Erica Driver

Why I Tweet

Posted by Erica Driver Sep 28, 2010

It's sometimes hard to explain to busy people why I use Twitter. "Who has time for that?" is a common question. They want to see data that shows spending time on Twitter really adds value. While I don't have data, I have anecdotes that I've been saving up. I've used Twitter heavily for the last couple of years to build my network and keep a finger on the pulse of what's going on in the communities I care about.

I often keep a Twitter dashboard up and running on a second monitor in my office. During the last couple months I've used Twitter to:

  • Learn. I'm relatively new to the BI software market. I've been an industry analyst for the past 14 years focusing on a wide variety of technology and strategy areas?but never focused closely on BI. So I use Twitter to listen and join in on conversations taking place among industry analysts, consultants, vendors, and users of BI software. Conversations are under way on lots of interesting topics including in-memory analytics, self-service BI, stack vs. portfolio approach, and BI in the bigger context of enterprise software.
  • Drive conversation. Since early August I've been working on white papers and writing blog articles. In just a quick glance at a one-line post, Twitter helps me engage with people with others interested in the same topics I'm thinking about. /cfs-file.ashx/__key/CommunityServer.Blogs.Components.WeblogFiles/theqlikviewblog/James-Kobielus-tweet-about-Twitter.PNG
  • Build my network. It's through professional networks that knowledge is shared, trust built, customers earned, and jobs found. (Speaking of: QlikTech is hiring!) When I have a few minutes, I see who my connections are following and add to the list of people I follow. This freshens my tweet stream?and helps round my perspectives. Twitter can also be helpful in leading me to more information about people I want to be connected to. Twitter can sometimes lead me to more information about people who comment on my blog posts, for example.
  • Spark creative thoughts. In my role in product marketing at QlikTech, I blog on themes like the consumer enterprise, the BI software market, innovation at QlikTech, and the associative experience. Sometimes I look up to my Twitter dashboard and see something float by?a link to a blog article, an announcement, or simply an observation?that gets my creative juices flowing. This tweet by Forrester Research analyst James Kobielus is a case in point. /cfs-file.ashx/__key/CommunityServer.Blogs.Components.WeblogFiles/theqlikviewblog/James-Kobielus-tweet-about-Twitter.PNG

 

 

In his 2009 book, The Design of Business: Why Design Thinking Is the Next Competitive Advantage, Roger Martin presented the concept of the "knowledge funnel." The knowledge funnel is the advancement of knowledge from a disorganized mystery state to well-ordered algorithms. This is achieved through trial and error or rapid prototyping. The knowledge funnel is a model for how businesses of all sorts can advance knowledge and capture value.

The knowledge funnel has three consecutive stages, each of which represents a simplification and ordering of knowledge. According to Martin, "As understanding moves from mystery to heuristic to algorithm, extraneous information is pared away; the complexities of the world are mastered through simplification."

QlikView Is a Tool for Mastering the Knowledge Funnel

QlikView technical advisor Elif Tutuk has identified ways in which QlikView is an optimal tool to help decision makers move through the knowledge funnel quickly and easily.

 

 

Stage

Description

The QlikView Solution

1

Mystery

  • The observation of phenomena-things we see but don't yet understand
  • A first stab at trying to answer a question
  • Load millions of records in memory from multiple data sources
  • Observe the data
  • Search associatively
  • Visualize relationships in the data
  • Select values and see what data is associated
  • Narrow down selections

2

Heuristic

  • A method or procedure that serves as an aid to learning, discovery, and problem-solving by experiment or trial and error
  • A rule of thumb that helps to narrow the field of inquiry and work the mystery down to a manageable size
  • An organized way of exploring the possibilities-of thinking about the mystery-that provides a simplified understanding of it
  • Analyze granular data with no pre-calulation or pre aggregation
  • Build metrics on the fly
  • Use charts to see patterns or catch outliers in the data by aggregating the data with any combination of dimensions
  • Narrow the business questions
  • Understand the root cause of an outlier on a chart by selecting that data point on the chart and digging in to explore its meaning

