I’m moving the blog to http://betterbizintel.blogspot.com/ for one simple reason: I can embed analytics published to Tableau Public directly in the body of the posts.
I do most of my data analysis work with Tableau, and believe it’s one of the most important tools to come along since the invention of the original 4GLs.
If and when WordPress provides this ability, or Tableau Public content changes into a form WordPress accepts I’ll revisit the decision.
Two months on the Panamanian Pacific coast. The beach is right below us, the surf a constant presence. The tides here are quite high-averaging almost 15 feet.
Living right on the ocean has it’s great benefits. The beaches here are strikingly beautiful; the surf intermingles black volcanic and white sand into fantastic patterns visible from space. Or at least in Google Earth and maps. The ocean has shaped the beaches into a steeply sloped upper section, and a nearly flat lower zone wonderful for all types of activities. Here’s a time lapse video of high tide in the early morning (it’s receding):
As you can see in the video, there’s not much room on the beach at high tide, and plenty at low tide.
It’s hot here. Not absolutely blisteringly hot, but hot enough to avoid strenuous outdoor activities, walking or running on the beach say, during the middle of the day.
It’s important to know the tides so that you can plan your beach activities. Fortunately, the local community has posted the tides online in a series of web pages, one for each month; March, 2012’s page is here. Unfortunately, the tides are presented in a table that, although it contains the data, is minimally useful.
A recently created discussion in the LinkedIn Business Intelligence group (here) prompted this note.
The original blog is favorable towards the IBM Cognos – Solution Implementation Method (C-SIM). This note disagrees and argues that C-SIM is in fact detrimental to the delivery of effective, high quality, valuable BI that really does help the business.
The overarching problem with C-SIM is its assumption that a BI “solution” is a discrete, finite thing amenable to the traditional Analyze–Design–Configure&Build–Deploy–Operate approach to building fixed-function software systems whose functionality is determinable prior to its construction.
C-SIM is a very poor model for delivering BI solutions. Its history is littered with embarrassingly low rates of delivering even reasonable levels of business value in the form of meaningful, timely, valuable data-based business information and insights.
You may ask yourself: “That’s a pretty bold claim. How can I evaluate it?”
I’m glad you asked.
Here’s a threshold test one should ask of any potential BI vendor partner: “How quickly will I or one of the other business stakeholders get valuable information from my business data?”
In the case of C-SIM a reasonable follow-on is: “Show me in your method/project plans the earliest point at which this happens.”
If your C-SIM vendor partner cannot or will not answer, stumbles, fumbles, hems or haws there’s a very large problem on your near horizon, should you choose to go down that path.
There’s a bright spot: C-SIM does have real value.
It’s attractive for large BI resource vendors because it’s road-mapped well into the future, and provides a firm framework for dedicating the amount of resources—people, and technology—that can be assigned to the elaborate project tasks structure, guaranteeing predictable revenue streams. It’s hard to argue with the Big BI resource vendors that their model isn’t working when it’s generating $Billions in revenue; their position boils down to: “Of course, it’s working. look at all the money our clients are paying us.” (A cynic might say that the method is only valuable for providing the revenues that the Big BI consulting company principals’ fortunes are built upon, and that they get to keep whether or not the business clients get the information and insights they’ve paid for.)
Although, with the increasing media awareness that Big BI initiatives are largely not delivering upon their promises, and the emergence of Agility in the practice of BI, along with 5 years of experience with tools like Tableau, Spotfire, QlikView, and their cousins, the tide might be changing.
Being a BI professional requires a full suite of intellectual skills. Our lives are spent helping people make sense out of their business data. In this pursuit we must be capable of making sense out of the things we’re told, what we read, and the data we’re presented with.
A big challenge, perhaps the biggest, is understanding what our business partners tell us. We are responsible for learning their language, their perspectives, their business ideas, their idioms. They are not responsible for learning ours.
We also must be capable of determining when what we’re being told doesn’t make sense. When we’re presented with something that doesn’t makes sense it’s our obligation to clear things up. Proceeding otherwise is very, very bad form, and almost guarantees big problems later.
