An excellent article in Health Data Management by Greg Gillespie gives some wonderful examples of the power of analytics to improve health outcomes, looking at data from some of the 2,000+ clinical trials that Cleveland Clinic is currently running.

I strongly encourage you to read the original article (also available in pdf format), but here are summaries of the three use cases highlighted: hand-washing analytics, central-line analytics, and blood-transfusion analytics.

Data Transparency and Hand-Washing Compliance

Cleveland Clinic uses SAP BusinessObjects Dashboards (Xcelsius) to display information, change behavior, and avoid infections:

Cleveland Clinic has developed a program where staff from compliance anonymously watch workers in different departments and record whether they do in fact follow hand hygiene guidelines. Their findings are uploaded into Cleveland Clinic’s enterprise analytics system and are accessible via a dashboard tab.

Four years ago, the system was showing a 40 percent compliance rate with hand hygiene guidelines. Now the compliance rate is staying well over 90 percent, staving off a significant number of hospital-acquired infections and other complications arising from hygiene issues.

“That’s the critical value of data transparency—you can show people what they’re really doing as opposed to what they think they’re doing, and we can show it on a department, unit-by-unit or individual practitioner level,” says Steve Davis M.D. “I’ve found that when you put that kind of information in front of physicians, their competitive streak really comes out. No one likes to get a ‘C’ on their report card, and if you don’t have data everyone assumes they’re getting an ‘A.’ When they find out they’re not, then they get moving.”

[By coincidence, a post in the Decision Factor blog also takes up the theme of hand-washing this week, arguing that data cleansing is the single most important means of avoiding bad decisions. ]

Reducing Infections While Saving Money

By carefully collecting and analyzing data, Cleveland Clinic has been able to reduce infection rates, spend less on equipment, and avoid costs of up to $30,000 per affected patient:

The Centers for Disease Control and Prevention estimate that nearly 250,000 of the bloodstream infections occur annually from procedural issues associated with inserting and maintaining central lines—tubes inserted near the heart or a large blood vessel that are used to give fluids, antibiotics, medical treatments such as chemotherapy, and liquid food.

Overhauling the health system’s approach to central-line infections had a significant financial return in addition to the clinical benefits.

Before clinical and business analytics were applied, each individual unit was responsible for ordering their own lines, which meant that more than 30 different lines (and more than 90 different PICC lines, another type of tube) were being used across Cleveland Clinic, which was not only financially inefficient but also clinically dangerous.

By streamlining the purchasing to one vendor, the equipment and maintenance costs dropped significantly. And standardizing the clinical processes resulted in major cost avoidance—it’s estimated by the Health Research & Educational trust that central-line infections add upwards of $30,000 in treatment costs per afflicted patient.

Best-Practice Blood Transfusions

A blood transfusion dashboard helps identify physicians that haven’t kept up with the latest information in health best practice, improve the supply of blood, and reduces costs:

Andrew Proctor, administrative director of medical operations for Cleveland Clinic has developed a blood utilization dashboard that enables department heads and others to drill down to a physician level how much blood is being used for transfusions.

Standard industry practice used call for ordering transfusions if a patient’s hemoglobin count was below 10 after surgery or due to critical illness, But about a decade ago, says Davis, medical research showed convincingly that blood transfusions given at those hemoglobin values, and even significantly lower, in nearly all cases did more harm than good, providing few benefits and increasing the risks of nosocomial infections.

“Blood transfusions is another area where physician behavior has changed slower than the evidence, and our data is helping drive that behavioral change by enabling us to determine where blood utilization still goes against best practices, and addressing the issue on a unit or individual physician basis,” Davis says.

The result has been a significant reduction in blood utilization, which equates to a significant reductions in costs associated with maintaining the blood supply, and an improvement in patient outcomes.

Health Data + Analysis Saves Lives

I believe we’ve only scratched the surface of what is possible using analytics. New developments in big data mobile, cloud, analytic, and collaborative technology are combining to create new ways of improving health care.

Examples include the new SAP Collaborative E-Care Management application that connects patients, care providers and their families through medical monitoring software and mobile devices to better manage their health with individualized treatment plans:

And the pioneering work being done in conjunction with Charité, Europe’s largest teaching hospital, to enable mobile access to health data anytime, anywhere, including the SAP HANA-based Oncolyzer cancer-research application.