3

Algorithm

  • A fixed formula
  • An explicit, step-by-step procedure for solving a problem
  • Means of simplifying and adding structure to the loose, unregimented heuristics so anyone with access to the algorithm can deploy it with efficiency
  • After finding the answers hidden in the data, formalize analysis and share findings with others using bookmarks and collaboration objects
  • Once a new way of analyzing the business has been established, easily merge the chart or the QlikView application into the production environment, under the control of IT


People need a BI tool that helps them pass through the knowledge funnel quickly. The tool should be flexible and nimble enough so that users do not get stuck in any stage. They should be able to build to think?ask a business question, build a chart to find the answer, make selections to see associations and outliers, and change the chart or create a new one. Decision makers should have access to a tool that helps them rapidly advance through the mystery, heuristic, and algorithm phases of the knowledge funnel quickly and easily, freeing them up to move on to the next business challenge.

For more insights by Elif Tutuk, see these related blog articles: "Build to Think: Applying Design Thinking to BI" and "QlikView Supports a Build to Think Approach to BI."

One of QlikTech's core values is "Take Responsibility." As part of taking responsibility for the products we deliver to market, our demo and best practices team tests each new software release candidate in a real-world environment and uncovers and shares QlikView best practices. We have a QlikView Technology White Paper coming out in a few weeks highlighting best practices the team has uncovered around scalability. To set the stage, I caught up with Shima Nakazawa, director of the team, and I've got some of her insights to share with you.

Erica Driver: What is the mission of the demo and best practices team?

Shima Nakazawa: At a high level, I'd say it is to ensure the quality of our products and share best practices with customers and partners.

Erica Driver: Why did QlikTech establish this team?

Shima Nakazawa: The team was established about three years ago with a few focus areas: operating our public demo site, improving the operations of QlikTech's internal QlikView environment, and sharing QlikTech's best practices with customers, partners, and prospects.

Erica Driver: What are some of the best practices you've identified for overseeing a QlikView deployment that consists of 200 applications and 900 users?

Shima Nakazawa: Wow, where do I start? By trial and error, we learned many lessons in the past few years. Here are a few best practices we've identified. Use a collaborative software development platform, and QlikView Workbench, for team development efforts. Test applications in a staging environment prior to production rollout. Craft a QlikView document deployment plan. Load data incrementally rather than all at once. Run memory-intensive services (such as QlikView Server and the QlikView Distribution Service) on separate machines. And consider virtualization for efficient use of machine processing power.

Erica Driver: What are some of the things your team has been working on lately?

Shima Nakazawa: One of the most visible projects is a demo app called Kick It and Qlik It, in support of a global marketing campaign. We designed the deployment of this application in such a way that we were successfully able to handle 8,000 users a day on this one application. We will be sharing some best practices we uncovered while doing this project in the upcoming QlikView Technology White Paper, "How We Supported Thousands of Users Daily on a Global Scale." Stay tuned!

SAP recently announced a set of pre-packaged, industry-specific BI applications. We feel compelled to weigh in because SAP's announcement exemplifies a philosophical approach that is fundamentally different from QlikTech's. Our philosophy is that software should easily fit the business?it must be flexible. Why is this so important? Because flexibility is a pre-requisite to fast deployment, quick time to value, and ongoing business relevance.

 

Pre-packaged, industry-specific BI solutions are by their very nature born out of a different philosophy?one that contends that all organizations in an industry share some portion of business processes and practices in common, and therefore flexibility isn't all that important. Not to say there's anything wrong with useful functionality provided out of the box?as long as that functionality can be easily modified to fit the business. Here's the catch:

  • No two businesses are the same. Organizations in one industry (say, healthcare) tend to operate differently from organizations in another (say, retail). It certainly seems logical that software vendors try to capture best practices around how organizations might analyze sales performance or inventory, for example. But even organizations within industries may be wildly different from one another. Business processes and practices vary based on culture, geographic location, business process innovation, how long the company has been around, and how centralized or decentralized the organization is. Small nuances often result in large differences in the data required for business intelligence, how that data is consolidated, the analysis required, and how that data is visualized and consumed.
  • Your choices: live with what's in the box, or customize it?over and over again. Put another way: "Modify your business to fit our software . . . or pay a lot of money to modify the software." Because no two businesses are the same, pre-packaged, industry-specific solutions nearly always require customization?which comes at a cost. BI tools that are not associative at the engine level (see figure) require long deployments while developers customize the application layer to manage the specific associations required to answer a particular business question. (See the related blog article, "QlikView Is Associative to Its Very Core.") Keep in mind that dashboards change over time as the business changes, so money spent customizing a packaged solution must be re-spent every time the dashboard has to be modified. When the BI application needs to answer a slightly different business question, a developer must alter the application layer again. This process is time-consuming and expensive.

In contrast, we designed QlikView around our philosophy that software should be easily molded to fit the business. Using QlikView, organizations can prototype with their data very quickly?in line with their own requirements, not ours. (See the related blog posts, "Build to Think: Applying Design Thinking to BI" and "QlikView Supports a Build to Think Approach.") QlikView is flexible and agile so as the business changes, so does the software. This makes it quick to deploy and modify.

This strength came out in a couple of recent analyst studies, which included survey responses from BI users in a wide range of industries. BARC, for example, found that QlikTech was among leaders in implementations within 3 months and had 85% of customers implemented within 6 months (see the August, 2010 report "The BI Survey 9: The Customer Verdict"). Aberdeen Group, in another example, found that QlikView users were on average able to drive a revision to dashboards from conception to completion in a single day, as opposed to an average of 3.5 days for all survey respondents (see the August, 2010 research brief, "QlikView Customers Outperform Best-in-Class with Dashboards").

I'd sum it up like this. It's important to pick software that fits your business, not just your industry, because no two businesses are exactly alike. With QlikView you spend your money and time not on software customization, implementation, training, and business re-engineering . . . but on attracting customers, bringing new products and services to market, and building your business.

CITO Research recently published a white paper, sponsored by QlikTech, titled "Quick Answers from Faster Dashboards." This paper explains how "living dashboards" empower business users. A living dashboard:

  • Is a business intelligence dashboard that is always changing, evolving, and maturing
  • Is implemented quickly and evolves iteratively, at the pace of business change
  • Morphs at the guidance of users rather than IT
  • Enables users to start their data exploration with a hunch and gain understanding by summarizing the data into aggregate metrics and charts

The white paper contains a graphic that I think is a useful depiction of how living dashboards lead to quick answers. Living dashboards close the loop, jumpstart innovation, propagate knowledge, and harvest low-hanging fruit.

http://community.qlik.com/cfs-file.ashx/__key/CommunityServer.Blogs.Components.WeblogFiles/theqlikviewblog/Living-dashboards-graphic-V2.PNG

 

QlikView certainly fits the bill as a tool for creating living dashboards. It can be implemented and modified quickly, which increases its power as a tool for jumpstarting innovation and propagating knowledge. In the recent BARC BI Survey 9, QlikView came out in the top two vendors, as measured by fast implementation time. In this study, QlikTech was among leaders in implementations within 3 months and had 85% of customers implemented within 6 months. And in a recent research brief titled, "QlikView Customers Outperform Best-in-Class with Dashboards," Aberdeen Group found that QlikView users were on average able to drive a revision to dashboards from conception to completion in a single day, as opposed to an average of 3.5 days for all respondents.

Business Application Research Center (BARC) recently published a report called, "The BI Survey 9-The Customer Verdict." From December 2009 through February 2010, BARC conducted a Web-based survey of 23 BI dashboard vendors' customers. The total sample size was about 2,660 individuals. Enough QlikView customers responded to the survey (n=146) for BARC to calculate all key performance indicators (KPIs) for QlikView reliably.

In this study, QlikView came out as the #1 or #2 BI vendor (in a field of 23 vendors) against four measures:

  • Intention to buy more licenses. On this measure, QlikView earned the highest score of all vendors studied. QlikView customers' intent to buy more licenses bodes well for customer satisfaction and our ability to deliver value?and our future prospects!
  • Customer loyalty. QlikView came out ahead of the entire pack as measured by customer loyalty. To calculate this aggregated score, BARC combined the standardization preferences, intention to buy more licenses, and renewal rates.
  • Fast implementation time. QlikView is typically deployed very quickly?often in a matter of just a few weeks. QlikTech was among leaders in implementations within 3 months and had 85% of customers implemented within 6 months. This is compared to less than 60% for SAP and MicroStrategy.
  • Low number of query performance complaints. BARC has found that query performance is the most frequently-reported BI product-related problem. Query performance problems can lead to reduced business benefits and goal achievements. Against this measure, QlikView came out at the head of the pack.

QlikView also came out strong (ranked among the top 5) in a few other areas, including:

  • Range (number) of applications customers deployed. QlikView ranked in the top three vendors in terms of number of applications customers deployed using the software. We interpret this score as evidence of QlikView's flexibility and ease of use.
  • Importance to the customer when deployed as part of a portfolio. In organizations that have multiple BI products, QlikView tended to be high on the list of BI dashboard products customers deem important.
  • Number of applications and business functions per administrator. QlikView was in the top 5 vendors as measured by number of applications per administrator and the top 4 as measured by number of business functions addressed per administrator. The higher the number of apps and business functions per administrator, the lower the cost of ownership.

We hear it time and again: our customers love QlikView, and for good reason. What do you think? Does the BARC survey data ring true for you?

Unanticipated Queries

Posted by mmy Sep 10, 2010

I viewed the illustration in the following link: http://files.qlik.com/blogs/theqlikviewblog/QV-Value-Prop-Illustration.pdf and thought it did a great job of communicating an example of how in-memory associative analysis differs from static reports that rely on pre-determined queries. If you ever prepared long and hard to present answers for a meeting only to have it turn into the nightmare depicted by the illustration in the link above, you probably gained a greater appreciation for more dynamic, flexible means of reporting.

There's no way anyone can pre-determine all the different possible dimensions, views, and perspectives that will satisfy a group of executives with different goals and vantage points. And there lies the advantage of in-memory associative analysis.

In-memory associative analysis works the way the human mind does, so a variety of unanticipated queries like the ones above can be calculated in random access memory and displayed in seconds to provide a more thorough understanding of what is really happening and what needs to be done. Answers to questions provoke new questions and so on and so forth. Traditional analysis and reporting tools that require hours and days of IT people hours to accommodate unanticipated queries are ineffective in providing answers that address the different perspectives executives and stakeholders reveal during the course of a typical business meeting.

In May 2010, QlikTech announced the launch of QlikView Labs. The Labs, based in Lund, Sweden, are a place for exploring new ideas and facilitating QlikView product innovation. We are prototyping and testing ideas and sharing them throughout QlikTech?and with the world. I recently touched base with Labs director Alistair Eaves for a sneak peek into what's going on in the Labs. Here are some highlights.

Erica Driver: What is the mission of QlikView Labs?

Alistair Eaves: It's a central strategic function at QlikTech designed to drive innovation by capturing ideas, nurturing creativity, and conducting focused research. At the core, it's about listening to problems and dreaming up solutions. Our goal is to turn the best ideas into working QlikView prototypes.

Erica Driver: Why did QlikTech establish the Labs now, this year, in 2010?

Alistair Eaves: This is a big year for QlikTech. We had a successful IPO in July, have grown our customer base to more than 15,000 compared to 11,000 a year ago, and have a new product release coming out. As the company continues to rapidly grow, we have to make a concentrated effort to innovate and stay on top of new technologies. Creating a laboratory-type environment is one way to accomplish this. We established the Labs to capture and nurture the wealth of creative thinking we see taking place in the greater QlikTech community, and bring new and innovative ideas to our products.

Erica Driver: How do you decide which ideas are worth investing in?

Alistair Eaves: The most important factor is, "Does it solve a real problem?" This is a must. We also evaluate ideas based on whether they have commercial value and are practical and simple. Simple is the QlikView way!

Erica Driver: What are some of the deliverables coming out of the Labs?

Alistair Eaves: We are working on technology previews, which are beta applications selected from the QlikView product roadmap. We are also working on experimental applications and prototypes. Some of these may graduate to production features of QlikView.

Erica Driver: How about a sneak peek at some of the projects you're working on right now?

Alistair Eaves: One interesting project we're working on is a universal touch client. When we started this project, I thought that what we were working on was a universal mobile client. But it's clear that it's more about touch than just about mobile. Touch interfaces will become pervasive?whether on mobile devices like phones and tablets, or on laptops or even larger devices.

Erica Driver: Why are touch interfaces important for QlikView?

Alistair Eaves: Because with touch screens, the interaction feels natural. On tablets and larger devices, which have more screen real estate than a mobile phone, QlikView gives the user an instant response and feedback. It's natural and intuitive to make selections in list boxes and charts. Touch screens really force you to think about a simpler way of doing things.

To learn more, please visit Stay tuned for the launch of the QlikView Labs web site, where you'll be able to play with the latest prototypes and participate in Labs projects.

In August 2010, Aberdeen Group published a research brief called "QlikView Customers Outperform Best-in-Class with Dashboards." Between May and June 2010, Aberdeen Group surveyed 400 end-user organizations about their BI dashboard implementation costs, use, derived value, and future plans.

Key findings from Aberdeen's research brief showed that QlikView:

  • Users have best-in-class forecast accuracy. QlikView users reported 100.9% accuracy of revenue to budget forecast and 100.3% accuracy of bottom-line budget forecast. Organizations that were in the top 20% of performers (best in class) were nearly 100% accurate in their forecasts. This was about five percentage points higher than non best-in-class survey respondents.
  • Beats best-in-class in one-year, per-user cost of dashboards. According to this study, organizations that deployed QlikView were able to deliver dashboards to end users at a one-year cost per user that was roughly three quarters that of the best in class. (Aberdeen asked survey respondents, "How much has your organization spent in the past 12 months on dashboard software licenses (or subscriptions), implementation services, and maintenance?") The average best-in-class, one-year, per-user cost of dashboards was $28.53. The average for all others was $35.58. QlikView users came out on top, with a per-user cost of $22.41.
  • Users' dashboard initiatives are driven by needs of line of business managers. Aberdeen found that 62% of QlikView users drove their dashboard initiatives based on the needs of line of business managers, whereas this was true for only 48% of best in class and 41% of all others. As such, QlikView users' dashboards were more likely than others' to report metrics that are relevant to the decisions these managers needed to make every day.
  • Supports rapid dashboard prototyping. QlikView users were on average able to drive a revision to dashboards from conception to completion in a single day, as opposed to an average of 3.5 days for all respondents. This finding verifies what we hear from our customers every day: QlikView is fast, flexible, and easy to use.
  • Makes it easy to identify data sources. Aberdeen found that QlikView users create a regular process for identifying and prioritizing data sources, which reduces the amount of time managers have to wait for IT to assemble relevant data. Half of the QlikView user respondents said they had a process for prioritizing data for user access, compared to 40% for best in class and 32% for all others.
  • Enables users to identify KPIs independently of data sources. More than half (55%) of QlikView users said they can define key performance indicators (KPIs) independently of data sources (compared to 39% for the best in class and 27% for all others). By inserting KPIs into dashboards before data sources are identified, line of business managers are free to experiment with modifications to dashboards. They don't have to go to IT for every change they want to make.

We're proud of these findings. They reflect the comments we hear from our customers every day. Do they ring true to you as well?

Filter Blog

By date:
By tag: