John Sands

The Key to Associations

Posted by John Sands May 9, 2013

I sometimes get asked to explain what QlikTech means when it talks about association.  Here’s how I explain it…


Too often we are compelled to adapt the way we think to the way software tools work.


This is the case with many business intelligence products, which optimise their performance by forcing us down linear drill down paths. But our brains don’t work in that way.  Our brains work by association.


For example, if we lose our car keys we don’t work through a pre-set drill down path. We don’t think to ourselves, “Hmm.  Let’s see, I was on planet Earth, in Europe, in the UK, in Hampshire, in Portsmouth, in Southsea, on Nelson road, at number 7 and in the kitchen.  Are my keys in there?  No, they’re not!”   After failing we’d then have to start all over again down another pre-set path.  If our brains were like most BI tools we’d probably have to wait for our consciousness to establish a new drill path!


In truth, what we do naturally, is think “What was I doing before I lost my keys? I remember; I was making a sandwich in the kitchen to eat while watching TV on the sofa.  A-ha! There are my keys down the back of the cushion!”  We find or make insights by associating non-linear data points.   (At the same time you might find the remote control you lost or if you’re really lucky some loose change, which you didn’t expect).  Providential discoveries rarely (if ever) occur through over structured thought processes.


We are used to being able to search freely for data quickly across billions of items on the web.  Many users of traditional BI wonder why their BI isn’t as simple as this.  Why do we have to change the way we discover things in everyday life just to suit a set of rules set down by a BI architecture detached from our  working reality?


To give a real life example of this I was trying to book a flight for my family this year somewhere warm and dry  (I live in the UK!) but on a budget. I went to the website of a well-known airline and selected my dates and destinations and then followed the path defined by the website. By the time I got to the 10th step and added all the taxes and baggage costs it worked out to be too expensive.  Wouldn’t it be a much more pleasurable experience if I could enter my budget and dates and automatically get shown the destinations that were available for me? Or a country and budget and see what dates were associated and available with those selections?


Life shouldn’t be that difficult.  We should all be able to just follow the associations.

Gartner recently published a new research note called ‘Market Trends: The Collision of Data Discovery and Business Intelligence Will Cause Destruction.’ (Data discovery is Gartner’s term for Business Discovery.)
A helpful team of operatives stood by to produce reports and make changes to the old BI platform’s semantic model…

The Gartner report sets out two possible scenarios:


1) "Data Discovery Becomes a Feature." In this scenario, Business Discovery would become subsumed in broad BI platforms as traditional BI vendors copy or buy lookalike functionality.


2) "Data Discovery: The Leading New Analytic Architecture". According to the report, "In this scenario, data discovery muscles out and replaces the BI platform for a majority of analytical/diagnostic use cases."


In this second scenario Gartner highlights a key shift in priorities "Whereas previously, users were looking to have a portfolio of: (1) a BI platform (semantic layers providing the end-to-end spectrum of reporting, ad hoc query and OLAP functionality); (2) data discovery in a tactical fashion, adopted individually or in workgroups; and in some cases (3) statistical/predictive analysis tools; the selection criteria has shifted to look like this: (1) data discovery (now the central norm for analysis on all data sources, and where most BI budget goes); (2) production and external reporting tools for systems of record; and (3) additional statistical/predictive analysis capabilities."


Business Discovery is now being viewed by industry analysts as a viable mainstream alternative to traditional BI, which should prompt thinking about the investments organizations are making to meet their analytic needs. Do they want to continue to spend the majority of their BI budget on top-down reporting led approaches, or switch over to user driven discovery and diagnosis? For many QlikView customers, Business Discovery is already the "Leading New Analytic Architecture."


Gartner gives no indication as to which scenario it expects to prevail in the long term.


It’s obvious what QlikTech believes will happen.


What do you think?

Every year, QlikTech executes several strategic initiatives. These are cross-functional programs that are funded out of a special budget and are blessed with executive sponsorship from the highest levels of the organization. An example of a QlikTech strategic initiative is QlikMarket, which we launched last year. One of QlikTech’s strategic initiatives for 2013 is called Customer Success.

Customer Success strategic initiative - four pillars.png

This initiative is very exciting for me on a couple of levels:

  • Focusing on the customer first is mandatory. Based on recent research from Temkin Group, most B2B (business to business) companies – whether they are in technology, financial services, aerospace and defense, or any other industry – are still mastering the basics of customer experience.* If in your role at work you deal with people from other companies, you probably see evidence of this yourself. But we live in a world where business customers have high expectations, often set by what they’re used to getting as consumers. B2C companies know that customer experience drives business loyalty. In my personal view, the software companies that will be around in 15 years are the ones that deliver a fabulous customer experience.
  • Your success is my passion. After more than 2½ years in QlikTech’s product marketing group, I have moved into a new role at QlikTech heading up a function called Customer Advocacy. The Customer Advocacy team is focused on measurement and accountability. As you might expect for a company that makes analytics software, we are firm believers that what you can’t measure, you don’t do. So we are launching quarterly customer relationship surveys and putting processes in place to act on your input. We expect the first survey to launch in May. If you see one come your way, please take a few moments to share your open and straightforward feedback with us.

So the focus of my posts here on the Business Discovery Blog will change. Rather than writing about Business Discovery and trends in the BI market, I look forward to sharing updates with you about the Customer Success initiative. Stay tuned.

* See “Best Practices in B2B Customer Experience,” Temkin Group, April 2013 (by subscription or for purchase).

I’ve been an advocate of the consumerization of BI for a number of years. Business Discovery platforms like QlikView that embody consumer oriented characteristics are more acceptable to a broad range of people, and so see higher adoption. At core, the consumerization of technology is about empowering people to do what they need, themselves. To be able to do so any technology must possess the three key consumerization attributes of speed, usability and relevance. (For more on these see Jeff Boehm’s 2011 post on my definitional research note whilst at Gartner).


However, driving back from taking my children to school on Tuesday morning I heard a discussion, on BBC Radio 4’s Today programme about the 2013 UK Design Awards, that brought home to me that there’s a fourth attribute of consumerization, as important as the others: aesthetics. For some people how something looks or feels will overcome the three other consumerization attributes, no matter how strongly they’re made available. It’s worth noting that while usability and aestethics are closely related they’re not the same. Something can be very usable, but not look pleasing. I suppose that’s the difference between satisfaction and delight.


During an interview on the radio show Professor Josh Silver, of the Centre for Vision in the Developing World (CVDW), described the Child ViSion self-adjustable glasses his organization is distributing through schools. The glasses have fluid filled lenses that are adjusted using pre-attached syringes. 



The glasses are a powerful example of the consumerization of an established technology:

• They’re fast; adjusting the lenses to correct someone’s vision takes a few minutes at most. 

• They’re usable; each glasses wearer self serves by fitting their own specs – no optometrist needed. 

• They’re relevant, directly to the individual life chances of the person that gets a pair, and broadly to the estimated 60 million young people that suffer from uncorrected myopia in the developing world. 


Crucially though (and unlike earlier versions) the glasses are designed with aesthetics in mind. Their shape is pleasing. They come in a variety of colors. The syringes on the arms used to alter the focal length are easily removable. Why does this matter? Because the target group for these specs is 12 to 18 years old! Image conscious teenagers the world over won’t wear something that doesn’t look good, no matter how useful it is. I’ve personal experience of this – when I was at school the UK National Health Service (NHS) provided one style of glasses frame in black for children. Nowadays these NHS specs, my 16 year old son tells me, are the height of ‘geek chic’. Back then, kids would do anything not to wear them, even if they couldn’t see the teacher’s writing on the blackboard from their desk. That’s why my short sightedness went untreated until I learnt to drive…


We can learn lessons from the development of CVDW’s inspirational project. 


It’s obvious that QlikView embodies the three attributes of speed, usability and relevance directly in how the technology works. The fourth one, the crucial aesthetic element, is largely up to the people who design the QlikView apps. It’s up to you to appeal to the consumers in your organization through the pleasing design of the apps you make available. Beautifully designed apps, styled to the users in question – a web site analytics apps for marketers should embody a different aesthetic to one for financial analysts – mean higher adoption, better perception of value delivered, greater return and more questions answered. The opposite means a less focused outcome.


A final thought on the Child ViSion glasses project: it’s a massively disruptive technology and extremely cost effective compared to the long established way of doing things. Sound familiar?

In the Harvard Business Review blog article by senior editor Scott Berinato, “Your Business Needs Insight, Not Just Pretty Pictures” (March 19, 2013), the author identifies an important trend in business communication. “Data comes first,” he wrote, “and it’s increasingly visual.” 

Berinato describes the two drivers behind this trend as Big Data and the democratization of tools for creating good data visualizations. I agree about Big Data. But I would describe the democratization trend a little bit differently; it’s not just about broader availability of and access to tools for creating good data visualizations – because as the title of the HBR blog post says, your business needs insight, not just pretty pictures.

An app includes multiple ways of interacting with data.png

What’s even more important here is the democratization of easy-to-use, interactive apps that anyone people can use to quickly and easily ask and answer the next question that comes to mind, and the question after that, without having to create a new visualization or report . . . without having to create anything at all.

By its very nature, a data visualization can answer only a limited number of questions. In contrast, a Business Discovery app provides multiple ways for a user to interact with information. It provides many different data visualizations – charts, graphs, list boxes (the most basic object in a QlikView app), tables and much, more more.

With each click, tap, or lasso, users can always see what data is associated with their selections and – importantly – what data is not. They make a selection in one chart or graph and all the other charts and graphs in the app update instantaneously to reflect the new selection. It is this rich, straightforward interactivity, with a full data set behind it (often drawn from multiple back-end systems), that empowers users to discover insights in their data.

I recently read an interview on the Harvard Business Review blog with Amanda Cox, who is a trained statistician and the graphics editor at the New York Times. I appreciate the perspective she shared with author Scott Berinato.

Power of Visualization.PNG

A few key takeaways:

  • Data visualization improves storytelling. “Most everyone here [at the New York Times] would agree the best way to tell some stories is through data,” Cox said. “Some think very rarely, some think most of the time, but they would concede telling the story with data is accepted.”
  • The primary data visualization skill needed: ability to ask good questions. I often hear people talk about data scientists. I find the term intimidating; it implies that to be able to create excellent data visualizations you need a PhD. I’m sure that’s true for some types of data analysis – but Cox contends that the primary skill needed to create good visualizations is the ability to ask good questions. Right on.
  • Visualizations for visualizations’ sake really don’t matter all that much. Data visualizations only matter when you can do something useful with them – when they are actionable. “There's an ‘Aha!’ moment sometimes,” Cox said. “Even on the most obvious things.”

Data analysis is becoming a bigger part of more peoples’ lives. Amanda Cox mentioned in the interview that the people she works with are journalists, biologists, urban planners, and mapmakers. If I think about my own role in customer advocacy (and formerly product marketing), I am similar—I come from a background as an industry analyst, not a statistician. No matter what our profession or background, we can all benefit from data visualization – a key aspect of data discovery software – to help us tell compelling stories, ask and answer more questions, and take the right actions.

Sometimes you have to practice what you preach. Last month I wrote and delivered the keynote at Gartner’s European BI & Analytics Summit, during which I challenged the audience to make some important decisions, and to choose a new path for BI. Well, I’m doing exactly that, by joining QlikTech.


Two paths in a wood.jpg


Some people have asked me, “Why go and work for a software vendor?” Well, I enjoyed my six years at Gartner. The analysts are great, and Gartner’s research remains fiercely independent and objective – you might even say that some of the analysts take pride in their robust dealings with vendors. (One analyst even said to me that if my first Magic Quadrant didn’t get escalated to Gartner’s ombudsman arbitration service by at least one vendor, then I wasn’t doing it right!)


But although I enjoyed being a critic, I recalled the excitement of being an actor. From my time in software development, and later at Hyperion, I remembered the excitement of helping to create and market technology, realized that I missed it, and so decided to go back to working for a software vendor.  But there’s more to it than that – it’s also the realization, made endlessly clear to me when talking to organizations about their BI strategies for Gartner, that the balance of power in the world of BI and analytics is about to tip, and I want to be part of the force that makes it do so.


As a follow up question – and knowing that I’ve had frequent dealings with almost all the BI vendors in the last few years – people then ask, “Why QlikTech?” I answer them with a question: “Do you remember what the BI market was like before QlikTech began to disrupt it?” The market was growing in revenue terms, but moribund in many other ways. It was seriously lacking innovation and was more known for failed projects than successful implementations. QlikTech changed that: It’s a rare thing to find a company and product that completely shakes up a market to the point where the existing market share leaders have to begin emulating its approach, and a bunch of new vendors emerge to follow its lead.


That’s all good, but QlikTech must continue to change the market; it remains a fact that too many people are still working with BI that doesn’t deliver, and they need to revisit how they do it. But for QlikTech to get as many organizations as possible make such shift, it will need to change, too, growing its platform capabilities and extending its reach. The company’s vision for “” (the code name for the next generation of QlikView) sets out how it intends to do so.  QlikTech’s plans are bold, and key in why I chose to join the company and to contribute my energies to transforming how people do analytics through discovery.


So, that’s why I’m here. I’m hoping, in the words of the Green Day song, that the move is “something unpredictable but in the end right.” I hope to have the time of my life.




Image attribution: (Creative Commons ShareAlike license.)

Analysts at Forrester and Gartner are seeing a rise in adoption of enterprise app stores. As Forrester looks ahead a few years, they see corporate app stores moving beyond distribution of corporate-approved mobile apps to provide content sharing, granular discovery, provisioning, and reporting and monitoring services. Forrester goes so far as to predict that app stores will become the primary way for individuals to obtain applications.* Gartner predicts that by 2016, 60% of enterprise app stores will be primarily composed of third-party apps rather than enterprise-developed apps and that by 2017, 25% of enterprises will have an enterprise app store for managing corporate-sanctioned apps on PCs and mobile devices.**

Cheese processing factory.jpg

I talked about this recently with Donald Farmer, our VP Product Management. A concept that’s on his mind when thinking about enterprise app stores is the information supply chain. How will all these apps be able to deliver the information users need so they can ask and answer streams of questions on their own?

In a traditional environment IT would build, own, and manage the entire experience of information, from sources to analysis and visualization. But in the supply chain model, an enterprise data warehouse is only one element of the information landscape, with many peripheral apps extending and augmenting it. IT seeks to publish as much data as possible through APIs, feeds, and other mechanisms – even reports. While some of this infrastructure requires IT support and maintenance, in a supply chain model, the goal is always to provide users with direct access and do-it-yourself tools wherever possible. (See the related blog post, “The IT Supply Chain: Making the IT Store Concept Work,” October 15, 2012 and the CITO Research white paper, “Putting the IT Store Ecosystem into Action.”)

Enterprise app stores and the information supply chain are first nature to QlikTech. Our customers have been calling their QlikView creations apps for a long time: lightweight, purpose-built, task specific applications. Our customers make these apps available to users via an internal web-based “store.” Today we call the place where users can search for and discover QlikView apps AccessPoint. In “,” the new AccessPoint will be QlikView itself. (See the blog post, “Compulsive Collaboration with ‘’ – Many Ways to Make Music Together.”) And as we set ourselves up to support customers deploying IT stores, we’re paying close attention to the information supply chain.



* See the Forrester Research report, “Build A Corporate App Store Into Your Corporate Mobility Strategy,” January 16, 2013 (available to Forrester subscribers or for purchase).

** See the Gartner reports, “Enterprise App Stores Can Increase the ROI of the App Portfolio,” February 4, 2013 and “There’s an App for That: The Growth of Enterprise Application Stores,” September 7, 2012 (available to Gartner subscribers).

“Learning is not a place, it’s an activity.” This is my favorite quote from a TEDTalk by Andreas Schleicher titled “Use data to build better schools.” Schleicher is the Deputy Director for Education and Skills and Special Advisor on Education Policy to the Secretary-General of the Organisation for Economic Co-operation and Development (OECD).

Image for TED video about using data to build better schools.PNG

This talk was about PISA, the OECD’s program for international assessment of 15-year-old students around the world. PISA studies education investments and outcomes and conducts international comparisons. The latest PISA assessment (2009) measured 74 school systems that covered 87% of the economy. The study measures skills directly, not whether students can reproduce what they learned in school. It “measures whether students are prepared for change,” Schleicher said.

This is a video worth watching for several reasons:

  • Great data storytelling. Schleicher used data to tell the story of how countries compare on education metrics and what the trends look like over time. He shared findings on spending per student, how countries spend their money (e.g., hours of schooling and teacher compensation), and value for the money. He augmented the data charts with specific examples that brought the story to life. For example, 9 out of 10 students in Japan believe that excellence in math is a result of their own investment and effort. In contrast, students in North America believe that to be good at math you have to be born a genius.
  • The transformational power of data. To be clear, it’s not the data alone that’s valuable; it’s the sum of the collection, analysis, and communication of the data. Schleicher gave an example of this power in Germany. In 2000, Germany’s scores on the PISA were low compared to other countries. This was a shock to many. For months the public debate centered on education. Policy makers got involved, the federal government increased its investment in education, and the beliefs people held about education began to change. Nine years later, Germany’s scores improved significantly.
  • It’s about one of the most important issues facing humanity. In case you don’t get a chance to watch the video, I’ll share with you some of the characteristics that high-performing school systems share, according to PISA. They place a high value on education. They believe all children are capable of success. (As examples, today every young Korean finishes high school. And in Finland there is only a 5% performance variation across education systems in that country; there, success is systemic.) And they invest in educators. They are careful how they select, recruit, and train teachers, and how they structure teacher pay. They provide an environment for teachers to collaborate, focus on growth pathways for teachers, and provide room for teachers and principals to be inventive.

Find this article and TEDTalk interesting? Check out the related blog post, “QlikView Is Playing a Part in Education Reform,” about how FirstLine Schools is using analytics to measure student and school performance and success and lighten the burden on administrators.

Who doesn’t like pizza? In every office where I’ve worked, free pizza is a great motivator. From a cost-benefit perspective, free pizza every day might be an effective employee incentive!

What does pizza have to do with Big Data? A recent white paper published by CITO Research and sponsored by QlikTech showcases many transformative uses of Big Data by businesses, including a large pizza

The pizza chain, like most retailers, needed to understand what products customers were buying and optimize their product mix to maximize profitability. Menu items that were unprofitable needed to be eliminated to make room for profitable items (which hopefully included one of my favorite pizzas - the Hawaiian, which is topped with ham and pineapple).

They faced the same challenges many retailers face when it comes to analyzing data – a complex organization comprising the corporate entity and franchisees, and the need to present the data appropriately, whether to the company board or a franchise manager.

With QlikView, the pizza chain was able to accomplish some great feats:

  • Analyze 57 million transactions
  • Consolidate data from 35 separate sources
  • Cover 500 pizza sales outlets across multiple venue types
  • Allow franchisees to view relevant datasets in a central location over a secure browser interface

What I found most impressive about the story is that this QlikView app took two developers only ten working days to complete.

With other BI platforms out there, after ten days, the developers would probably be still designing the requirements for a data warehouse just to consolidate all the data. The project would have taken months or years to complete and cost the company hundreds of free pizzas to motivate the developers. 

I think fundamentally this is why BI developers are so passionate about QlikView. They can deliver incredibly useful apps within days to business users. Not only does that make them the hero, it frees them up to tackle other interesting business problems.

The moral of the story? With QlikView, companies could offer free pizzas to their employees and still be profitable!

I recently read a New York Times article called “What Data Can’t Do” by columnist David Brooks. The gist of the article is that while data can yield important insights to drive better decisions, it’s not the only input. He gave an example of the chief executive of a large bank that had to decide whether to pull out of Italy. While the data showed a series of downside scenarios, the executive decided not to pull out of Italy based on other, non-data-related criteria: the relationship, trust, and values.

This really resonated with me. Holistic decision making relies on multiple sources of input, some quantitative (hard numbers, GPS info from mobile devices) and some qualitative (e.g., others’ opinions, observations, questions, and ideas — sometimes gleaned while out “on location”). Conversation and collaboration, as well as indicators and information from the world around us, help create the context around data and drive better decision making.

People make decisions based on multiple sources of input.png

With QlikView we embrace this reality that people make decisions based on multiple sources of input. Users collaborate on creation of analytic apps and can define and answer their own questions―in formal or informal groups. They communicate with each other to collaboratively explore data, forge paths to discovery and insight, and arrive at decisions.

How? For asynchronous collaborative analysis, when people can’t be in the same place or online at the same time they can initiate and participate in threaded discussions or send each other bookmarks that retain the selections they have made. When they are available at the same time but can’t be in the same place they can interact simultaneously with an app using shared sessions. People who are creating QlikView apps can send the entire app to others, who can pick up where they left off on development or analysis. This is just the beginning of collaborative Business Discovery. Click here to learn more.

This week, InformationWeek executive editor Doug Henschen published an article about the 2013 Gartner Magic Quadrant for Business Intelligence and Analytics Platforms, which came out in early February. (You can read the report in its entirety here.) We’re proud to be positioned in Gartner’s Leaders quadrant for the third consecutive year!

Gartner MQ.png

InformationWeek quoted Gartner analyst and report co-author Kurt Schlegel as saying, "Almost every user organization I talk to now is looking at making data discovery a more significant part of their BI and analytic platform architecture.” He commented that the benefit is improved agility because business users are freed to explore data and find new insights without having to put in requests to IT for new cubes or reports. QlikTech CEO Lars Björk was quoted in the article as saying, “Others are now confirming that [data discovery] is where the puck is moving, and it's a great testament that we're in the right place.”

The most exciting part of this year’s Magic Quadrant is that in our leadership of the data discovery category, QlikTech has helped transform the entire BI industry.  But we’re not resting on our laurels. There’s more to come. Check out the series of blog posts about “,” the next generation of the QlikView Business Discovery platform.

We all like stories. Why? We can lose ourselves in them for a time. Stories can make us feel as if we are experiencing something new. This also explains why movies and, to an even greater degree immersive video games and virtual worlds, are so compelling – but that’s a topic for another day.

I appreciate the insight about the human brain and storytelling in the December 5, 2012 Lifehacker article, “The Science of Storytelling: Why Telling a Story Is the Most Powerful Way to Activate Our Brains,” by startup co-founder Leo Widrich. With data storytelling one of the product scenarios of “” (see the related post, “Data Storytelling with ‘’”), the article grabbed my attention.

Brain cortext lights up when you see images.png

Widrich pointed out that when we read words on a PowerPoint slide, for example, our brain goes into language processing mode; the brain is trying to decode words into meaning. In contrast, when we are engaged in storytelling (either on the telling or listening side), not only are the language processing parts of the brain activated—but also any other part of the brain that we would use if we were experiencing the events in the story.

Wait, it gets even cooler. Because this brain activity happens in both the storyteller and the person listening to the story, storytellers can synchronize their brains with the recipient of the story. Whatever the storyteller is experiencing, they can induce the listener to experience too.

What does this have to do with Business Discovery? A whole lot. The same principle applies to numbers on a page or screen as to words. If we just see the numbers in black and white our brain goes into processing mode. We try to figure out what the numbers mean.

With numbers, how do you get your (data) point across? How do you convey the emotion behind your discovery or proposed decision? How do you get others on board with you? If someone is telling or listening to a story about the numbers – how the numbers came to be, why they matter, what their implications are, and what should be done about it – more peoples’ brains (and more of their brains) are engaged. Telling stories with data requires a connection to the data being analyzed.

And, to take this idea all the way to its conclusion, isn’t brain synchronization the nirvana of business intelligence? The nirvana of BI is alignment – getting everyone on the same page so the organization can move as one in the right direction, based on facts.

See these related blog posts:

Thanks to Tom Mackay, principal solution architect at QlikTech, I recently came up on the article, “Seven Dirty Secrets of Data Visualisation” by programmer Nate Agrin and data visualization developer Nick Rabinowitz. In addition to shining light on some best practices in data visualization, this article helps illuminate the difference between standalone data visualization tools and a Business Discovery platform.


The article covered seven dirty secrets of web-based data visualizations; I have thoughts about a few of them:

  • Secret no. 1: Real data is ugly. The article points out that before data can be accessible to, and useful with, data visualization software, a data expert has to find, acquire, load, clean, and transform it. This problem is compounded when the data comes from multiple sources. A lesson? Look for Business Discovery software that enables you to begin working with and analyzing even imperfect data without requiring a whole boat load of external (ETL extract, transform, load) capabilities. Discovery begins with the data. Being able to immediately see the mis-typed, out-of-range, or missing lets you begin tackling the problem from the start.
  • Secret no. 3: There’s no substitute for real data. When getting your hands on software to help you make sense of your data, try it out with your own real data – not dummy sample data. It’s in quickly and easily being able to see the associations in your own real data that the “a-ha” moments occur. This experience may be especially powerful with data sourced from disparate systems. “There is nothing like the moment when a user sees associations between data they have never been able to bring together before,” says Tom Mackay, principal solution architect at QlikTech. He’s speaking from experience; he’s seen it happen time and again at customer sites.
  • Secret no. 6: Visualization is not analysis. Agrin and Rabinowiz argued that visualization is a tool to aid analysis, not a substitute for analytical or statistics skill. True. But it’s not either/or; an analytic app can contain both detailed data, with complex statistical expressions applied, and visualizations that help simplify the picture. Users can analyze the data in the way that is most comfortable for them. And: clear, well-labeled, interactive charts and graphs that are part of an analytic app make it easy for even non-data analysts to explore information and derive insights and meaning – critical in a world where information is strategic and we all have to know how to work with it.
  • Secret no. 2: A bar chart is usually better. “The coolest looking visualisations are often the least useful,” the authors wrote. They pointed out that the cost of the novelty and visual appeal of some data visualizations is clarity of meaning. This potentially leads to comparisons that distort the data and takes viewers to false conclusions. The authors made the same point about animations (“secret no. 5: animate only when appropriate”) and recommended making animations simple, predictable, and replayable. I’d add that another complicating factor is the variety of devices people are using to interact with analytic apps, including tablets and smartphones. We offer some design suggestions of our own in this Technical Brief: “Mobile User Interface Design Best Practices.”
  • Secret no. 7: Data visualization takes more than code. “. . . Creating visualisations that offer real insight or tell a compelling story still requires a particularly wide range of real skills in addition to coding, including graphic design, data analysis, and an understanding of interaction design and human perception,” the authors wrote. A complicating factor is that data experts tend not to be graphic designers. To get past this dichotomy, check out the tips our experts offer on the QlikView Design Blog and take a look at how we are focused on closing the gap between data and design with the next generation of QlikView (see the post, “A Gorgeous and Genius ‘’ – The Best of Scandinavian Design”).

Visualizations are just the tip of the iceberg – the iceberg being a person’s understanding of the data. To be able to derive meaning and insight from data, especially complex data sourced from multiple systems, the user requires not only well-designed, clear, concise data visualizations, but the ability to explore the full dataset on their own. They need to be able to ask and answer their own streams of questions without having to go back to an expert for a new visualization every time they have a follow-up question. This, in a nutshell, is the difference between a standalone data visualization tool and a Business Discovery platform.

During this time when student test scores convey doom and gloom and education budgets are pinched, I jump on positive news when I get it. I came upon some good news recently during a conversation with FirstLine Schools in New Orleans, Louisiana. I spoke with Sia Karamalegos, director of data management, and Rebekah Cain, director of development and communication. FirstLine Schools is a charter management organization managing five public schools. The organization has approximately 2,500 students and 300-400 employees and is a grant recipient in QlikTech’s Change Their World program.

Part of the student dashboard at FirstLine Schools.png

In the words of Gillian Farquhar, QlikTech’s customer ambassador, “Like in most urban environments across the United States, lots of kids in New Orleans come from broken or economically challenged homes and the public education system has had a hard time producing the success it hopes for. Charter management organizations like FirstLine Schools intervened to make a difference early in children’s lives.”

Sia Karamalegos of FirstLine Schools said, “A lot of schools are trying to reform across the United States. A big aspect of reform in public education is using data to drive instruction. A challenge with this, though, is that school leaders, teachers, and special education and intervention staff are generally not data experts. They need to be able to easily access and make use of the data. But the diverse nature of the source data systems we use presented a challenge. Some systems permit users to export data while others don't. The data is in many different formats and the systems each have their own login. For teachers and administrators and even network administrators, this situation created a barrier to being able to make good use of data. People needed a better way to get a complete view of school and student performance.” 

Part of the admin dashboard at FirstLine Schools.png

Rebekah Cain added, “School leaders have a big job. They are already working really hard. So any way we can make it easier for them to access the data they need to do their jobs, the better. Then they can spend their time on things that will improve instruction instead of looking for data.”

Using QlikView, FirstLine Schools is focusing on:

  • Student and school performance and success. FirstLine Schools is collecting data about students such as grades, classes taken, attendance, demographics, behavior (e.g., merits and demerits, suspensions, etc.), and whether the student is in intervention (performing below grade level). The data team assembles data from myriad back-end systems into a cohesive QlikView app that gives constituents a well-rounded perspective on a scholar’s “at-riskness,” trends in a student’s progress over time, and whether the approaches the school is taking with the scholar are the right ones. With QlikView, stakeholders can explore the data, toggling among students or schools or selecting a particular quarter—all the while seeing only the data they are supposed to see. “If a student is not progressing the way we expected,” said Rebekah Cain, “we know what we are doing isn’t working – so we need to add additional resources.”  In addition, a lot of the data the schools track is not about student grades. Schools conduct progress monitoring and interim tests to see if they are improving across large student populations.
  • Administrator dashboard. New Orleans is an open enrollment city; students don’t have to live in a particular neighborhood to attend a school there. What this means for FirstLine Schools is that it has to recruit students. To aid in recruitment and support people who are writing grants and performing other compliance tasks, FirstLine Schools uses QlikView to track and present statistics like enrollment, year-to-date attendance, tardiness, and truancy. Stakeholders can get the info they need when they need it. They can drill down into school or grade level and can select factors like special education status, homelessness, gender, or others to get just the relevant stats they need. QlikView logs into the various source systems automatically using scripts so the data in the apps is always up to date.

I asked Sia about FirstLine Schools’ “a-ha moments” with QlikView. In one example, they noticed that for different schools in the system, the rate of in-school suspensions vs. out-of-school suspensions had changed. This realization sparked further investigation. Moving forward, FirstLine Schools is planning to tie in additional data to identify which intervention programs are having the greatest success so they can spread best practices.

Are you inspired? I am.

Filter Blog

By date:
By tag: