SATURN 2011: John Favaro, IEEE Plenary Talk, Managing Architecture for Value

Managing Architecture for Value
John Favaro, Associate Editor in Chief, IEEE Software 

Abstract and presentation materials

These are notes from John Favaro’s IEEE Software plenary talk on May 19, 2011 at SATURN 2011.

Elephant in the room metaphor—deliberately ignoring an impending issue. Everyone knows it’s there, no one talks about it.

One such elephant is business value. Mentioned everywhere, not clearly defined or is pushed onto others to resolve.

In 2005, CEO of Diageo asked at a conference in London, What business value we are getting from IT? It represents half of all corporate spending, yet failure rates are high. 20% spent on issues that don’t achieve objectives (Gartner). Two reasons:

  • Finance – lack of discipline to deal with technology investment. More that 2/3 of CIOs never have to evaluate investments they make in IT. Have no idea if they are making good use of capital.
  • Strategy – IT is easily replicated, and competitive advantage erodes quickly. 85% of investment is back-office, infrastructure stuff. 15% funds front-end innovation, and even that is easily replicable.

What is business value? Buffet = intrinsic value.

Three perspectives on economic profits and tensions: profitability x revenue growth; short-term + long-term, whole vs. parts. Successful management of these three tensions leads to value creation. These are three tensions with which a manager must deal.

Economic value is created by combination of growth and profit, short term and long term, and whole and parts. Yet they seem to function as opposites.

Patterns are a common bond that resolves tension in design and architecture, e.g., time vs. space, flexibility vs. efficiency, extensibility vs simplicity. Patterns make it unnecessary to choose.

How to achieve both growth and profitability? Customer benefit unites both. Customer benefit is the reward that customers receive. The delight that Bosch talked about yesterday. Measure of customer benefit is willingness to pay. That can be measured. Customer relationship management is an advance of the second half of 20th century. Customer focus, though, can result in unprofitable growth. Customer benefit is different. Don’t confuse features with benefits. It’s too easy to add features without adding benefit.

Short-term vs. long-term. Markets always reflect the effect of all available information. Efficient markets separate the past from the future. Only truly new information can influence market movements. There is an Agile metaphor here—an Agile project is supposed to be an efficient project. You ain’t gonna need it (YAGNI) is about implementing information available at a given time. Common knowledge is created in efficient products.

A project with technical debt is an inefficient project. Doesn’t implement everything implied by available information. This is related to broader problem of not balancing short and long term. Market values both the short and the long term.

Common bond to manage for both with success is sustainable earnings. Those earnings not influenced by borrowing between time frames—borrowing from past, present, or future.

Borrowing from present: cutting back excess investment is a way to increase sustainable earnings.

Borrowing from past = milking the cash cow. Take every technological development in your area seriously. Track it and give it funding. Substitution is a key factor in IT development.

Borrowing from future = technical debt, start/stop investing. Invest when prospects are good, not when times are good. Put money aside when opportunity, not times, are good. Don’t make crazy predictions, do a real analysis of the business you are in. Make realistic analysis of prospects for growth in your business.

Borrowing from present = bad project management, e.g., spending use-or-lose end-of-year money. Funding out of proportion with desired goals. Strive for correct funding level.

Manage for short and long term by managing for sustainable earnings. Manage long term by managing how short term is produced. Know where earnings come from.

Where is it really important to be fast today? Speed and efficiency of decision-making process has a real effect on sustainable earnings. Not so important with technological development, because that is so easily replicable.

Despite common beliefs, capital and talent are not scarce. One management resource that is truly scarce is time. 80% of time is spent on 20% of value. Managers let the urgent crowd out the important. How managers allocate and spend their time is a hidden determinant of their ability to grow sustainable earnings. Measure sustainable value of every item on the agenda.

Whole vs. parts. This is about centralization vs. autonomy. Should operations be centralized in a corporate center? This kind of issue comes up a lot in field of architecture.  In product lines this is characterized as centralized asset development vs. individual product development. Much interest in agile approaches to architecture.

Vertical value = relationship between corporate center and business unit

Horizontal value = value added from coordination with other business units. Traps for both. Centralization trap: tendency of people running central functions to focus more on their efficiency than value to customers. Autonomy trap: accountability is defense from any kind of central interference; equation of accountability with authority (decision rights over inputs). Abdicating responsibility for product success.

Problem with mergers is always culture. We need better understanding of culture clashes. Culture of autonomy vs. culture of centralization. Goal must be to add both horizontal and vertical value at the same time. Need both.

Diagonal assets – sense of connectedness, common aspirations, shared sense of “how we do things around here,” cultural values. Centralize and decentralize at the same time. Centralize policy, decentralize operations. This can be in form of standards. Companies are enterprises, not federations. Company keeps firm control of strategy planning, cultural values, controls, performance management.

“Balancing” and trading off can be a sign that something is wrong. Balancing perpetuates corporate cycle, bouncing back and forth between extremes. Concentrate on strengthening bonds. Resolve tensions and make is possible to have both. This is managing for value.

Saying “managers and management are bad” is an instance of pushing failure on others. Good management, not good technology, creates value for business. Innovative business changes are too difficult for competitors to replicate.

Architecture can be managed for value, but it’s not automatic. The challenge of great management must be accepted.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s