In this spirit, I’m providing a snippet from an article on Bicycling.com (below) that claims to correct several fallacies associated with sports, specifically bicycling, nutrition. Your job as a BI professional is to spot the problem with the material. You don’t need to be a sports nutritionist or bicyclist. (but aren’t you happy I told you there is a problem?)
Feel free to use the comments section for your observations, or contact me directly.
The full article is on Bicycling.com here.
Problematic section follows:
A calorie is a calorie
This might be the biggest weight-loss misunderstanding in existence. For years we’ve been told that weight loss is a simple calories-in, calories-out equation, and 3,500 excess calories will put on a pound whether they come from soybeans or banana cream pie. That’s simply not true.
“There are three key types of calories: carbohydrate, protein and fat,” says sports nutritionist Cynthia Sass, MPH, RD, CSSD, creator and coauthor of the Flat Belly Diet (published by Rodale, Bicycling’s parent company). “They’re as different as gasoline, motor oil and brake fluid in terms of the roles they play in keeping your body operating optimally.” Sass says that many of her clients might eat the perfect number of calories, but they have cut their fat intake too much. So the jobs that fat does, such as repairing cell membranes and optimizing hormones, go undone, and the surplus carbs are stored as fat. By correcting her clients’ balance of carbs, protein and fat without changing their calorie intake, she says, she has helped them lose weight, improve their immune systems, gain muscle and boost energy.
1. Is this funny? Insp. Clouseau checks in
2. Why is this funny?
3. What does this have to do with BI?
2. See #3.
3. Quite a lot. This joke goes straight to the heart of BI.
As BI professionals we’re responsible for helping our business partners gain the information and insights they need from their data.
Sometimes they understand precisely what they want to know and how it’s represented in the data. In which case we’re responsible for removing the barriers between them and their data, given the opportunities and constraints of the environment we’re working in.
This is the idealized scenario that’s the outcome of the traditional approach
More commonly our partner, call him Jim, asks us: “Does your dog bite?” when he really means: “Will this dog bite me?”
It’s our responsibility to recognize this when it happens and work with Jim to get to the point where two conditions are true: Jim asks the right question, the one he really wants the answer to; and he receives the correct answer to that question, in this case, most likely some variation of: “Yes”, “No”, or “I don’t know.”
Now, you might be thinking: “Why not cut to the quick, take the short cut, and just answer Jim’s real question for him?” And this might seem like a reasonable thing. It’s what the naive BI practitioner may well do. After all, isn’t Jim getting the information he’s looking for, and maybe avoiding getting bitten in the bargain?
The world of BI isn’t the world of Clouseau, where it would be perfectly OK to reply: “I don’t know, it’s not my dog”, where the only real problem is that the joke wouldn’t work.
Our responsibilities as BI professionals are multifold, and include helping our partners, Jim included, in multiple dimensions. Our primary responsibility is, of course, to ensure that they get the information they need. But we also need to help them learn to help themselves (Here’s to you, Mrs. Robinson), and this means helping them develop the awareness of the subtleties, complexities and consequences of precision in thinking and communication. This is in some circles considered heretical, crossing some boundary of propriety where we’re not permitted to presume that our partners’ cognitive and intellectual skills aren’t up to snuff. This usually gets phrased as something along the lines of: “You can’t insult our clients, you just need to give them what they need.” To which we say: “Dammit, Jim. We’re BI Professionals, not mollycoddling data weenies.”
The more we help our partners understand that they can, and should, simply ask for the information they want, and how to do it, the better off they will be. They will be able to get the information and insights they need faster, easier, and with less effort. This is better for everyone.
The BI industry—is—an industry. Over the past 20 years it’s grown into a multi-billion dollar ($US) business largely on the strength of a symbiotic relationship between the luminaries who designed and advocated the classic enterprise BI paradigm and the vendors who created and sold the software and technical staffing to build the solutions.
Who are today’s BI thought leaders? Who should you listen to? Who has the knowledge, insights, and experience to help you find your path to BI success? Most importantly, who holds your interests first and foremost?
