Frank Buytendijk Blog

A reader comments about my blog "I wasn't in the room..."

"Executives are treating the financial element of the four as an outcome, rather than an essential and fairly equal indicator to the other three dimensions, we should agree that financial performance, over the course of developing the remaining tree elements should take a parallel growth, hence finance should become a forth indicator rather than an outcome"

I agree. In fact, how you describe the four perspectives was exactly the way how they were visualized in the original balanced scorecard between 1992-1996. Four perspectives grouped around the overall objectives. Then it changed, when the strategy maps came, and the perspectives were visualized as having cause-and-effect relationships. Growth/Learning is needed to keep your processes up to date, which leads to happy customers, and financial results are the outcome.

What I like better about the original visualization, is that the 4 perspectives can have conflicting requirements, instead of cause-and-effect relationships. There is a clear conflict between growth/learning and process, called "innovation vs optimization", or between financial and customer, called "value or profit".

Thinking about these conflicts is much more profound than figuring out the cause-and-effect relationships (if these even exist, instead of correlations). This is one of the main topics of the book I am currently working on. Interested? Check www.frankbuytendijk.com and take the survey.

frank

(I recently posted a blog on www.oracle.com/innovation, and I thought it would be nice to copy it here too...)

Usually when we use the word innovation, it refers to improving products or services. Sometimes this is a gradual process, like adding functions to a software product or adding more customer service options (like self-service check-in for airlines). Sometimes innovation can be radical, such as the Nintendo Wii, which changed the game console market almost overnight, or flying cars that are currently being prototyped (a process that will take many years).

But innovation doesn’t always have to be products and services. Business models can be innovative as well. Look at the Apple iPod. Although the design and the user interface are very slick, the true innovation is not in the hardware, but in the integration with iTunes. Apple solved a fundamental dilemma in the music business. The existing business model in music--selling CDs--was declining in popularity. The alternative model, downloading music via peer to peer networks, had legal issues. Apply made downloading easy and integrated enough for people to actually prefer this over downloading music for free.

Another example of business model innovation is software-as-a-service (SaaS), or on demand. The true innovation is not in the license versus subscription structure (Capex vs Opex), but in opening up of new capabilities. Traditionally, only large companies can manage full-functionality software and deal with the global complexity. However, globalization and regulation do not differentiate between large and smaller organizations. The software that a smaller organization could manage would not provide enough functionality to handle their own complex globalization and regulatory issues. SaaS and related models allow smaller organizations to run fully functional software. The dilemma has been ‘broken.’

Business model innovation usually is about ‘breaking the code,’ where code would mean the way you usually do things--how you believe things work. Innovation is finding a way to eliminate those limiting beliefs.

This is the same way of thinking I’d like to apply to strategy, which is full of limiting beliefs as well. One of the most common misconceptions about strategy is that it's about making choices, about the either/or questions. As Porter says, either cost leadership or differentiation. Or, according to the value disciplines of Treacy and Wiersema, either customer intimacy, product innovation or operational excellence. Strategy innovation is about finding ways to do and/and, and eliminate those limiting beliefs.

I believe the Oracle strategy has done so. Remember the ‘$1 billion saved’ advertisements from a few years ago, where Oracle saved that much money by implementing its own software in a single instance? The cost savings turned out to be not even the most important result. Oracle’s efforts turned out to be the basis for the current innovative product strategy: ‘best of breed from a one stop shop.’ A complete and open product set, integrated using a common middleware layer. Choice for customers, without the burden of having to integrate yourself. Another dilemma broken. And with the single view of the customer that Oracle has established in its own business, we have global insight into what customers are doing.

There's operational excellence and product innovation and customer intimacy, and we can’t say one is more important than the other. They’re all equal parts in the Oracle mix. Strategy innovation in action.

Recently I attended a presentation by a member of the board of a large insurance company. He discussed the "lean" implementation the company has gone through. Lean is a methodology aimed at continuous improvement, and it originally comes from the manufacturing world. Obviously it can also be used for administrative processes.
One of the key elements of Lean is to define the customer value that the organization needs to deliver, and then to eliminate everything that doesn't contribute to that goal. The problem with that could be that you stop innovating. Customers usually don't ask for innovation, they just want to do business hassle-free. The insurance company took a great approach. They company didn't simply ask customers, it didn't just analyze data, the company observed customers closely.

One example that was given was the process around customers moving house. Simply optimizing that process is simple: once you know the customer's old and new address, you can immediately change that in the system, and send out a confirmation letter or email. Process optimized, costs saved, unnecessary steps eliminated, on to the next process. But hang on, what if you think this through? How to add customer value? Moving house from an health care insurance point of view is much more impactful. If you move to another city, you may need a new dentist, a new general practioner, new specialists, etc. Why not offer the customer to register them at a new doctor, and point them to the web site where other customers rate their doctor, dentist etc. Customers become a community. Another innovative thing, was that in its scorecard the insurer doesn't only measure the improvements in time, cost and quality for the company itself, but it actually has performance indicators on what it has achieved from a customer perspective. As logical as this sounds, this is a rare thing.

Lean and Six Sigma do not always have to be 'in the box', and simply about cost savings and operational excellence. Really thinking through what adds customer value is the key to innovation, and this presentation I attended showed that clearly.

"Infrastructure is everything below the level I care about". For developers that would consist of the network and the hardware, for the CEO of the company, probably all of IT would fit under that definition. And given the fact that data warehousing is such an established discipline, it is an attractive thought here as well. All databases can handle queries, it doesn't matter which one you use to build a data warehouse. Right? Well, not really...

Data warehouse architectures have greatly evolved the last years. Where they used to aim at providing the management information for the managers in an organization, now they serve a multitude of stakeholders, also external to the organization. It is very common to give customers access to parts of the data warehouse, so they can manage their customer relationship with the organization better. The same goes with suppliers, who have access to information to improve their administrative and logistical integration. In short, the data warehouse has become a platform to service the complete value chain. This means that all of a sudden, database security becomes a major requirement. Not every database has the same capabilities here.

The second trend in data warehouse architectures has been to support closed-loop integration. Instead of weekly or daily batch updates, data is flowing in continuously, and is being integrated continuously. Moreover, the transactional systems that feed the data warehouse also become a target for the data warehouse, feeding back integrated information. Think for instance of call center and e-business systems that need integrated customer data coming from multiple product systems. The data warehouse increasingly serves as that central data hub. Updating and querying at the same time asks for database capabilities such as multi version read consistency and fine grained locking (an advanced form of row locking, originally meant to support transactional systems), that ensures that users get consistent results from queries even during updates to the data.

Further, with data warehouses becoming more operational of nature, performance tuning becomes a different game. It is no longer acceptable to have a star-schema-based structure without at least a normalized layer underneath. Closed-loop architectures require normalized data to match transactional data models, when data is denormalized the link with the operational data is lost. Further, performance tuning shouldn't come from denormalization, but from more advanced database and hardware techniques.

At the same time, traditional uses of the data warehouse, analytical queries requiring massive table scans, haven't gone away. The type of queries has become much more mixed of nature. Sometimes short random records are needed, sometimes large joins. Query management has become much more versatile.

Both trends combined (greatly expanded user constituency spanning the complete value chain, as well as operational use) also demand high availability techniques. Database does matter.

Interestingly enough, these exact points are made by vendors of specialty databases optimized for data warehousing only. They use specific techniques such as column-based storage, and proprietary hardware to deliver 'appliances', allowing what is often called 'active data warehousing'. Active data warehousing comprises the forementioned continuously inflow and integration of data. Still, the reasoning that this requires a proprietary database is flawed.

First, and I mentioned this already, because of closed-loop architectures. Transactional systems need to access the data warehouse as well. The data warehouse is not a one-directional street, or an information silo only accessed by BI tools. It has become an integral part of both the IT infrastructure, as well as IT operations processes and skills. Having non-standard hardware and database makes it harder to integrate the data warehouse in such an environment.

The second problem with specialized databases is the need for specialist skills. DBA skills are rare. And with growing demands for scalability, availability, and performance tuning, the required skills have become more advanced. Building specialist skills for the data warehouse only isolates the data warehouse DBAs, and makes it even harder to integrate the data warehouse in the IT infrastructure and IT operations. Having mutually exclusive DBA skills for operational databases and the data warehouse places a burden on IT budgets, headcounts and skills that are under pressure already. Advanced self-management techniques in the database helps bringing the ratio DBA per terabyte of data down. Again, database does matter.

The trends in data warehousing (more types of users, and a mix of operational and analytical use) favor Oracle. With extensive experience in both operational and analytical use, mixed query loads, security, availability, self management and advanced row locking applied to data warehousing are not a problem. That leaves one thing: the advantage of an optimized hardware appliance, for advanced performance tuning. And that's the Oracle Exadata Database Machine, however running the same Oracle database, leveraging the skills you already have.

More than 8,000 attendees (mostly academics), 35,000 papers submitted, and 1,600 papers accepted. Amongst which a paper on 'scenario-based strategy maps' written by Pietro Micheli of Cranfield U, Toby Hatch of Oracle and myself.

The start of the conference couldn't have been better. I check into the hotel, get into the elevator and am standing next to... Henry Mintzberg. He is one of my heroes, his work "Strategy Safari" was one of the best strategy books I've ever read. I was so perplexed I didn't say "hi" or anything. Then again, can the gentleman please have 30 seconds of peace and quiet in the elevator, people are bothering him all the time.

The number of presentations was overwhelming, covering every conceivable management topic. I was happy to see that the conference guide has contact details of all presenters, so we can ask for their papers.

So Pietro, Toby and I showed our work on scenario-based strategy maps in a bit of a different style. Not a formal presentation, but almost a market. We had an 8"-4" board where we attached a number of slides to, and people were wondering around, stopping at stands they thought were interesting. We got good traction, have been busy all the time.

The bookstore was unbelievably big, with a full range of books from most large publishers. Shopped till I dropped!

Oh, and lastly, Henry Mintzberg did a book signing, so I got my autographed copy of his latest book and was even able to give him a signed copy of my book. Silly, but I had to...

frank

Ethics is the philosophical discipline that studies morality. Morality is about what is fundamentally right and wrong, towards yourself as well as to others. I wonder how often 'what is right and what is wrong' is being asked in organizations. It should play an important role in strategic decision-making, particularly in these days, as making ethical mistakes can lead to serious consequences. Most organizations have a code of conduct all employees need to sign. 'Doing the right thing' has become more important than risk management alone.

Many have discussed the idea of what constitutes an ethical organization, or ethical targets and performance indicators. If aggressive cost saving targets leave a procurement officer no choice but to work with suppliers that use environmentally unfriendly materials or even use child labor, that is clearly unethical.

But I have never heard of ethical considerations when designing a process. On the philosophical level, for me, a process is a promise. A process promises that if you use it, the outcome will be timely, predictable, and correct. Processes are often obligatory. You have to use it. If the process itself cannot live up to the promise (because it is for instance too slow), it creates frustration, anger, and in the end lethargy. And it drives people to think of ways to circumvent it. Unethical behavior, yes, but driven by an unethical process.

Considering what is right and what is wrong, ethics in other words, shouldn't only be a strategic discussion, they should be part of every business case, or systems implementation.