For more information, visit the Healthcare area of SAP.com

I will be the host of the next #SAPChat, an interactive panel discussion carried out entirely over Twitter. It’s scheduled for 9am PST / 18:00 CET on Wednesday February 29th, and the subject will be “Business Intelligence Best Practice”. We will be discussing how best to implement business intelligence projects successfully, with an emphasis on people-focused hints and tips rather than technical discussions. And with only 140 characters available, the advice is guaranteed to be succinct!

You can participate in the discussion by asking your question over twitter, by including the hashtag #SAPChat, or just tune in on the day to watch the information flow. You can use your favorite Twitter tool, or join the action at http://tweetchat.com/room/SAPChat. The goal is to make the discussion as interactive as possible – don’t hesitate to answer questions from other people, or expand on the points made.

If you’d like an idea of what the chats look like, take a look at the transcripts from previous sessions such as SAP’s CIO Oliver Bussmann on the Consumerization of IT and a deluge of questions around SAP HANA.

For the session, I will be accompanied by two world experts on the topic of successfully deploying business intelligence projects:

Alys Woodward

Alys Woodward is program manger for European Business Analytics, Enterprise Collaboration and Social Solutions at IDC. Alys has been in the business analytics space since 1995, and an analyst since 2003. I’ve shared many stages with her over the years, most recently in Russia, and I have been consistently impressed with how much I’ve learned in each presentation. You can find out more about IDC’s research on making BI more pervasive with this research sponsored by SAP:

Improving Organizational Performance Through Pervasive Business Intelligence:

An increasing number of organizations are making BI and analytics functionality more broadly available to all decision makers inside and outside the organization. Internally, more pervasively available BI solutions lead to greater accountability by all employees and greater consistency in performance management. Externally, relationships with supplier and partners can be strengthened through effective sharing of key performance indicators (KPIs). However, having pervasive BI means more than having the appropriate BI tools distributed to all stakeholders. In pursuit of pervasive BI, organizations should focus on the five key factors that can be directly influenced to increase diffusion of BI. They are:

• Degree of training on the data, tools, and analytic techniques
• Design quality of the BI solution
• Prominence of data governance
• Nonexecutive involvement in promoting the design and use of BI solutions
• Prominence of a performance management methodology

You should follow Alys on Twitter: http://twitter.com/alyswoodward

Cindi Howson

Cindi Howson is the founder of BI Scorecard, a resource for in-depth BI product reviews, based on exclusive hands-on testing. I have known Cindi ever since she was the manager responsible for rolling out business intelligence at one of Business Objects’ largest customers, many years ago. Cindi carries out a key industry survey each year which gives insights into how successful organizations are with their business intelligence efforts. Cindi is a TDWI (The Data Warehousing Institute) faculty member and a contributing expert to Information Week. She has been now been advising clients on BI tool strategies and selections for more than 15 years, and wrote my favorite BI book:

Successful Business Intelligence: Secrets to Making BI a Killer App.

Business intelligence has the power to change people’s way of working, to enable businesses to compete more effectively and efficiently, and to help non-profits stretch their dollars further. The book draws on exclusive survey data and real-world case studies of BI success stories to identify proven BI best practices you can put to use in your organization, including:

• Gaining executive support and aligning your BI strategy with business goals
• Organizing BI teams and experts for success
• Choosing the best BI tools to meet user and business needs
• Improving data quality so decision-makers trust the BI solution
• Finding the relevance of BI to all employees, including front-line workers
• Using agile development processes to deliver BI capabilities and improvements at the speed of business
• Measuring success in multiple ways

You should follow Cindi on Twitter: http://twitter.com/BIScorecard

As you’re thinking of questions to ask, take a look at some of the great blog posts on the SAP Analytics blog and the Decision Factor blog.

Finally, if you’re interested in this session, you’ll be interested in this presentation I did late last year called “Why BI Products Fail and What To Do About It” (also available as slides: Why BI Projects Fail — PowerPoint Version)

Here’s a great example of a BI competency center (BICC) explaining and promoting BI to the rest of the organization. In this case, it’s SAP’s own BICC that created the video, destined for SAP’s 5,000+ employees worldwide. It was uploaded by Matthias Wild of SAP’s Global IT group (NOT marketing): BI4ALL in Plain English (but with a slight German accent ).

