Facilitating the Spread of Knowledge and Innovation in Professional Software Development

Write for InfoQ


Choose your language

InfoQ Homepage Articles Intelligent Evolution: Making Change Work

Intelligent Evolution: Making Change Work


I have been working for nearly twenty years in changing organizations. Over this time, I have come to the conclusion that some 80% of all improvement and change programmes fail. Failure means that they did not achieve the expected results, the investment in the change programme was greater than the value achieved, “improvements” were seen as mostly bureaucratic, or changes were abandoned soon after the implementation.

In this short paper, I would like to make some recommendations as to how things can be done so as to ensure long-term business success rather than short-term satisfaction of a standard or theory.


Over the past two centuries, a new life-form has evolved: the multi-national organization. This is a new creature, whose prime purpose is to survive and grow, assuring the perennity of its genes is more important than the product or service its creators had in mind.

When these organizations try to change, they frequently fail because we forget that they are complete creatures and need to be changed in depth. Trying to improve the practices of the engineering department without paying attention to the rest of the organization is similar to trying to change an organ within a body without understanding the impact on the rest of the body.

Imagine that a surgeon performs a liver transplant without considering whether there is a potential difference between the livers of different people. We would have a high probability of organ rejection and potentially death of the patient.

The same principle is true for the change of work practices: if you are trying to change the way things are being done without understanding the culture of the organization, the inter-relationship between the engineering departments, sales, management, policies, relationships with other departments, suppliers and customers… The result must lead to failure.

If you want to change the way things are being done, you need to understand how to perform this in an intelligent evolutionary manner rather than through a revolution.

What is Evolution?

Natural evolution is generated through a large number of random mutations of the genetic make-up of living beings. These mutations happened in every creature throughout history. Some of these have little or no effect and disappear without being noticed in the next generations. Others are destructive and appear as genetic diseases. Finally, a few make the creature more resilient, more attractive, stronger or give it some advantage over the others; these are then prized and rapidly become common as the creature involved gets to reproduce more than the others.

Evolution is always linked to transformation, to movement, to a process of gradual change – and that is what we want when we try to change the way we work.

What is Intelligent?

Intelligence is defined largely through knowledge and understanding, based on an effective exchange of information – efficient communication, whether internally (my thought process, the speed at which I can reach a decision) or externally (how well I am able to listen to you and respond in a meaningful manner).

Intelligence is always linked to the ability to obtain information and use it efficiently – and that is what we want to achieve.

Why Do You Want to Change?

The success of evolution can only be measured if it is generating an improvement rather than a regression or “genetic illness”. In the case of intelligent evolution, that objective needs to be identified early on – otherwise the evolution, like natural evolution, is random and it is only after the fact that we will discover what should have happened.

Evolution can only be successful if it is generating something which is valuable and beneficial. Intelligence and reason are needed to support it. For intelligent evolution to work, it needs a clear motivation and a sense of urgency to encourage it and to judge, step by step what the objective is.

One of my most successful recent customers was a Romanian software development firm, developing software for the Western European market, a subsidiary of a Dutch company. This was a successful company: they were busy, their customers were reasonably satisfied, they were making a profit, their Dutch management and sales team was bringing business in, their staff was stable… When I first spoke to them, I asked them why they wanted to change anything, what was their motivation for implementing an expensive and disruptive improvement programme. The answer was short and clear: they told me they were starting to lose business to the Ukrainians who were cheaper.

That was motivation.

They followed my suggestions and implemented the principles of intelligent evolution (even though I had not yet thought up the name). The result was that after a couple of years, they saw a 46% increase in their bottom line (during a period of recession) and the highest rating customer satisfaction survey in the history of the company.

In Practice

This organization decided to use the models, standards and theories as guidelines to help them consider what they could or should do. They combined the principles and suggestions of CMMI (Development and Services), ITIL, Agile, and more and weighted them according to the business benefits they could get out of it. Some of the things they implemented are listed hereunder.



A strong effort of communication was put into place so that everyone was trained and educated in basic practices. This included risk management, measurement, peer reviews and more. This allowed them to start speaking a common language, using terms such as “quality assurance” “validation”, “verification” and “configuration management” to mean very specific activities rather than the confusion over these terms which they had previously – as do many other organizations1.

The communication continued by having results from all the projects continuously displayed in the halls on flat-screen televisions. This included measurement such as the progress on the project, the number of defects, the cost of quality and others, as well as general information regarding the company, meetings, courses, visitors, new staff, etc. This encouraged projects as they could see their status and compare to their colleagues.


All activities were measured based on the time spent on the activity, as precisely as was reasonably possible. Every activity in the WBS was related to a cost of quality code, identifying whether this was a good quality cost (training, planning, risk management…) or a bad quality cost (rework, customer support…). These data were analysed so as to identify the root causes of the costs.

The measures are collected automatically and made available in real time. This means that there is no temptation to cheat, to pretend that things are better than they are – because the reality will come out in any case.

By automatically collecting these data, they had the opportunity to truly understand the impact of a review on the testing effort, the value of effective quality assurance on customer satisfaction – in fact, this last measurement was highlighted, when, having decided that they were producing good quality, they could reduce the QA effort – the impact was immediately visible and they returned to the previous level.

Quality Assurance

An effective, proactive quality assurance is implemented on every project. Their purpose is not control whether projects are following the standards, but whether they are doing what is most efficient and effective to increase customer satisfaction. One of the most difficult battles within the company was when I tried to persuade them that they needed to have QA people at senior management level and not only at the project level. Senior management obviously was not interested in having their own board meetings attended by a representative of the engineering department, reviewing and commenting on their processes. It is only when things started going wrong and some decision were made a the senior management level which created issues throughout the company, that they realized that QA and a formalized decision and analysis process were required at the top level.

The value of QA was measured (see Measurement) and demonstrated and management soon realized that they could also benefit from an independent, objective insight into their methods. It always surprises me when senior management believes that they do not need these elementary practices…


Some time ago, I developed a presentation called “Forget the Process, Focus on the People” or FP2 for short. The idea was fairly simple: people produce quality, customers want quality. If you cannot demonstrate that your tools, processes, principles, ideas, or whatever can make your people happier, more productive and/or more creative, than you should not be using them. This led to an in-depth analysis of all new practices based on their value rather than their cost.

One of the approaches they used for this was through a regular process improvement team meeting at which the people responsible for any particular task or activity came together to write a one-page set of rules and principles for it. These were translated into posters and distributed widely throughout the company. Every poster had the e-mail address to suggest changes and improvements to practices. Every e-mail was read and answered within a couple of day.


Customer feed-back was used to underscore the need for improvement. Customers were invited to the site, saw the measurements and the internal communications, could see the progress, the results, and the risks on the projects which they were financing. The fact that a customer was going to see the results of your project encourages you to make sure that those results will bring you praise, not blame – and the good results need to be in place at all times, before the customer walks in the door.

Many organizations I see do not feel the motivation or the urgency. They see their revenue and feel comfortable, they do not consider that the amount of time and money they are wasting is significant, they continue to cut costs in the wrong places (actually, they frequently cut investments like training, communication and quality, while spending more money on costs like corrective actions). The time to improve is now!

Companies that are doing well do not consider the potential additional savings and income that a more efficient work process could deliver. The lack of urgency means that nothing will get done for the foreseeable future. Companies that are doing badly feel they cannot afford the investment of a change programme, so they try other means to change things. This typically involves imposing new practices, rules and tools which force the staff to forget what they were doing successfully in order to implement what amounts to little more than bureaucracy. The resulting excess of process, standards and documentation ends up just drowning the staff who can no longer do their work.

We live in a global market. Whatever you are doing, you need to understand that your customers can find the same products and services cheaper elsewhere. The only thing you have to sell is your quality. The quality or your products and services is the only thing which differentiates you from your competition. Your quality is your trademark. If you are delivering high quality products or services (which produce high customer satisfaction, based on your strategy – quality might mean very cheap or zero-defect or cutting-edge technology,… it needs to be defined), your customers will remain loyal to you, will recommend you to potential customers, will not go and look for a cheaper alternative. However, if you do not continuously improve your quality, remember that your competitors are improving theirs and will be happy to take over your customers.

Focus on Quality

Quality has an impact not only on customer satisfaction, but on everything else which is important to the organization:

  • Quality means less time testing and correcting defects
  • Less time testing and correcting reduces the effort to develop
  • The reduction in the effort reduces the time to market
  • The reduction in correction reduces the cost to develop
  • The increased quality means improved customer satisfaction
  • Increased customer satisfaction means increased trust
  • Trust means increased customer loyalty
  • Customer loyalty means you keep your job.

The difficulty most companies have is understanding what they mean by quality and how it can be measured (other than by counting the number of customer complaints).

Why Revolutions Fail

Obviously, the easiest way to improve the way things are done would be to document the ideal approach (or buy it from a company promising you miracles). So you bring in the experts, the consultants, the gurus, the wise (wo)men, who create documents and charts and graphs and roll out procedures and tools and processes and methodologies and rules and metrics and a whole lot of stuff you don’t understand. They then tell everyone to implement this new approach and meet significant resistance. The people who have been creating your products and delivering your services for years are not interested in changing what was a reasonably efficient approach to follow these theories. The consultants now implement the “Quality Assurance” police, who will come, armed with checklists, and make sure you are doing what they told you to do and report any deviation to management. Their role is to make sure you follow the rules. Contrary to their name, they are not seeking to assure the quality of your products and practices, they are only controlling conformity.

Quality is now understood to mean conformity and bureaucracy.

Many times, I have heard people tell me that quality is respecting the standards. When asked to justify this statement, I am told that if they follow the standard they will produce better quality: they see me as the CMMI auditor and they need to tell me that CMMI will solve the problems. Unfortunately, I cannot accept that statement.

“How will following the standards increase customer satisfaction?” I ask. The answer is a blank stare.

“What is the commonly held definition of quality here?” I ask. They don’t really know, they have no answer.

“How can you assure quality of you don’t know what it is?” I ask. They now appear to be very uncomfortable and wishing they were somewhere else.

“How do you measure quality?” I ask. They tell me they measure customer complaints.

“You mean you only measure the quality of your work after the customer has received it and complained? How do you measure the quality of your work during the development phases?” I ask. They tell me they measure the number of non-compliances identified.

“What is a non-compliance?” I ask. They tell me a deviation from the theoretical standard.

And we are back where we started.

After the French revolution in 1789-90, there was a period known as “La Terreur”, in which pointless rules were established and people were killed because someone denounced them. This is fairly typical after a revolution. It is the principle which is put in place by Quality Assurance, management’s process police force.

The result is demotivation and the departure of key members of staff – and it is always the best people who will find a job and leave first.

Intelligent Evolution

The Principles

We need to look at achieving results when trying to change an organization: we need to take our time, act step by step, with a strong emphasis on understanding and communication.

The first step in implementing change is understanding what and why we want to change, what the objective is, where we are trying to go.

I am always a little baffled when customers tell me that they want to be CMMI level 3 or that they want to implement Agile methodologies, because, then they will be more efficient or productive. This is a common tendency of confusing the result with the tool: it is not because I have an expensive paintbrush that I can paint. One of my customers, an international insurance company told me that they had exact data on how many minutes their computer system had been down during the year. They also knew how many independent insurance brokers connected to their system at any time. If the brokers could not connect, they did not wait (they were usually dealing with a customer when they tried to connnect), but went directly to another site, that of a competitor. The CIO could tell me exactly how much computer and software failure had cost the company. He had a very clear idea of what he wanted to change / improve and how much he could invest in this.

That needs to come from management in the form of clear communication as to the path to follow and how progress along that path is to be measured, throughout the life of the change programme, the projects and the processes. What is the purpose, why are we being asked to change things, what are management’s expectations? Defining these expectations is not a simple task, it involves documenting what management expects to get out of the practices being implemented. In most cases, I see communication of what people are expected to do, rarely why.

The second step is to underline the desired outcomes with clear measurements which will encourage the right approaches. This is best done by simple measurements which can be clearly understood, and preferably over time:

  • Reduce the number of requirements changes due to misunderstanding;
  • Reduce the number of defects which are allowed to pass from one life-cycle phase to the next;
  • Reduce the number of hours over-time or rework;
  • Reduce the deviation between estimates and actuals.

A clear quantified goal is always better, but I tend to encourage a proportional goal (reduce by 10% per year, every year) rather than an absolute goal (no more than 5 defects per KLOC).

Most management teams know what the right answer should be, so when questioned, they always tell me that they are focused on quality, they advertise quality, they have posters about quality in the work places, they talk to the staff about quality… When asked what they measure, I am usually told that they measure time taken and cost of projects or activities. Just for clarification, I ask them to confirm whether quality therefore means cheap and fast.

If you don’t measure it, you won’t achieve it.

The third step is to implement a team which aims to assure the quality of the proposed, planned and performed practices – not control the compliance, but assure the quality. The team does not have to be a heterogeneous team, but needs to be a team of people understanding the different jobs concerned. Typically in the team, I would expect to see an experienced project leader or scrum-master, a software developer or tester, someone who has experience managing people, etc. But the team needs to have a common reporting structure and regular communication and exchanges. This means that they understand the pressures, the customers’ needs, the skills and abilities of the team members, the business requirements, management’s expectations, the standards and processes being used, the successes and failures of other teams and able to counsel, support and guide – they should also be able to challenge unjustified requests. They can answer questions, make recommendations, offer coaching and training and they can understand why a practice failed so as to bring back the lessons learned and improve the standards. Quality assurance will regularly remind staff that the objective is not the task, but it is the end deliverable and the satisfied customer.

The Practice

When trying to change the way people work, we need to consider how to get them to change the way they think rather than just telling them to do things in a particular way. This means, we need to be able to reach their complete mental process: reasonable and emotional, reflective and active – and we need to understand the intersections and interactions of these elements.

Reasonable + Reflective = Acceptance

The combination of rationality and intellect is probably the area we are most willing to advertise and believe to be our own dominant trait. It focuses on a logical interpretation of the known facts. This is the part of us that understands and accepts the need for change. The demonstration of the value of a change, the statistics based on similar organizations and the evident return on investment should be enough to demonstrate the need to change, just as the fact that I am breathless after coming upstairs and the view of my stomach in the mirror makes me accept that I need to exercise more.

And yet, we don’t do it.

Reasonable + Active = Ability

This combination covers the ability to do what is needed. It includes knowledge, skills, understanding and access to the necessary equipment and practices. We can show you what you should be doing, explain how to perform it. We can educate and train you to achieve the results, provide everything you need to be successful. I have a gym near my home, I have a bicycle and I have the time to exercise.

And yet, we don’t do it.

Emotional + Reflective = Aspiration

Viktor Frankl explains, in his theories about logotherapy, how people are willing to put up with amazing levels of suffering, if only they understand the purpose or the objectives. For effective change to happen, there is a need for the people involved to have a shared vision of what the end will look like, what the image of completion for which we are aiming is. This must be inspiring and encouraging, and it must be shared by the people involved: when asking about the vision, we should hear that “our vision is…” rather than “management’s vision” or “they”. This is something we all believe and want to achieve. Personally, I would like to be able to run a marathon.

And yet, we don’t do it.

Emotional + Active = Attitude

The reason so many change efforts fail is because the last quadrant is not being considered: people’s attitude. This is deeply ingrained and is what provokes your reaction more than anything else. This is also the cultural aspect of the organization, but also of the teams and the individuals. If we want to change anything, we need to change the attitude. I was recently in an organization which advertises its sustainability and quality but focuses internally only on delivering projects on time. The amount of bureaucracy they had put in place to manage the projects, meant that the project managers did not really have the necessary time to do the work. When I suggested a few simple changes, measurements which could be put in place in order to reduce costs and increase efficiency, the reaction from management was largely a statement that “they will never accept that” and “it is too difficult”. This is the culture, the attitude which leads to failure. Interestingly, the engineering staff wanted to do the things which management decided they would not accept, they were stopped by the negative attitude.

By reducing the requirement on generating reports and teaching managers to perform data-base enquiries, we freed up a significant amount of time spent on paperwork and managers found they had access to the real data, when they needed it rather than when the project manager reported it at the end of the month. It is estimated that we saved an average of half a day’s work per person per month, and increased the reliability of the management data.

Moving On

Changing the attitude or the culture is something very difficult and delicate. It requires careful management and continuous encouragement. Some techniques need to be implemented at different levels in order to ensure sustainable improvement.

Shared Vision

Management needs to establish and communicate a vision of what the organization is going to be in the future. That vision needs to be communicated in such a way that the team members buy-in and commit to the vision; they need to consider it as “our shared vision”, rather than management’s dream. That vision must be sufficiently challenging to motivate those who want to make a difference.

Visions that have shown success are usually focused on people (customer and employees) rather than on shareholders and finance. Some of my customers have used concepts of increased measured market-value of the employees (implemented through training and support), or becoming the preferred employer within a region (meaning a strong increase on employee satisfaction and interesting projects).

Established Expectations

Expectations, in business and result terms need to be clearly communicated across the organization. People need to know and understand what they are being asked to deliver and why this is critical to the realization of the shared vision. The expectation is not that more defects are found in reviews and tests, but that less defects are missed; the expectation is not documentation, it is traceable communication; the expectation is not to perform tasks, it is to deliver results.


Most organizations I visit focus on training people to use tools, to perform tasks, however there is a lack of education – helping people to understand the purpose of the activities they are being asked to perform and guiding them in thinking for themselves, judging and improving continuously at every level. Nearly all the companies in which I participate show me training in their tools, their processes and techniques and (sometime) their customers’ business. Few of them, unfortunately, appear to encourage staff members to think for themselves, based on their understanding of the company and the customers. They should be encouraged, not to follow their standards, but to challenge their standards and continuously prove that there is a better approach. I remember a CEO of a relatively small firm (about 250 people), who called a full company meeting on a monthly basis, during which he gave the usual management progress reports, but also mentioned by name the people who had communicated the best efficiency improvement ideas, and responded to any and all rumours in the company. People who had challenged him particularly successfully were frequently rewarded publicly with a shopping voucher, or some other token of appreciation.

Constructive Analysis

Measurement is the most powerful tool at the disposal of any management team. However the measurements are not useful without appropriate analysis. The focus needs to be on understanding which measurements identify problems in the development or delivery system. The purpose is not to identify the person who is under-performing, but the roadblocks in the whole system: where are you wasting time and money? What is frustrating the people doing the work? Where are the earliest points at which errors can be identified and removed efficiently?

This is the foundation of methodologies such as Lean and Agile. A simple method is the one mentioned earlier on about measuring the cost of quality of various activities and the time they take to be completed. This can be automated through an intelligent time reporting system and analysed accordingly.

One aspect of the analysis (see illustration) is understanding how good we are at estimating. If you are good at estimating, no matter which technique or tool you use, you should be overestimating and underestimating about as much. Ideally, I would like to see an “expert” defined as someone who overestimates 45% of the time by less than 10% and underestimates 45% of the time by less than 10%. That leaves 10% of her estimates which are affected by random circumstances. Also, I would expect to see that average improving over time, continuously. If you are always meeting your deadlines, I think you are probably overestimating…

Risk Management

An often forgotten management technique, which consists in identifying how much effort, time and money should be consecrated to a problem which might never happen. Everyone within an organization occupies a unique position and can therefore see things that are not visible to anyone else. They are able to identify risks which cannot be identified by their management. People who identify risks and report them are demonstrating an interest in the success of the organization, not criticizing management.

A few years ago, I was did a risk identification and analysis workshop in a Belgian company. The first reaction was resistance: the project manager assured me that he and his architect had already identified all the risks on the project (about ten of them). I insisted that I wanted the full project team participating. Two hours into the exercise, we had identified some 250 specific risks which had been originally missed. We estimated these risks, by giving them real numbers2. The top risk had a probability of close to 70% with an impact which was estimated at over twice the total budget of the project, considering loss of market, reputation, etc. This is not difficult, but it does require active participation and understanding.

Reward Mechanism

A reward mechanism should support the people who are trying to move the organization forward. There is an old maxim that the person who manages a project well is seen as “lucky to have an easy project which ran smoothly”, while the one who messes up and struggles to deliver anything gets rewarded for their efforts. Understanding who should be used as an example and rewarded is one of the more difficult management tasks. Rewards should not be financial, they rarely have any lasting impact. Most efficient people are happy to be rewarded instead by being given time to attend training, or an all-expenses paid trip to a professional conference – this also means that their reward will keep on giving as they bring back their knowledge to the company and share it with their colleagues.

Frequent, small rewards are always better than rare large rewards. Salary raises rarely motivate for long, as people realize after two months that they are not living more comfortably than before.

A couple of companies for which I have worked (very well-known world-leaders in their industry) have taken out patents in the name of their employees when an idea or improvement justified it. The employee now got to receive an ongoing commission every time the idea was used. But people who identify problems and risks should also be rewarded. If you fail at something, there is a lesson to be learned by the rest of the organization. People need to be encouraged to share their lessons learnt and their mistakes.


Change is possible and needs to be encouraged. It is not done by throwing out babies and bath-water, it is done by building on the skills and experience which already exist in the organization. You need to trust the people who are doing the work, they got you were you are, even if they did not have the necessary training, skills, tools, processes, techniques… necessary for continued improvement. Listen to them at least as much as you listen to any consultant or “expert”. They need to be based on progressive transformation and clear communication at every level.

Change management requires:

  1. Establishing a challenging objective
  2. Measuring progress towards the objective
  3. Encouraging staff to find better ways to achieve the objective
  4. Sharing lessons learnt
  5. It’s not easy, but if it was easy, it wouldn’t be fun.

About the Author

Peter Leeson is a process and quality improvement consultant, a CMMI instructor, coach and lead appraiser, an SEI partner and a regular speaker at conferences the world over. He is also the director of Q:PIT Ltd, a network of professionals focused on raising the standard of quality. He can be reached at


1 Quality assurance is often confused with quality control, testing and others. Configuration management is frequently used to reference version control. Validation is seen as customer acceptance or testing, while verification is reserved for reviews.

2 I prefer a risk to be identified as “25% chance this will cost €150,000.00” to a risk identified as a “high-probabilty type A” risk. One of them will get management’s attention.

Rate this Article