Erica Driver

Data Needs Poetry

Posted by Erica Driver Dec 27, 2012

I recently watched a TEDxUtrecht video, “Data by Itself Is Not Enough, Data Needs Poetry,” by digital artist Brendan Dawes. How could I not, with that title?!

Two main points from his talk stuck with me:

  • We entice serendipity by going non-linear. Brendan Dawes gave the example of looking for information on the web. Normally, browsing the web is relatively linear process. You go to a search engine and put in some words and get some results, or you click links on a navigation bar. The typical web search interface does not provide a great way to just bump into things. So Dawes made his own experimental way to surf the web in hopes of inviting serendipity and discovery, called Doodlebuzz. Doodlebuzz is a chaotic draw-to-explore system to help people discover things they didn’t even know they were looking for. “DoodleBuzz,” the tool’s website says, “was born out of an idea to create an entirely new way of exploring information - one that allows for a kind of 'quiet chaos' that gives people the opportunity to explore unthought of paths and connections along their news gathering journey . . .” The user types in a search term and then draws a shape on the screen with their mouse. Search results are ordered along the shape. They can explore, following any line of inquiry that seems interesting, by drawing more lines or shapes.
  • We remember the unexpected. In this talk, Dawes attributed a quote to structural engineer and designer Cecil Balmond with regard to the Pedro e Inês footbridge in Portugal: “We wanted to break the traditional, continuous sight lines of a bridge to create a structure that provokes exploration and questioning of accepted practices and methods.” This footbridge makes an unexpected bend in the middle (see image below). Dawes said, “When you do something that goes against the unexpected, then beautiful things can happen. . . And that’s also what people remember, as well . . .  They are used to the expected.” A bridge with a curve, instead of straight across, can lead a pedestrian to wonder who they are going to bump into as they go around that curve. It makes them think differently and provides a memorable experience.

Pedro e Ines footbridge.png

These ideas are beautifully relevant for Business Discovery. QlikView Business Discovery apps invite a non-linear experience. Users can explore the data any way they want to – up, down, or sideways. With each click or tap in their app – or by back-tracking and removing selections they have made – they generate new on-the-fly views of their data. With each step along the way, they expose relationships in the data and derive insights they may never have thought to pursue directly. It is in the unexpected nature of discovery that delight awaits.

What happens when you search for answers in dramatically changing circumstances? Often, you will find significant new questions. In this video, QlikTech’s VP of Product Management Donald Farmer describes how a global cosmetics company used QlikView to examine its supply chain after the terrible tsunami in 2010, and was able to mitigate significant risk to the company.

 

 

As decision makers at the cosmetics company began to explore the data, they made a discovery: a small Japanese supplier of a highly specialized ingredient had been severely impacted by the tsunami. This ingredient was due to be used in a major product launch which would now be impossible. A big part of the discovery came from seeing the associations between the product launch and a large impending advertising spend. Using QlikView, the brand manager was able to take action and cancel the ad campaign.

Want to hear more amazing tales of Business Discovery? Check out these other videos with Donald Farmer:

Sometimes I get questions from QlikView evangelists asking for tips on measuring the business value of QlikView’s social and collaboration capabilities. I came across a Forrester Research report that does a great job doing just that: the November, 2012 report, “Mapping the Value of Social Business and Collaboration” (available to subscribers or for purchase).

Forrester - sample chart showing business processes on social and collaboration value-viability matrix.jpg

Decisions are social events. As a result, BI and collaboration are inseparable. This philosophy, in combination with empowered consumers bringing social networking sensibilities into the workplace, led us to bring social and collaboration capabilities into QlikView. With QlikView, information workers co-create, communicate, and explore together by:

  • Building and sharing charts and graphs–or entire apps
  • Modifying apps together while you’re online at the same time
  • Discussing an app’s progress throughout development, right there in context
  • Sharing analysis with bookmarks, which preserve your selections—recipients can pick up where you left off
  • Authoring threaded, in-context notes and comments
  • Exploring QlikView with others via co-browsing sessions, for a fast path to decision making
  • Interacting with QlikView in the context of enterprise portals and collaboration platforms.

(For more information about QlikView’s social and collaboration capabilities see the Social Business Discovery page on our web site.)

Forrester’s recommendation to IT executives looking to measure the business value of social business and collaboration is to focus on specific business processes where information workers need better access to information, easier access to expertise, and the ability to take collective action. Forrester’s model for assessing the value of investments in social business and collaboration comprises three steps:

  1. Assess business value factors. For each business process you are thinking about addressing with social and collaboration capabilities, quantify these factors: “What percentage improvement do we expect in the process?” “Does the process directly increase revenue, reduce expense, or improve customer experience?” “Are key participants highly compensated?” and “What is the level of risk that this won’t work?” Assign each of these factors a score of 1 to 5.
  2. Score viability factors. Likewise, for each business process you are thinking about addressing with social and collaboration capabilities, quantify these factors: “How difficult will it be to get people to work differently?” “Are worker goals and objectives aligned with the success of the initiative?” “Are business process owners on board and ready to address potential change?” and “Will existing processes or systems need to be re-architected?” As with the business value factors, assign each of these factors a score of 1 to 5.
  3. Plot the business processes on a chart. Plot the business processes (e.g., sales, marketing, training, HR, operations, manufacturing, new product development, or customer service) on a chart with business value of the investment in social business and collaboration on the horizontal (X) axis and viability on the vertical (Y) axis. You’ll see business processes that have the highest likelihood of being improved quickly with investments in social business and collaboration on the upper right side of the chart.

Have you tried this approach to deploying social and collaboration capabilities in your organization—in particular, the social Business Discovery capabilities of QlikView? It seems like a simple and straightforward way to assess where to prioritize your efforts. I’d like to hear your thoughts.

As a kid I really enjoyed playing video games (and still do!). Almost as enjoyable was to watch other people play. What made it so enjoyable was to share their emotions and experiences as they tried to beat the game. These experiences often turned into elaborate stories that we shared over and over again amongst ourselves and with other friends.

In my job as a user experience researcher in QlikTech’s product management team, I have the privilege of going through a similar process on a regular basis. I get to listen to and share people’s experiences and feelings about using QlikView. I get to learn what their jobs are like in general. They tell me about what they are trying to achieve, how they try to achieve it, and why (or why not) QlikView was helpful during this process. These are passionate people who hold many valuable experiences. Us understanding them is fundamental to “QlikView.next” (the code name for the next generation of the QlikView Business Discovery platform) fully supporting their needs and expectations.

Blog1 Lovquist.jpg

To me, QlikView.next is not about technology and features. It is about addressing people’s needs in the best way possible. I strive to ensure that our designs help people achieve things in their working lives effectively and efficiently. Understanding peoples’ abilities and expectations helps me to guide our designers to create experiences that are pleasant and enjoyable and at the same time fit people’s working lives. We want to maximize peoples’ achievements and minimize their efforts in reaching their goals with Business Discovery using QlikView.next.

I’m very excited to be a part of the QlikView.next team. Our driving force is designing for real people’s needs. Stay tuned – I’ll be writing more about our people-centric product design approach here on the Business Discovery blog.

Yesterday we announced QlikView 11.2. The main new capability in this release is QlikView Direct Discovery. Since we announced this new capability in October, we have seen great excitement in our customer and partner community as they see the potential of QlikView 11.2 enabling Business Discovery with Big Data, without any data size limitations.

Capture.JPG

Today organizations have the capability to store and process more data faster and cheaper than ever before. Currently, Big Data sources such as Hadoop are thought of as a sort of “staging environment on steroids”—a place to stage and dump a massive amount of stuff.

But as I talk with our customers about Big Data, I keep asking myself, is the more the better? The fact that IT can now store petabytes of data does not mean much to business users unless the business users can benefit from greater relevance. Until business users can seamlessly analyze Big Data along with their other decision support data, asking and answering their own streams of questions (even questions no one anticipated they might have—without having to go back to an expert for a new report or a new visualization), the more does not bring any value to them.

With a single QlikView app, even the least technically savvy person can answer question after question after question, moving along their own path to insight. Users are empowered to explore information freely by clicking on field values, doing searches, and more.

The Direct Discovery capability of QlikView 11.2 brings the same unique Business Discovery capabilities to analysis on Big Data. QlikView Direct Discovery is a hybrid approach that leverages both in-memory data and data that is dynamically queried from an external source. QlikView maintains all the associations in the data, regardless of where the data is stored.

For example, a policy analyst who uses a QlikView app to analyze regional loss and revenue information on a daily basis can now also see the policy-level claim payments for billions of policies (stored in a Big Data system) in the same QlikView app. He does not need to remember policy numbers, he only needs to select the regional info, as he would do every day, and QlikView displays policy-level information from Big Data sources in the same app as in-memory metrics.

With QlikView 11.2, QlikView once again conquers the hearts of business users and IT pros alike—this time by making Big Data user friendly. QlikView Direct Discovery makes IT organizations Big Data heroes. With the rapid and easy development of QlikView Direct Discovery, IT can now make Big Data sources available in existing QlikView apps in mere days and let business users incorporate that data into their Business Discovery. In addition, with QlikView’s Data Governance Dashboard, they can see how the existing enterprise data is being used with the new big data sources. This knowledge of high use data will allow IT people and data administrators to build a more comprehensive enterprise data roadmap. They can easily identify the usage patterns, metrics definitions within the QlikView environment and decide how to position big data and enrich their existing enterprise data sources with big data if necessary.     

The value of the Big Data will not be explored by the pre-determined path of questions.  With Big Data, the business users need to follow the information scent, and be able to ask and answer streams of their own questions to find what is relevant to them. QlikView 11.2 opens a new area for QlikView where the unique Business Discovery capabilities will enable users follow the information scent and chase the value from the Big Data.

When I hear stories like one my colleague Paul Van Siclen told me recently, it sets my mind at ease about the doom and gloom of the impending data analytics skills shortage you hear about everywhere.* Paul is a Director of Market Development-Financial Services (Americas) at QlikTech. "I always try to find new uses for QlikView,” Paul said, “so when my 12-year-old son said he was doing a science experiment my ears perked up. Science experiments are all about data."

A middle school science project.png

The purpose of this science experiment was to see what color vehicle tends to go the fastest; the 7th grader had put his money on black. As the student was preparing to capture the data, his dad guided him to ensure he set up a table in Excel that had field names at the top so the data could be brought into QlikView easily. The young scientist tracked how long it took for the vehicles to pass between two points on the street. He also recorded each vehicle’s color and whether or not the vehicle was coming out of a turn.

After the data collection was done, Paul and his son loaded the data into QlikView and in 15 minutes had a simple app including a couple of basic formulas, like count and average speed. They popped up a couple of charts and went on to dinner, happy with their progress.

When Paul and his son returned to the desk after dinner, they realized that they hadn’t saved their work. AAAARGH!!! They had to do it all over again. But the second time around, Paul’s son created the app on his own with very little help from his dad.

I asked Paul how technically savvy his son is and he said, “He’s your typical 12-year-old. He has some Microsoft Word experience, a little bit of Excel experience, and much more video game experience.” He found that teaching his son how to create an app in QlikView was easier than it would have been to teach him how to use Excel. “You are thinking differently, with QlikView, and it's easier. Getting started and producing meaningful results happens very quickly."

Now, with his app created—and saved—Paul’s son was able to explore the data any way he wanted. He explored it many ways; for example, looking for patterns in sports utility vehicles vs. trucks vs. cars. And insights he found. The youngster’s experiment proved his hypothesis correct: black vehicles did drive faster, on average, then other vehicles.

However, his research turned up an unexpected surprise, which he was only able to uncover because of the ability to click through his data, following the “information scent.” He found that the average speed of black vehicles was equal to the average speed of other vehicles, but only when not coming out of a turn. (They were faster than other vehicles on a straightaway.) Discovering this important fact without QlikView would not have been feasible! 

In the conclusion of his research paper, the student wrote, “This experiment could help society because traffic control may be able to know who is more likely to speed (or accelerate faster). Also, this could help vehicle insurance, vehicle manufacturers, and/or state and local governments.” Right on! Oh: and he got an A+.

* See these articles about the looming data analytics skills shortage:

In my previous articles on the Business Discovery Blog, “Metadata Management the QlikView Way” and “Overcome Challenges with the QlikView Governance Dashboard,” I described metadata’s importance as well as two types of metadata that QlikView and QlikView Expressor handle.

 

I recently talked to a large QlikView customer in the manufacturing industry about its QlikView data governance practices and challenges. The company uses QlikView companywide to empower business users to ask and answer their own pic1_valid.pngstreams of questions. The company’s Business Intelligence Competency Center (BICC) has mandated that any new development must observe a set of policies.

For example, the BICC established a policy of using only approved and conformed column definitions across the QlikView estate. Column definitions can contain quantitative values (metrics) or textual values (dimensions). Each label must use pre-approved descriptive names that can be easily recognized by company decision makers.

 

To support this effort, the BICC created their own version of a metadata repository that stores specific metadata characteristics ranging from column names and data types to expressions and validation flags. As business teams develop new KPIs, a QlikView administrator enters values for metadata into a database, giving particular care to column names and validation fields. For example, if the policy requires that company income should be referred to as “revenue,” the admin enters “revenue” along with the validation flag of “Y” for yes. Other possible variations such as revenues, sales, income, earnings, etc. are also entered and flagged as “N” for no.

The Challenge: Reconciliation

The BICC encountered a challenge when they wanted to reconcile this data across all the labels already used in the existing QlikView applications. They had no means to analyze the entire QlikView environment and did not have time to build a custom solution. In addition, they needed to examine what QlikView documents, data files, columns, and expressions were or were not compliant in order to take the appropriate actions.

The Solution: QlikView Expressor and the QlikView Governance Dashboard

I suggested a QlikView Expressor solution that would leverage the data already created by the QlikView Governance Dashboard app. A byproduct of the QlikView Governance Dashboard scanner is the creation of data files specifically used by the QlikView Governance Dashboard app. These data files are available to be used in conjunction with this customer’s defined metadata to facilitate the solution they were looking for. The customer was very pleased with this approach as QlikView Expressor performed both the reconciliation of the incoming data and the data preparation to be used by QlikView. Stay tuned for part two of this article, where I will dive deeper into the technical details of the solution using QlikView and QlikView Expressor Desktop.

 

Michael Tarallo

Senior Product Marketing Manager
QlikTech

Erica Driver

The Way of the Genius

Posted by Erica Driver Dec 4, 2012

About 12 years ago I went to Lithuania with my father to visit the village where he was born. We drove through Kantwainai—a rural area populated primarily by cows and storks. My dad wanted to find the farmhouse where he was born.

Kantvainai map.JPG

As we drove around the village, my father recognized buildings that had once been the blacksmith shop and the kindergarten. We got out and took pictures at the tiny hill where he went sledding as a little kid. We tracked and backtracked, going back to the center of the village to try a different route, only to find that we were driving back and forth on roads we’d already tried. After a while, we drew up a map and tried to mark out the streets and landmarks, to avoid redundant efforts.

Unfortunately, it didn’t work; we never did find the house. Though when we met up with relatives later we learned that we had probably driven past the house a dozen times; my father just hadn’t recognized it because it had gained an addition or two over the years, had different siding, and was much closer to the now-paved road than he remembered. Perfectly understandable considering that he left Lithuania as a six-year-old and had not been back in 60 years.

Little did I know, this behavior—exploring, then backtracking to explore previously disregarded paths—is an indicator of genius. Genius! In my case it was more like desperation. But in an interesting article in the November/December 2012 issue of Scientific American, Dean Keith Simonton wrote about BVSR (blind variation and selective retention) and how it is common to people who exhibit “outstanding creativity,” or genius.

In “The Science of Genius,” Dean Keith Simonton described how geniuses try a bunch of things—they generate “a variety of ideas, one or more of which turn out to be useless.” When one thing doesn’t work, they “return to an earlier approach after blindly going off in the wrong direction.” “Exploring the wrong track,” he wrote, “obliges a return to options that had been originally cast aside.” Simonton wrote, “. . . a genius hunts widely—almost blindly—for a solution to a problem, exploring dead ends and backtracking repeatedly before arriving at the ideal answer.” This is BVSR.

This article struck a chord with me not only because it reminded me of the search I did with my dad, but because it puts words to a way of exploring for which the QlikView Business Discovery platform is perfectly suited. The explorer sits down with their data, brought together from multiple sources, and what happens? They come up with a question. And another question. And a question after that. With QlikView, explorers:

  • Choose their own paths. With QlikView, each selection users make in an app (e.g., clicking or tapping a list box, or lassoing a section of a chart of map), they go further down a path of inquiry. For example, if in my sales management app I select the year 2012 and the product category “bikes,” I see which data in my app is associated with bikes sold in 2012—and which data is not. The data that is not associated with these selections is highlighted in gray. It is in the visible nature of these associations that insight, discovery, and pure genius lurk.
  • Start anywhere, go anywhere. In QlikView, my selections are cumulative. I can refine my exploration to see just red bikes sold in Canada in 2012, for example, or can change bikes to components. If I reach a dead end in my exploration, I can hit the back button to backtrack, or can clear out all my selections to get back to the beginning with a single click. (See the related blog post, “100,000 Questions.”). I can practice BVSR.

Simonton wrote, “Domain expertise, such as the knowledge of advanced problem-solving strategies, supports thinking that is routine, even algorithmic—it does not inherently lead to the generation of novel, useful and surprising ideas. Something else must permit a person to go beyond tradition and training to reach the summit of genius.” Simply put, “a creator must try out ideas that might fail before hitting on a breakthrough.” With QlikView, the focus is empowering the user to find their own genius.

Thanks to Paul Van Siclen, QlikTech Director of Market Development-Financial Services (Americas), for sending this Scientific American article in my direction.

Filter Blog

By date:
By tag: