Superman can take it easy... because he has actionable information

When Superman goes into action, does he usually know only that truth, justice and the American way are under attack OR does he know that truth, justice and the American way are under attack AND who the bad guys are?  I think you’ll agree that he generally knows both that there’s a problem and who to go after.  For, you see, Superman is working with actionable business intelligence.  (BTW, I’ll have to check our records but I believe that Dataspace was the developer of his BI system )

Actionable has been a hot term for a while now.  Everyone wants their BI to be actionable BUT, how many folks actually step back and try to understand what ‘actionable’ means?

To me, actionable means that BI shouldn’t just deliver information but should deliver information that drives toward a reactive or proactive plan of attack.  For example, which of the following is more useful:

Here is a list of our YTD sales by salesperson

OR

Here is a list of our YTD sales by salesperson compared to their YTD sales quotas?

The first one gives you information.  The second one gives you information you can act on – you know where your problems are and you know who to call.

It’s important to note that actionable data is different at the field level than at the executive level.  Field level managers are focused on addressing immediate issues; answering the question, “Where are my problems today?”  For example:

• In a distribution facility: What shipments are late?
• In a sales organization: Which salespeople are selling below their targets?
• In a hospital: What maintenance must be performed in order for us to remain compliant with relevant laws?

On the other hand, rather than solving a problem with today’s details, executives focus more on managing their subordinate managers.  An executive might ask a set of questions like:

• What is the general trend in our sales, or distribution, or maintenance performance?
• How does that trend compare to the goals I’ve set for the period?
• Is the trend headed in the right direction?
• Who’s the manager who’s responsible for that part of our business?

Even though executive-level and field-level managers have different ways of viewing data, if you capture enough detail, the same data can serve both audiences.  Ralph Kimball, father of the star schema, suggested that we capture a lot of detail because we can always summarize up to the executive level.  He also knew that if we store only aggregate data, we’ve lost the detail; we can’t get back down to the field level.

The key to success here is to ask: what does my user do with this report once he has it?  Does he take some action?  If so, are we directing him toward that action or just giving him information with no idea of what to do with it?  (BTW, if he doesn’t take some action, then why are we spending money to give him reports?)

So, to be the Superman of BI requirements, don’t look just at what data you can produce but work with your users to understand their business processes and how the reports you deliver can facilitate them.  Remember, you’re not delivering data, you are fighting to support truth, justice and the _______ (your nationality here) way.

Chances are that you can make a small change to your BI data that could provide a huge increase in value.  How?  By including data from outside your company.  In this post I focus primarily on customer and sales data but the concepts apply equally well to a wide range of analytic situations – do they apply to yours?

BI Systems Are Internally Focused

The vast majority of business intelligence systems are internally focused.   They show data from within the company – what products did we sell, who is our least expensive delivery vendor, what is our average salary.  These systems might be able tell you where performance is strong (e.g. what customers are buying more than they did last year) AND where problems exist (e.g. which customers are buying less than they did last year).  But, any opportunities they show are based on things you already know – on data you already have.

Huge Opportunities Beyond the Borders of Your Organization

In other words, your BI systems show only opportunities embedded in the data you already have.  In many cases this represents the proverbial tip of the iceberg.  For example, how much do you know about good and bad customers besides their names and addresses?  Do you know their income levels?  Do you know how many people are in their households?  Do you know their average ages?  Do you know how a good customer differs from a weak one along these dimensions?  And, finally, do you know who is not a customer but fits the profile of a good customer?

Chances are that this kind of information is not captured by your systems but, it can provide huge value.

Two Types of External Data

External data breaks down into two basic categories: Attibutes of things you already know and new things you know nothing about.

Attributes of The Customers You Already Know

Consider additional knowledge about customers you already know to be attributes of those customers.  The list of possible attributes is infinite.  Things like income level, household size, number of cars owned, level of crime in the local area, and average level of local twitter usage are all examples of attributes.  Getting this information, and tying it to your existing customer records, can give you a much more nuanced idea of what represents a good customer – and how to reach that customer.

Information About Prospective Customers That You Don’t Already Know

So, you’ve taken the step of broadening your view of your existing customers but how do you identify new prospects?  By understanding your attributes of good customers and acquiring lists of prospects filtered through that “lens”.  Then, make your pitch to those prospects.  Given that you know a lot about these prospects, you can even tailor that pitch.  For example, offer minivans to families and sports sedans to yuppies (wait, did I just date myself?  Does anyone use the term ‘yuppie’ anymore?)

Where You can Find External Data

An amazing about of external data is available to those who look for it and,  it’s becoming easier to find.  Here are just a few places where you might want to look:

GeoCommons is a really interesting concept – public domain data sets.  There are really interesting data sets here and they can be yours.  The site requires that you use it for non commercial purposes but GeoCommons parent, FortiusOne can provide commercial licenses.

Microsoft is now in the game with their SQL Azure service Codename “Dallas”.  The site contains a number of free and for-pay data sets.  For example, here is the description of one of the free sets: Cumulative American Recovery and Reinvestment Act of 2009 (ARRA) Prime and Sub-recipient award reports from 02/17/2009 to 06/30/2010 updated as of 09/22/2010.

Another place where you might find interesting data sets is DexOne.  Dex is the largest publisher of yellow pages phone books and a provider of online search and marketing services.  The company is a Dataspace client and I can attest to how seriously they take their data.

There are, of course, a million other places to look, these are just a start.

How You Can Get Started

If you have a marketing background you probably recognize that we’ve stepped firmly into the realm of campaign management.  We certainly have but remember that these techniques can be applied to areas far afield from traditional campaign management.  For example, site selection or fraud risk analysis or investment valuation.  What it takes is creativity.  Start by thinking about what you’d like to know but isn’t accessible from within your company.

Of course it’s not quite as simple as I note here but, like many things in BI, you can make it hard or you can make it easy.  Make your first cuts manageable.  Start by experimenting on a small scale.  Come up with an idea and attack it in a week or less.

So, what do you think?  Are you using external data?  Write your comments below and feel free to contact me directly if you’d like to kick some ideas around!

– Ben

What is business intelligence?  It’s the set of tools, abilities, and data required to create a picture of the world and how your organization operates in that world.  In building your business intelligence capabilities, you can learn a lot by looking at a historical precedent, the rise of radar, and applying it to your problems of getting to your data.

The Battle of Britain

The fate of the western world was largely determined by the actions of a few people in Britain in the late summer and early fall of 1940.  Poland had fallen, Belgium had fallen, France had fallen.  Great Britain was next.  Prime Minister Winston Churchill lobbied President Franklin Roosevelt for help.  But, Roosevelt was reluctant to fully support a Britain that looked doomed to defeat.

Air power had come into its own and the Nazis planned to use it to clear the way.  Key to their plan was destroying Britain’s air force.  Once it was gone, a seaborne attack would follow.

Winston Churchill knew what was coming and famously announced to Parliament, “We shall fight on the beaches, we shall fight on the landing grounds, we shall fight in the fields and in the streets, we shall fight in the hills; we shall never surrender…”

The air assault started in July with attacks on convoys in the English Channel.  Shortly thereafter, relentless attacks on the airfields.  The British airmen were outnumbered and exhausted.

Eventually, they were also victorious.  The Nazis gave up their plans for invading Britain and focused their attention on Russia.

A number of  factors led to Britain’s success.  One especially important asset was radar.  Britain was an early innovator in Radar and early warning technologies.  But, as we shall soon see, radar alone wouldn’t win battles.

7 December 1941 – Pearl Harbor

Around 7 am on the 7th of December, 1941 the Japanese arrived.  In a brilliant piece of seamanship and aviation skill they sneaked  up on Pearl Harbor, Hawaii.  A few hours later the US fleet lay in shambles.  While a few US fighters managed to take off and tangle with the attackers, their effect was negligible.

Attack on Pearl Harbor

After the attack, US President Franklin Roosevelt gave a stirring speech that contained the following, memorable line, “Yesterday, December 7th, 1941 — a date which will live in infamy — the United States of America was suddenly and deliberately attacked by naval and air forces of the Empire of Japan.”

It was over a year since America’s ally, Britain, had prevailed in the Battle of Britain.   The Americans also had radar.  In fact, American radar detected the Japanese attack on Pearl Harbor.   So, why was America so soundly defeated at Pearl Harbor?

Type of Radar at Pearl harbor

The answer lies not in radar but in the military’s ability to take the data from radar, assemble it into a meaningful picture, and then act on the result.  The British had developed an entire information network that gathered data from radar, and other sources, assembled it at a central point, and then distributed it, as necessary to commanders in the field.  Each member of the team knew their job and knew how to use the information delivered to them.

At Pearl Harbor, on the other hand, the Americans hadn’t built the information infrastructure.  They captured the raw data but didn’t have the mechanisms for rapidly assembling that data into a coherent picture.  And, they certainly didn’t have the plans for how to act on that picture.

Business Intelligence

This brings us to the concept of business intelligence (BI).  What makes for effective BI?  It’s the ability to gather the  relevant data, integrate it into a coherent picture of reality, and the processes necessary to execute based on what that picture shows.

In large part, organizations already have the data – they have the radar.  Where?  In their operational systems.  In the ERP, sales, manufacturing, legal, HR, and other systems that run the business on a day to day basis.  Each entry, each transaction provides a pixel in the overall picture.

But, most organizations are more like the Americans in 1941 than the British in 1940 – they don’t have an integrated view of that data and they don’t have business processes to take advantage of it.

Which leaves me with a question for you… 1940 Britain or 1941 America, where are you?

Next post… Using the Data From Business Intelligence Radar

In the Battle of Britain… radar was the eyes of Fighter Command… its effectiveness was greatly enhanced by being only one element of, and integrated into, a sophisticated command and control network which received the raw information of radar plots and rapidly applied it to direct the use of precious resources of pilots and aircraft to the best possible effect.

Courtesy Imperial War Museum, London

GM’s bankruptcy wasn’t a big surprise – it simply represented the nadir in a twenty year decline where the company’s shrinking market share and excessive labor costs had been well publicized.  In contrast, Toyota’s current sudden acceleration disaster seemed to come from out of nowhere.  In January the company was the industry darling, protecting its customers and our environment with safe, reliable vehicles.  Come February, the brand had become toxic, rapidly obliterating billions of dollars of shareholder value with a devastating impact on long-term goodwill.

But, were there really no warning signs?  I think that highly doubtful.  If Toyota is like the companies we work with, the indicators were there, in corporate systems, though probably lying dormant just out of executive visibility.  Imagine if six months ago an astute manager had correctly interpreted them and stakeholders had taken corrective action.

How does an organization develop an early warning system?  As a last resort, it’s easy to find signs of impending trouble in a company’s legal case matter management system.  These systems support corporate legal departments by capturing information about the content and status of the company’s legal case portfolio.  While each case will follow its own trajectory, in the aggregate the litigation for particular conditions will follow identifiable, predictable patterns.

A great example is the litigation trend that followed the auto industry’s introduction of standard passenger airbags.  Shortly after the technology was widely deployed, trends became visible in the product liability litigation that followed.  Tracking these trends, Federal standards were amended and companies introduced second generation (low powered) airbags, weight-based deactivation and instrument panel indicator lights.  They were also better able to plan and budget their legal costs.  Similar data must have been available at Toyota.  Imagine the billions of dollars, not to mention the innocent lives that could have been saved, if only someone had been looking.

But before troubles hit the legal case management system, they are often captured in a variety of other systems used for daily operations.  Great places to look include warranty claim systems and call center systems.  Just as radar provided the ability to foresee and react to threats during the Battle of Britain, wise companies will integrate information from these systems to create corporate early warning systems that will provide executives with the ability to foresee and react to threats on their company, preempting the worst fiascos.

Developing an early warning system doesn’t even have to be expensive.  Companies already own the data.  What’s needed are the critical steps of highlighting trends that predict trouble and making those trends visible in the form of management dashboards and automated alerts.  Start small, but start now – you can always expand later.  Can you imagine the ROI on this relatively small investment?

Create your own early warning system, transforming your call center, warranty and legal expenses from overhead into investments in a valuable strategic asset that can stave off crises before they occur.  It doesn’t take advanced analytics to know that Toyota’s current calamity could have been avoided had they had the foresight to do so.

Give us a call to talk about how an early warning system can help you reduce the risk to your organization.

A quick example of how BI & BP (business process) should go hand in hand…

Working with statisticians from Central Michigan University, we’ve jointly developed PVForecaster, a product that allows hospital administrators to forecast their patient volumes on an hour-by-hour, specialty-by-specialty basis.  Tactically, the result is better control over scheduling and costs.  Strategically, PVForecaster helps administrators forecast changes in demand and react accordingly (e.g. build more capacity, etc.)

The PVForecaster  business process design is a good example of how BI (and predictive analytics) applications should integrate with business processes.  In our case, rather than just accepting data and ‘spitting’ out forecasts, users drive the system as follows:

1) A planner uses PVForecaster to develop a number of forecast scenaria for the hospital.  For example, what is expected to happen if we increase our advertising budget?  What if we reduce it?  What if our operating territory’s population increases by 2%?  Stays flat?  Decreases?

2) The planner uses the BI tool to  review these scenaria with hospital executives.  Together they decide which scenario will be the one that the entire hospital will work from.  They use a routine in the BI application to designate this as the ‘working assumption.’

3) Department managers from across the hospital are given access to the ‘working assumption’ forecast in the BI tool.  They can use the tool to slice and dice through this forecast along all the relevant dimensions (e.g. date, specialty, time of day, etc.).

4) Using this information, the department managers can plan accordingly.

Yes, we could analyze data and build forecasts without implementing this business process.  But, the tools may not ever get used.  They may become the unused toy of some isolated planner.  On the other hand, implement a business process wherein all the relevant parties have, and need, access to the BI tool to do their jobs and you’ve developed a powerful tool that will get used and will provide enormous value.

Do you have any other good examples of business process / business intelligence synergy?  Put them in your comments.  Thanks!

– Ben

It’s a great feeling helping a client understand their data and working with them to analyze it to get to an ‘a-ha’ moment.  Since Dataspace’s founding 15 years ago, our leaders have seen pretty much every technology that helps us help our clients.  And until recently, our CEO would comment, “they’re all pretty much the same.”  Well, he’s got a different set of talking points now.

You may have seen a few of our posts on the merits of QlikView, and now I’m proud to announce we’re Michigan’s newest QlikView partner.  Let me tell you why I’m excited.  Trite as it sounds, QlikView really is different.  Well, maybe it’s not QlikView that’s different, maybe it’s that using QlikView is a completely different experience than using other leading BI tools.  I’m not talking about features, technical architecture,  enterprise deployability or things like that – I’m talking about how, at the most basic level, using QlikView is different, and here’s how I sum it up: QlikView allows analyses that follow the way your brain thinks, not the way the data is organized.

With traditional tools you get some data, format it a certain way, and then use some kind of analysis and reporting tool to view it different ways.  If you find that you missed something, you need to go back and get more data.  If you find you have the right data, but it’s not formatted so the tool is optimized, you need to reformat it.  All this means that to use the tool, the user must bow to the data.  It makes free-thinking difficult, because if you find you want to look at the data a new way, you need to jump through hoops to get the tool to do what you want it to.  Even worse, if you need to rely on IT to reextract and reorganize the data every time you want another analysis, good luck making friends with them.

With QlikView and its database structure, you load all the data at once.  You don’t have to create cubes or other views on which to perform your reporting and analyses – QlikView’s application lets you drill down, up, sideways, it doesn’t matter – it’s all there from the start.  So, if you’re investigating which products are most profitable, and realize it would be great to see which customers buy those products, with one click they’re identified.  Want to see which products one of those customers buys?  One click to reset the products and one click to select the customer, and all the information updates again.  No more cubes, no more incremental fetches, no more bowing to the way the data is structured, no more IT SOWs.

Let your BI tool help you uncover the facts as your brain dictates.  Give QlikView a once-over.  Contact us if you’d like to discuss further.

Last month Ann Arbor’s long established newspaper, the Ann Arbor News, shut down.  But, it was replaced with a web / twice weekly hard copy publication.  Little did I know that Dataspace would be one of the first companies profiled in the new business section.  To see what they have to say, click here.

I, also, wanted to add a little ‘subjective’, business intelligence quiz here.  The answer is really open to interpretation so it will be interesting to see what people think.  The question is: What was the first business intelligence tool?  Why?

Feel free to add your opinion as a comment.

Thanks for following us!

– Ben

Business executives already know our secret.  The best business intelligence systems are not about the system.  The best business intelligence systems are about the business.  Sometimes it’s up to us, as IT professionals, to remind ourselves of this reality.  When talking to a business executive, don’t use keywords like analytics, data mining or data warehouses.  Instead, talk to them about their business.

Ask them if they know which of their customers are the most profitable and which are actually draining resources.  Ask if they want to find ways to reduce the amount they spend in legal costs.  Ask if they want to be able to predict the future demand for their products and services, so they can match capacity and staffing levels accordingly.

As experienced IT professionals we know that the way to provide these answers is to use the data already being collected by organizations in their operational systems (like ERP, CRM) and present it in new, visually appealing ways, with Business Intelligence (BI) tools.  But, occasionally we need to remind ourselves that no matter how cool the technology (to us, even that first program we all wrote that displayed the words “Hello World” was really cool), that’s not what sells BI.

Executives appreciate technology, and many are quite savvy, but when it comes to how they spend their day, they’ve got problems to solve, opportunities to capitalize on, and stakeholders to please.  To them, the best systems are like dishwashers – tools that get a job done.  Executives are not interested in how the dishes get clean, just that they do get clean in a fast, reliable, budget-friendly way.

In other words, executives are interested in the benefits of BI, not how it gets delivered.  So, the next time you’re discussing BI with an executive, sell the benefits, not the tool.  Sell the value of sales force ranking, not the BI system.  Sell the patient volume forecast, not analytic algorithms.  Sell the ability to direct your valuable purchasing dollars to the lowest cost vendors, the ability to have your sales executives use their limited time to court the most profitable clients, the ability to gauge the effectiveness of your latest promotion… you get the idea.

I’ve gotten some great feedback on my Blog posts and I’d love to hear your input.  Feel free to add a comment or email me directly at btaub@dataspace.com.

Just a quick note here, a few weeks ago I wrote the IBI was a sensible player in the BI space that is sometimes overlooked in BI evaluations. Well, to follow that up, Dataspace has just joined the Information Builders (IBI) partner program. For more information, check out the link to the press release on our home page: www.dataspace.com.

In my discussions with customers and others in the BI field I am hearing a lot of anguish about software vendors.  Many of the larger BI players have been sucked up into bigger companies and I’m hearing that these BI vendors, which were never customer-friendly to begin with, are now becoming downright customer-hostile – even with customers who have paid maintenance for the privilege of obtaining technical support.  Take a look at this article from CIO Magazine for more detail on the turmoil, and this article on Business Intelligence vendors to watch.

BI software vendors used to be grouped into the top three or four players, with open source and smaller players taking a far back-seat.  Things have changed.  Every week I’m seeing product announcements from new entrants.  Some of the tools look really good.  One I’m a big fan of, despite some limitations, is QlikView.  They’ve been around for a while, but are still relatively new compared with long-time BI players.

As consultants, we have the opportunity to work with a great group of clients, and we apply our expertise regardless of the technology they have already deployed – even if its one of the aforementioned customer service-challenged companies.  What to do, though, when a client asks us to recommend a new technology to help them rid themselves of their BI headaches, or make a new start in the field?  We’ve learned that the company behind a technology is as important to a successful implementation as the technology itself.  After all, you can get management dashboards, alerts, statistical analysis and basic reporting from virtually all of the established players.

So, who to choose?

Of course it depends on the application, but if you’re looking for enterprise-class reporting, dashboarding capability, the ability to email basic business intelligence to a broad audience, reasonable licensing models, and an independent vendor partner with a great reputation for customer service, the answer could very well be Information Builders, IBI, the creators of Focus and WebFocus.

What?  Isn’t IBI an old-line mainframe reporting solution?  That’s what I thought.  But, then I saw the technology demonstrated to one of my clients.  Is it perfect?  No.  Other major vendors like Business Objects, Cognos and MicroStrategy each have their unique advantages and disadvantages.  But, IBI really seems to ‘get it’ in two very important ways:

1) WebFocus provides flexibility but also comprehends a fact I’ve been talking about for a long time: MOST USERS DO NOT CREATE REPORTS FROM SCRATCH.  In most BI environments reports are created by IT and, perhaps, a few power users.  The vast majority of users don’t want to and don’t need to create reports from scratch.  (For more on this see my post on tying the BI tool to the user. )

2) IBI customers are genuinely happy both with the tools AND with the vendor.  As with any piece of software, tech support issues do occasionally arise, but the feedback I’ve received is that when they occur, they are given top priority.  Can you say that about your BI vendor?

We were so impressed with the research we’ve conducted, that we’re considering partnership with IBI.  If you are evaluating Business Intelligence tool sets, I recommend that you consider them.

– Ben