Frank Buytendijk Blog

Check out www.biguru-online.com from Jorgen Heizenberg. In a recent post he discusses the 10 Truths about BI.

1. We need one version of the truth

2. BI projects require a business driven approach.

3. BI development should be done incrementally.

4. BI projects needs high level sponsorship.

5. BI = decision making process.

6. BI is going to be real time (BI 2.0).

7. Information democracy: BI for everyone.

8.  BI cannot be outsourced.

9.  Merging: BI, CI, EI, EPM, CPM, BIM...

10. BI is the answer to everything.

In this post Jorgen contradicts the bottom 5. Let me pick a few and provide some input too.

BI cannot be outsourced

Many BI professionals claim BI cannot be outsourced, it is too specific and too nteractive. I think that is not the case. In fact, if you lack the skills needed for BI, outsourcing can be the most strategic thing to do, to benefit from BI. In fact, if you outsource things, including BI, it forces you to think how to manage the process, and keep control over it. In many organizations where BI is not outsourced, there is no such clear view.

BI is about Information Democracy

Do you really think people are looking forward to a transparent business where your contribution to the overall results is really visible, and where everyone has access to the right data? I would argue most managers do not. Knowledge is still power. It is not in their best interest to live in a transparent world. Other than that, you can argue that for REALLY strategic BI you need to have insight that others don't have, so that you have a first-mover advantage. BI for the masses may contain all the bread-and-butter information, the real stuff is still very exclusive of nature.

All BI will be real-time

Most problems take care of themselves, if you give them some time. Think about when you have been out of the office for a few days. First, you'll find an email saying "Hi, I really need your help urgently," and a few mails later "Never mind, problem solved." There is a hierarchy in an organization for a reason, to deal with stuff that never reaches the level of senior management. Let's assume they all of a sudden have a real-time BI system in which they see all these little details. The shock will be enormous, "where does this mess come from?" Once problems reach the executive level, executives are used to firm responses, and this will be the case here as well. But these little problems were always there, and were dealt with in a normal way. Too much real-time information will not give you additional control, you'll lose it.

BI supports the decision-making process

Yes, but amongst other things. BI also supports justification, external reporting to shareholders, regulators and society. And the 3^rd purpose of BI is to influence people's behaviors. Providing feedback motivates people. Purely using BI for decision-making processes greatly limits your business case.

In general, I don't believe in "truths."  I believe in thinking for yourself. My best practice could be your worst nightmare. Insight comes from challenging your beliefs and assumptions. What is your Top 10 of BI Truths and Myths?

frank

I was a member of the jury of the Data Quality Award recently. One particular thing did strike me in reviewing the cases that were entered and in listening to the presentations of the nominees. When asked what they thought they could further improve, almost all nominees responded that they thought data quality wasn't on the executive radar screen enough. There was not much executive commitment.

When you think about it, this is actually rather understandable. First of all, every project in the organization claims to be strategic and in need of executive commitment. In a larger organization, I bet that if you add up all projects that claim time on the executive calendar, you end up with working hours of more than 24 hours a day! Second of all, I think it is also the nature of the subject. There is 2 ways of doing data quality: perfect, which is normal, or bad. There's hardly any way to "score" with the subject, other than when the status quo is really horrible and blood is coming out of the project in every direction. It's like the salary administration, crucial, but you can't really score points with it, perfection is the norm. Fortunately, perfection is usually the case with salary admin, however unfortunately, it is rarely the case in data quality initiatives.

I think data quality professionals mostly have themselves to blame for not being on the executive radar screen (unless they fail). Data quality projects are usually too focused on getting the customer addresses right. As important as this is, it doesn't "wow" people. Data quality should evolve beyond the customer address. Data quality professionals have a specific set of skills. They are good in detecting incorrect or incomplete data, they are able to fix it, and they are able to connect different data sources. Here are some ways to leverage those skills in more strategic ways:

• 
  
  Make a customer value proposition out of customer self-service. Think of advance check-in for flights, promotions to register for, order status checking, anything that makes it the best interest of the customer to deliver perfect data. The moment data quality is connected to new ways of interacting with customers, it becomes more strategic.


• 
  
  Broaden up the scope of data quality. Master data management is a topic very close to data quality skills. Make sure your skills are indispensable in communicating throughout the complete value chain, including partners, suppliers and customers. The moment perfect master data optimizes the value chain and contributes to operational excellence, it becomes more strategic.


• 
  
  Make sure you are part of Business Development initiatives. If with your specific skills you are able to piece together data during merger and acquisition due diligences, and your insights on previously unintegrated and "soft data" lead to better decisions, or equally important, faster execution, your skills have become extremely strategic.

These are your career tips for the week!

 frank

Not even the best performance management initiative can predict the future. However, it can help you getting ready for it. One of the techniques that help becoming future-oriented is scenario analysis. According to wikipedia, scenario analysis is a process of analyzing possible future events by considering alternative possible outcomes, called scenarios.

Scenario analysis is becoming more important as part of performance management. Organizations do not operate stand-alone; many decisions made elsewhere in the value chain or in the market impact our performance. Today, in most organizations we cannot budget and plan based on just our resources. Our external stakeholders ask not for our budget, but for our projections, our forecasts, and our guidance to the market. If you miss external trends that impact your business, and you do not make your guidance, this has serious repercussions. Your stakeholders lose confidence in the capabilities of the management, the share price may be affected, and ultimately the agencies lower your ratings, leading to an increase of the cost of capital. How's that for a business case for enterprise performance management and scenario analysis?

There are many techniques for scenario analysis. Some use narratives, stories that describe "the day in the life of manager John" in 2009, going through different experiences based on what he decided in 2007. Narratives can also be a bit bigger, such as the future of the telco industry 2020, when (if?) literally every device is connected, or the world at large in 2100 if we don't address global warming. Other techniques use complex quantitative techniques, based on stochastic or Bayesian modeling, assigning probabilities on certain outcomes and outcomes of outcomes.

Whatever "school of thought" you belong to, scenario analysis is about thinking the unthinkable, questioning the most foundational beliefs in your business, and see what happens. It is not even important if one of those scenarios becomes reality. The real benefit of scenario analysis is flexing the mind and being ready for any change, understanding that the future begins now and we should come into action, or seeing today's word simply in a broader scope. Predicting the future can be done best by shaping it yourself!

In this series of blogs I will describe three ways of "predicting the future":

• Trend - Counter trend. Every action leads to a reaction. Every thesis leads to an antithesis. There is never simply a single trend that extrapolates endlessly. There is always a complexity of trends in different, sometimes opposite directions, that lead to a balance, an equilibrium. Every time the balance is disturbed, different trends compete for a different balance and the equilibrium is restored. What happens is we look at today and describe the opposite situation? I will post on how a counter-reaction to user-friendly software could lead to "power to the nerds."
  
• Extrapolation. Although there is never a single trend extrapolating eternally, trends do sometimes extrapolate for a while. What happens if we take today's situation, understand how we got there, and logically think through the next steps? I will reason how Service Oriented Architectures could make business consultants out of IT vendors and business integrators out of system integrators.
  
• Comparison. Global trends are not restricted to a single industry. Privatization in the energy-business will most likely lead to the same effects as it has in privatizing the post office, or the railways. Or privatizing the police or the army, or the treasury department, all would be unthinkable today. What happens if we apply the dynamics of another industry on our own? How can Search benefit from understanding a human shopping experience?

Sounds a bit too conceptual? Read the following blogs over the coming weeks. I have chosen a nice contentious subject, "The Future of IT." See if you agree, disagree, if it makes you smile of even upsets you. If so, scenario analysis has worked.

frank

In this series, I use scenario analysis techniques to "predict the future." Please also read the introduction for the background of this series. Smallprint: these blogs on how the future of IT will look like do not necessarily reflect the opinion of Oracle Corp.; they don't even necessarily reflect my own opinions. They are an exercise in scenario analysis, to broaden the mind and to help you shape your own future.

IT used to be the arena of experts. With a double PhD, wearing lab coats. They would talk in a language no one understood (although it vaguely resembled English), and computers were off-limits for most people. Today's different. Most people I know have more bandwidth at home than in the office. We all use computers. We're all online. But we don't program a lot ourselves. We just use 'em. In fact, we expect them to work as our iPod: out-of-the-box, perfectly and without reading the manual. Software is being promoted as easy to install and to use.

Well, the future of IT could be different. As a reaction, the trend could reverse. Power back to the Nerds. With all software being easy to use, a low cost of ownership, everyone being able to reap the benefits, where's the exclusivity? Where is the competitive advantage? If we continue like this, Nicholas Carr's words on "IT doesn't matter" become a reality.

In the future, software should become harder to use, require more programming, some software will become much more expensive, as companies require some exclusivity. Bespoke systems will make the difference, which is the ultimate price of software, being built for a single user or company only.

In this future, Open Source can go both ways. Either open source becomes a commodity and people use it just for non-competitive stuff such as printer servers. Or it is just a basis, and open source developers adapt it to bespoke software with unique and exclusive features and functionality, not used by anyone else.

The advantages of such a model are evident. Whoever has the information has the power. In an information democracy, the people have the power. In an information monarchy, the rulers have the power. And in an information-centric business, this is true competitive advantage. If transparency is forced by legislation and regulation, the need to exclusive systems is even bigger. If everyone has the same information, only the ones who have superior insight benefit.

It's the nerds who hold the power in this future. They are the ones making more money than the CEO. They are the only ones who can build and control these complex systems. The lab coats are back, this time just with some pizza stains added. It's not money or muscles that count, it's brains. The most popular guys in school, admired by all cheerleaders? You got it, the nerd! The guy or girl who can get you the concert tickets you want, who can help you with your homework, who can get you connected with everyone and everything, and who will make a ton of money.

In this series, I use scenario analysis techniques to "predict the future." Please also read the introduction and "Power to the Nerds"  for the background of these series. Smallprint: these blogs on how the future of IT will look like do not necessarily reflect the opinion of Oracle Corporation; they don't even necessarily reflect my own opinions. They are an exercise in scenario analysis, to broaden the mind and to help you shape your own future.

IT is changing. Service-oriented architectures break up application silos into many reusable software components. Within a certain framework, these components can be used to put together new systems, and making changes is done by simply reassembling these components in a different way. This is a partial reality today, but it is not hard to imagine this happening across a complete domain of business applications. Let's extrapolate this trend and see where it could further lead.

First of all, in this future this will create a different class of workers. Not IT people, as they are business experts in their industry, and not business people, as their task is to build and maintain systems. Where's the difference between building and maintaining systems? It's not really there. Lines between these two IT disciplines will blur. Prince2 as a development methodology and ITIL as a maintenance methodology will need serious updating as application development and management change dramatically.

At the same, time, business models and business innovation is becoming more information- and IT-centric. Mass customization principles (mass production, in which every item may have different specifications) today already make that consumers can configure their own car, or their own sneakers, or their own health insurance. And why stick to combinations of one producer or supplier? It is easy to imagine that customers (first business-to-business, then business-to-consumer) would want to assemble their own systems as well, for instance for integrated travel planning (air, hotel, car), multimedia experiences from various sources, games spanning multiple "worlds," or personal administration. All through standard services. You could say that IT has become the business itself.

In this world, the role of software vendors will be changing. First of all, they don't sell applications anymore; they sell a platform on how to compose, decompose and recompose services. Their brand is not about the company anymore, but represents the overall community of small vendors adding components (why would they come from one vendor) and best practices from consumers on how services are being used. In creating and running these communities, the role of the software vendor most likely even turns into that of a business consultant, with an intimate understanding of industry-specific processes. Even more, the software vendor monitors these processes running at customers and provides advice on how to optimize them, helped by benchmarking based on comparable processes. This exists today already on how well systems are running, but this would evolve into process monitoring.

The role of systems integrators changes too in this scenario, they could become business integrators. Often system integrators have a vertical focus already, they are organized in industry units. They often have large outsourcing practices, in which they run their customers' systems. They often serve multiple customers in an industry, and in case on value chain integration, act as an project integrator between all parties. From system to project to business integrator is just a small next step. Business integrators have a central spot in a certain industry or value chain, and make sure the partners connect, to run a seamless process. They are not an IT service provider, they are deeply embedded in the vertical itself. Think for instance of an IT company to which a banking system is outsourced too. The only thing missing being a bank itself is perhaps the banking license, something that could be fixed by a cooperation with a bank or with multiple banks. This may very likely happen today already, although I am not aware of any specific examples and what legal possibilities or boundaries there are. A system integrator working in the telco industry could come up with services to connect networks better, and could start sell these services to consumers directly. How many internet providers have gone into "content" already?

Stock investors always warn that results of the past are no guarantee for future results, but for extrapolation the opposite counts. The signs can usually be spotted throughout the complete trend. First attempts may not always be successful, it usually takes a few rounds to get the next step in business right. Not every internet provider that went into content was successful. But success is the summary of all failure. Extrapolation of trends helps you keep that faith.

frank

In this series, I use scenario analysis techniques to "predict the future." Please also read the introduction, "Power to the Nerds," and "Software Vendors as Business Consultants for the background of these series. Smallprint: these blogs on how the future of IT will look do not necessarily reflect the opinion of Oracle Corporation; they don't even necessarily reflect my own opinions. They are an exercise in scenario analysis, to broaden the mind and to help you shape your own future.

Everything has to be search-driven when working with computers. The effect of search engines on IT, and specifically user interfaces, has been enormous. It has even led to enriching the language, as we "google stuff." I wonder how long it takes until the verb "googling" is used even for physical searches. "Honey, the cat walked away, I'll be googling the neighbourhood now."  Too bad the airlines can't google my luggage yet.

Is keyword search the ultimate way of finding information we need? Comparing search with a human shopping experience says it's not. The future of search could benefit from lessons from today.

Let's first look at how a human shopping experience would be like if it worked like keyword search. Imagine you are going shoe shopping. You go into a shop and show the salesman your shoes, and you tell them you want the same shoes in black. If it were search the salesman would either immediately say "no" (without even saying sorry, or that he could order them), or would come back with those black shoes in every size between 6 and 12, because you forgot to specify your size (how dumb was that?). In both cases the result is the same, you leave the shop. You no new shoes. The salesman no revenue. Loss-loss situation.

How does it really go? If the salesman has the shoes, he will ask for your size, or will measure it, and then will go. He'll also try and sell you shoe polish, a set of shoe spanners, and matching socks, although you didn't  ask for it, but wanted it anyway. If he doesn't have those shoes, he'll tell you he'll check for a minute, and come back with something like it. In the meantime, you were walking around the shop and found another pair of really cool brown shoes too. The salesman will point out that the shoes you asked for were last season's anyway and that he has brought the latest model, something that fits much nicer with the fashionable clother you are wearing (hmm, what a nice person, you can't help thinking). You leave the shop with two pair of shoes, both better than the ones  you thought you wanted, shoe spanners, socks and shoe polish. Win-win situation.

Search today is unidirectional. The supplier of information is the slave, the customer decides everything, is master. This is not in the best interest of both. Information demand and supply won't meet in an optimal way. The future of search is like in human interaction, bidirectional. This means that search engines need to become aware of what search process they serve, and ask the right questions back. It means that the search engine needs rules on how to interpret the answer, as much as how to interpret the question. Searching for an LCD TV with certain specifications, could lead to the search engine producing the result of a Plasma TV, because for the same price and specifications, the margin is higher (in the best interest of the supplier of the information). If the desired vacation in Greece is not available, perhaps a similar location in Turkey is a good option. If you search for a second-hand Jaguar in black, with not more than 100,000 kilometers, perhaps the dark blue one with 101.000km is fine too.

Innovation, to shape the future, often comes from comparison with other industries or similar processes. The future of Search is no exception.

frank

We looked at three different ways of how to create a futuristic scenario, in this case the future of IT. I described how the value proposition of IT could become more exclusive again, instead of aimed at user-friendliness. How a trend can lead to a counter-trend. I described how the role of software vendors and systems integrators can change, if we extrapolate the trend towards service-oriented architectures. I lastly described how the future of search engines could look like, if we compare it with a shopping experience.

Are the "Power to the Nerds" and "Software Vendors as Business Consultants" conflicting? The first scenario describes that IT is becoming harder to use, for it to make a difference, while the second scenario stated it becomes easier to use to make a difference.

First of all, great if they are conflicting. It shows there are multiple futures, and it should spark a discussion which future you like best. And trends and futures can coexist. Different strategies to reach the same. IT becomes the business itself. Some car manufacturers, like Volkswagen, are very successful in producing for the masses, others, such as Lamborghini, produce for an exclusive group. Oh and by the way, the Volkswagen and Lamborghini brands belong to the same group.

Of course, this exercise is not complete. I only used one example and one technique at the same time. Multiple techniques should be used to draw multiple scenarios. The possible futures that I described have in common that they all point in the direction of an increasing importance of IT. That in itself is nothing more than an assumption, that could be questioned. Anyone up for producing that scenario, as I couldn't imagine it?  ;-)

frank