I love the use of low-tech graphics and great editing — and I’m sure this will increase interest in using BI internally within SAP…

You can hear more from Matthias and the goals of his team in this post on the SAP community network: The Concept of Business Intelligence for ALL (part 1 of 3)

I was asked this question last week at an internal meeting for the BI teams of a large Nordic telecom company, who had invited me to do a presentation on BI trends. Here’s my answer:

“It’s one where that successfully changes
the information CULTURE of the organization”

Business intelligence is about business and people, not information and technology. Information is useless unless you actually change something in the way your organization does business. And technology is useless unless it actually gets to the people who should be using it.

A truly successful BI program is one that not only provides value to the business with every project, but also inspires the company as a whole to push to the next level of information use.

I regularly present on the topic of best-practice BI, with topics like “Why BI Projects Fail and What to Do About It”, where I go through a long list of the BI problems I’ve seen repeatedly over the last two decades. In this post, I’ve extracted the top five that I think make the biggest difference in the long run:

1. Focus on Changing the Business

If you’re in charge of BI, your job is not providing a technical infrastructure, nor information, nor keeping internal customers happy – it’s using information to improving the way the company works.

BI projects aren’t delivered when you have built the data warehouse and started providing the reports to the business people – that’s just the start of the real project of changing the business.

Yes, of course the business people think that’s their job, but it’s the mindset that is important. Focusing on the end goal – even though you are not directly responsible for it – leads to the types of behavior that correlate to BI success:

• It helps you ask why people want information, and what they’re going to do when they get it – which in turn helps focus business people who may only have a vague idea of what they’re really trying to do, and what is possible
• It helps you learn and understand your company’s business
• It helps you prioritize projects, make the right tradeoffs, and dedicate resources more intelligently
• It helps you explain the value of your projects in business benefits, not just cost or productivity savings

As an example: somebody comes to you with a strong business need for better data, but the only way to achieve it is through manual information gathering and spreadsheets, and it doesn’t have anything to do with your existing DW or BI infrastructure. Is it still part of your business? Yes! (note this doesn’t mean that your team necessarily does the work).

You should be THE go-to person in the company that best understands both the business information needs and what’s feasible. You should be a clearing-house for best-practice “better run business through better information”, using whatever means are necessary.

2. Focus on People

The figures to the left came from a old IBM survey about IT in general, but ring especially true for business intelligence projects. We spend over 90% of our time on data and technology, while 75% of project success or failure depends on people, process, organization, culture, and leadership.

BI is a crucial interface between the tens of millions of dollars invested in your information systems over the years, and the people who are in a position to unleash some of the value in that investment.

Ultimately, BI projects never fail because of technology alone. Things go wrong all the time, of course, but it’s only if non-technology factors like leadership and expectation setting have been neglected that a BI project truly fails.

There are myriad signs that indicate underinvestment in people: the intended audience is disappointed with the solution(and IT replies “but that’s exactly what you asked us for!”); user adoption is systematically under-funded, with little ongoing training; executives don’t understand “why it all seems so hard – I just want these numbers!”; business teams end up downloading information into Excel because that’s what they’re used to; etc. etc.

If 75% of success is about people, why aren’t we spending 75% of our time on it? If your job is world-class BI, you should be spending a lot more time listening, explaining, evangelizing, and leading than you do “implementing”.

3. Provide Some Simple Data Access for Everyone

You must, of course, focus on the BI projects that provide the most value to your organization. B, but there’s one thing we have to learn from the consumer world: the most effective way to build demand for your product or service is to provide something that’s “too simple”, and then create a community around it.

For example, the iPod wasn’t the first MP3 player, it wasn’t the most advanced, and it certainly wasn’t the cheapest – but it was the simplest to use. And that’s why it’s the first MP3 player that most people ever heard of.

We design our BI implementations for our power users, and implement what they want – typically lots of complex data and “features”. Apple designed something for everybody, but keeping the number of features deliberately, even artificially low.

I’m convinced that Apple employed somebody whose job it was to say “no”: “Can we add some more buttons?” “No!” “Can we add search?” “No!” “Custom playlists?” “No!”.

Once the iPod was a success, more features were slowly added, and it’s a formula that Apple has repeated with newer devices like the iPhone and iPad – launching with fewer features than the competitors and aiming for volume first, and then extending.

The Web 2.0 world has followed a similar model: Facebook and Twitter did one simple thing well first, then built a community, then provided more features. Even video games follow this model. They start out easy: level one is about understanding the basic controls, and then you slowly build up skills level by level.