When you’re considering which BI thought leaders to follow, you might want to take into consideration their position in the BI universe.
Many of the people mentioned as BI thought leaders in the techno-business-IT media are the old guard, who made their reputations creating, promoting, and profiting from the framing of BI as the big, complex, complicated, enterprise-class industrial undertakings that have given us the current mainstream BI world. If you think that these people are the ones you want to follow into the future, they’re the leaders for you.
On the other hand, the world of BI is finally breaking free of the shackles of Big BI. It’s no longer a case of “Let’s build the all-inclusive data warehouse so we can divine the single version of the truth that we’ll then allow the business to see.”
NoSQL, big data, appliances, in-memory databases, intimate analytics, direct data access, high visualization quality, agile development, and other advances have dramatically reduced the barriers between business people and their data.
If you are interested in seeing where BI is capable of going, and how it might get there, beware the BI pitchmen (and women) who brought you the past.
Look to Stephen Few, Edward Tufte, Colin Ware, and their peers, for the principles of analytic information design. Run away from anyone who’s written a dashboard book or published thought-leader materials featuring pie charts, dial gauges, wet-look anything, bulb thermometers, radar charts, 3D inverted cone graphs, 3D anything, excessive color, bouncing bars, or any of the other look-at-me examples of the eye-candy-rules school of Sexy Marketingware Sells BI.
Look to Christian Chabot for the democratization of data. Look to the other vendors who are bringing tools to market that eliminate barriers between people and their data. These are the people who say “your data, your information, your way”. Pay no attention to those who say “First thing, let’s put in a comprehensive program of data governance to assess and ensure data quality. Then we’ll build us a comprehensive enterprise data warehouse and front it with the universal answer-anything analytical platform. Then, once we’re confident that we’ve fully identified, confirmed, vetted, and validated the essential KPIs, metrics, dashboards, scorecards, strategy maps, self-service reporting requirements, and other analytical elements and artifacts we’ll open the doors and let the business in for the show.”
Look for and follow the innovators who understand that the best BI architecture is emergent, the result of continual refinement responding to the aggregation and consolidation of the multiple inputs from live analytics created collaboratively with business stakeholders, and the synthesis of cross-source common business domains into coherent information models that then, in their proper turn, direct the design and development of the enterprise analytical data stores, e.g. data marts and warehouses, that provide the high level comprehensive information.
Follow those who understand that BI is not a project, with discrete inputs and deliverables, but a process of discovering, satisfying, and supporting business understanding of the information, insights, and stories that the data can tell. There’s little value in using your time to follow those who tell you that BI is all about building big data cathedrals and the monstrous software platforms to interrogate them. Be skeptical of those who are selling an architecture, highly polished universal schemas, their BI recipe book.
Follow those who understand that BI is a professional practice requiring intellectual and cognitive talents and abilities, education, experience, training, and the passion for figuring things out. Shun those who cast BI as an industrial, automated process well-conducted with the proper acquisition of sufficient technology along with the right blend of onshore/offshore cost-contained resource allocation and custom strategic very expensive consultant leadership.
Follow those who commit to delivering value in the form of real, live analytics immediately and constantly throughout your organization’s lifetime. Dismiss those whose idea of agile BI is taking a month or more to deliver a report identifying low hanging, high value strategic metrics that will provide your company with an enduring competitive advantage along with the very large and impressive MS Project plan manifesting the “Special BI Methodology” laying out the detailed step-by-step task structure and timeline that will deliver the results “our very experienced industry expert consultants” have determined you need to dominate your industry as an analytical competitor.
Dashboard Insights announced its winner of the best dashboard of 2011 contest
This post describes some of its flaws, and adjustments that improve its ability to communicate the salient points.
The winning dashboard was created in Tableau and published to Tableau Public, where it’s available for viewing and downloading.
Note: Tableau Public dashboards can be functionally embedded; this one’s too wide to show it all, so the Dashboard images in this post are reduced reduced to fit. Click on any of them to connect to them on Tableau Public.
The following image shows the dashboard zones identified for redesign. Descriptions of the adjustments made, and the rationale behind them, follow.