Analysts to quickly answer tactical business questions.

Five Business Intelligence Predictions For 2014 (from the CEO of Paxata)
Summary
The dust is finally beginning to clear from the big data explosion, which is a good thing. One of the problems with big data is that it’s been led by technology, not business requirements. And business requirements will be the focus in the 2014 business intelligence (BI) ecosphere—to enable enterprises to achieve results with data mining and analytics and to prove those results.

Stephen
I found this article a fascinating glimpse into the strategic thoughts of a CEO of a promising, second-generation, cloud-based data integration company- Paxata.

Continue reading

Top 5 Stories Worth Reading in
Data, Data Warehousing, Analytics & BI

January 26th, 2014
Stephen McDaniel
Chief Data Officer Advisor at Freakalytics, LLC

Finding it hard to make time to keep up with the rapidly changing world of data, data warehousing, analytics, data science, business intelligence and visual analytics? We understand! Each week, I read through hundreds of stories in this space and share the five most worthwhile articles with you. Each article includes a summary and link to the full story.

 

1-Netflix-analyticsHow Netflix Got Analytics Wrong, Then Right
Two entertaining and informative articles about Netflix illustrate how to be smart about using analytics. In one instance we see where Netflix went wrong, and in another we see Netflix doing the right thing.

Netflix launched a high-profile crowdsourcing project in 2006 to develop a better recommendation engine, offering a $1 million prize to any person or team who could improve Netflix recommendations by a modest 10 percent.

 

2-BI-dashboardSigns That Your BI Dashboard Needs a Comeback

Everyone loves a good comeback. Stories about celebrities like Robert Downey Jr. and Britney Spears climbing back to the top after falling so far capture our collective imagination. Movies like Rocky and Cinderella Man – about underdogs making a comeback – inspire us to think we ourselves can rebound from any setback.

Is your business intelligence dashboard the underdog at your organization? Dashboards have been around for decades, with some companies not putting the time and effort into updating them regularly to keep pace with the innovations in BI and the growing expectations of users.

 

3-analytics-movies"Pitch Perfect" And How Analytics Are Transforming Movie Marketing

When Universal released the cult musical film Pitch Perfect in 2012, they did what any self-respecting studio would do: They commissioned marketing reports and forecasted ticket sales for the Anna Kendrick-starring movie. Among them was an analysis by a company called Fizziology which data-mines social media to see how the film would play out with audiences. Continue reading

Top News – Data, Data Warehousing, Analytics & BI

December 30th, 2013
Stephen McDaniel
Chief Data Officer Advisor at Freakalytics, LLC

Finding it hard to make time to keep up with the rapidly changing world of data, data warehousing, analytics, data science, business intelligence and visual analytics? We understand! Here’s our curated summary of relevant news that could help with your future data and analytic projects. We also add commentary on the topic, a summary of the article (in orange) and the link to read the full article.

There are four articles in this update:
Why Do Forecasters Keep Forecasting?
2014 BI Outlook: Who's Hot, Who's Not
Where do you sound like you’re from? (How Y’all, Youse and You Guys Talk)
Aided by Data Analytics, Internal Auditors Dig Deep
 
 
 
 
 
01-01Why Do Forecasters Keep Forecasting?

I found this article quite interesting, a wide range of investment advisors missed the forecast for the growth of the S&P stock market index in 2013, all of them by a huge amount (a range of being off by -53% to -109%)!  This shows how incredibly difficult forecasting even one year out can be, in spite of these companies having entire teams of experts constantly studying every machination of the underlying data that affects the stock market including economic factors, political climate   and more. In fact, if forecasters at many retailers were off by so much on their overall company sales growth forecasts, they would likely be sacked or at least pummeled with a bag of oranges in the garage one dark evening!

Why is this article so important? Because there has been tremendous hype around the power of predictive analytics to steer the business. The reality is that forecasting is an incredibly hard job, regardless of the overall intelligence, toolset, access to data and experience of the team. Does this mean we should give up on forecasting and using advanced methods to predict future outcomes and behaviors? No! However, it does imply that forecasting should be fluid and adjusted, sometimes rapidly, in response to changing external factors.

All of this begs the question, how can someone do this? In my opinion, it points back to visual analytics, dedicated teams of business analysts with clear missions and good data management and warehouse practices with an agile approach, so that major misses can be caught, reviewed and corrected mid-course.

01-02-Freakalytics-Bad-Forecasts-Investment-Firms

I created my own summary dashboard of the growth forecasts (image above.) Imagine if you were a client of these firms and you were told in late 2012 that stocks would lose money in 2013 (Wells Fargo and UBS), so you buried it all back into low-yield bonds! In the words of Warren Buffett, “Buy when everyone else is selling.” Continue reading

Strategic analytics, forecasting & visualization for smarter choices & greater profits