How does this apply to business intelligence? You should aim to roll out some simple analytic information to everybody in your organization – such as travel and expenses, or breakdown of mobile phone bills, or budget spending, or time management. And it should be incredibly simple to use – basic reports, with every fancy option turned off, and with no extra logon required.

Once you’ve done this (and promoted it widely), you’ll find that people soon come and ask for more information, other types of information, and more features. People get used to having information, their expectations get higher, and you’ve started changing the information culture from the bottom up.

Unlike rolling out BI to power users, widespread information to everyone sets up a long-term virtuous spiral of people accessing, using, and demanding information.

4. Tell Stories

Obviously, people need to believe in the benefits of BI if you stand a chance of changing the culture of the organization.

It’s notoriously hard to predict the return on investment on business intelligence projects, because “you don’t know what you don’t know”: having better information reveals new areas for improvement.

As a study by IDC showed, the majority of BI benefits are typically in “business process enhancements” which can be very hard to determine in advance.

By following #1 above, you’ll have a better idea of what the business benefits of your project really are, and be able to take credit for them (Sadly, if a Marketing VP, say, improves campaign performance thanks to improved business intelligence, they don’t often include a big ‘thank you’ to the IT organization when they tout their performance to the board)

But this can only take you so far. For all the insistence on “hard numbers,” executives – like the rest of us – are surprisingly anecdote-driven. Just as charities know that focusing on the plight of one child is more effective than a series of statistics, you need to be able to tell the stories behind the numbers. You need to collect real-life examples of how your projects have helped individuals in the organization transform the way you do business.

Here’s the crucial test: if you suddenly find yourself in the elevator with the CEO of your organization, do you have a compelling, 30 second story to tell about how somebody spotted something in the data (a risk, an opportunity, a problem), and was able to act on it? If you don’t, start researching how people are using the data you’re providing. I guarantee you’ll find a great story (and if you don’t, then you really need to rethink the foundations of your BI initiative).

5. Stick With It

With twenty years of experience to draw on, it’s not that the industry as a whole doesn’t know what best-practice BI looks like. But it can be very hard to put into practice.

BI remains a stubbornly complex, hard-to-simplify technological and business problem. It takes time, the environment is ever-changing, and there are no real silver bullets or shortcuts. Every successful project I’ve seen was the result of professionals doing the right incremental things day after day.

You have to roll with the punches and stay pragmatic. It’s not about having a perfect data architecture (nobody will ever achieve this). It’s about making endless iterative improvements, making the right painful tradeoffs, and picking yourself up after every business complaint.

If it helps, you should realize that while you only ever hear about the problems, business people do appreciate the power of the information that you’re providing. For example:

• According to a CIO magazine review, business executives are actually more likely than CIOs to believe in the importance of technology to the business(they just aren’t that sure that the CIO is qualified/able to deliver)
• Cindi Howson’s book on Successful Business Intelligence includes a survey that showed that 92% of business people said that BI had contributed somewhat or significantly to company performance (which was higher than the number of IT people that considered their BI projects to be moderately or very successful).

Ultimately, the best way to change the information culture of the organization is to lead by example, and being a tireless advocate for better run businesses. Good luck!

Inevitably, Wolfram|Alpha suffers from the same problems that have plagued every attempt to provide answers to human questions using technology since the dawn of computers: semantics, and specifically meaning and syntax. These are nicely illustrated in some of the questions posted in the forum.

Meaning:

• Is Scotland a country, or a town in Connecticut?
• Does “Sun” mean a planet, or is it short for “Sunday”?

Syntax:

• circumference earth / distance london auckland
• Cross referencing doesn’t work

There are technology vendors in both these areas that could bolster Wolfram|Alpha’s capabilities. For example the problem of syntax, although considerable, is now largely solved at least for the types of instructions that Wolfram|Alpha is interested in. And I’m sure it won’t take long before a better database helps the company discover Scotland.

The long-term problem of understanding metadata remains challenging, however – especially within organizations. Indeed, most of the effort of implementing business intelligence projects often goes into trying to determining what people are trying to measure – i.e. which data sources need to be connected to each other, and how common business terms should be calculated. It’s one of those areas that exasperate business users: “how hard can it be to give me sales revenue by product?!” – but the IT department understands that the devil is in the details.

For example, at the SAPPHIRE user conference last week, I attended a session where a presenter from Kroger explained that during their BI implementation they had found six different definitions of “sales” and four of “gross profit” – all of them legitimate in certain circumstances. Trying to figure out which one to use when was one of the key challenges of the project.

Once this hard work is done, though, it becomes a very valuable resource that can then be leveraged throughout the organization. For example, a new prototype is available that automatically suggests corporate charts as people type in a PowerPoint slide. It uses the metadata from the existing BI installation, and combines it with SAP BusinessObjects’ Text Analytics technology to provide something that looks a little like magic to the average business user – see the demo below.

Use whatever resources you have

Sometimes implementation teams have access to internal resources for brochures and other materials – e.g. see the example below for an example of a brochure that was produced for a large French public sector BI project.

Even if you don’t have any budget or access to official resources, you may be able to find and persuade some marketing or internal communication colleagues to give you a hand creating materials and advice about how to reach your internal “prospects.” 

Find your gifted amateurs

Interestingly, the most successful BI communication efforts I’ve seen were not from official teams. Instead, they leveraged the spare-time expertise and hobbies of the staff members (e.g. amateur actors and family film-makers making spoof "ads").

These were far more creative than the average internal materials. Sadly, they were also slightly subversive, and blatantly misused other people’s intellectual property, so while they might be OK for internal use (I’m no lawyer), I can’t post them here on the blog.

Think like Madison Avenue

Play on the existing difficulties faced by potential users of the BI solution, emphasize the benefits of the new system, and give very simple, concrete steps to get people started. And do it in a way that stands out from the ordinary and combines an awareness of natural cynicism about anything to do with IT with the natural optimism of progress and increasing ease of use..

You should plan out a full “campaign” that slowly eases people into the idea of using the solution, building on existing knowledge, one step at a time. It’s better and more effective to build awareness over time, rather than trying to explain everything at once.

Keep it simple

It’s absolutely essential to make the prospect user’s very first step as easy as possible – if you can get people to come to your internal portal portal page, you’ve already had a big win. The next most important step is to make sure they don’t leave before you’ve had a chance to hook them.

Ultimately, branding is about the solution itself, not the brochures. No amount of “marketing” will help if the BI project doesn’t meet people’s needs.

Initially, you should make the product interfaces themselves as simple as you can possibly make them. Turn off anything that might possibly confuse the users – and note that if you’re reading this blog, it’s almost impossible for you to imagine how even the simplest of things can trip people up. If people complain they don’t have enough features, that’s great – you can always unleash more power later.

The power of dashboards

Sexy dashboards, such as those built with Xcelsius, are a great way of generating interest in internal data. I’ve seen internal campaigns that included “mock-up” interactive dashboards that were spectacularly successful – to the point where the project team had to scramble to manage expectations back down. In particular, dashboards can help you reach executive audiences that might not otherwise find the time to pay attention to your work.

Track your progress, and keep communicating

Use business intelligence to track your business intelligence efforts, including adoption rates, and regular surveys to get people’s opinion of the solutions. And don’t assume that you can stop communicating as soon as people are using it. The BI solution will be updated, people will change roles – make sure that ongoing communication is an integral part of your long-term strategic BI plans.

But do you need a BI theme tune?

Jill Dyché of Baseline Consulting posted on the subject of BI branding on her blog, and mused about taking this to the next level, with “audio branding”. If you want to do this, you might be able to find an amateur musician who can help you put together the ultimate BI theme music…

Talking of BI theme music reminded me that I haven’t yet posted the lyrics of Andreas Bitterer and Nigel Rayner’s rendition of “I’ll BI Watching You” featured at the start of the keynote of the latest European Gartner conference.

Get ready to groan, and see previous post for the video. (Here’s a video of The Police doing the real version)

I’ll BI Watching You

Every warehouse you make
All data replicate
Each data mart you create
Every schema you shape
We’ll be watching you

Every report you spool
Every query that’s uncool
Every OLAP you cube
Every consultant you confuse
We’ll be watching you

Oh, can’t you see
You need BI strategy
Performance management takes
Many steps to make

Every user you placate
Downturn you navigate
For the costs you save
And the profits you crave
You need BI

Data’s lost without a trace
It’s somewhere inside your database
You looked around
But there’s nothing you can replace
No need to start another vendor chase
Just keep trying, maybe, maybe, please

Oh can’t you see
You need BI strategy
Performance management takes
Many steps to make

Every user you placate
Downturn you navigate
For the costs you save
And the profits you crave
You need BI