For this Quality Engineering interview, I welcome Christian Sayegh, specialist in automation consulting and strategy, co-founder of CloudNetCare.
With 35 years of experience in IT, with quality always on his radar, Christian has plenty of anecdotes to share.
We were able to discuss his current perspective on quality in organizations through the various missions he was able to carry out.
A good listening, reading, or viewing at your convenience, do not hesitate to comment.
Join the QE Unit to access regular exclusive content via the newsletter.
About Christian Sayegh
More than 20 years of experience exclusively in the world of service. Holder of a Paris VI degree in IT and a Masters in management, he has devoted half of his career to working for prestigious clients such as: Total, CCF, BNP Paribas, EDF, Pfizer Laboratories.
Beginning as a developer to end up as a project manager. The second half of his career will be dedicated to customer relations with the objective of responding as closely as possible to requests within companies such as: BUT, HSBC, Laboratoire Aventis, ATAC, BNP Paribas, Fondation Claude Pompidou, Société Générale, Banque Rothschild , MFA, Robeco bank, Cortal bank consors, etc.
Healthy relationships with integrity have enabled him to maintain relationships within these companies.
Antoine: Can you start by introducing yourself?
I could describe myself as an “old computer scientist”, I started as a developer.
The IT sector has always interested me, I went from developer to project manager, project director etc.
In this first one part of my career I have been confronted with the quality of software which has always interested me, to provide useful and stable software.
Moreover, I would like to point out that traditional software has always lagged behind this quality wheel, in comparison to the industrial sector.
In the second part of my career, I moved to software edition with CloudNetCare, to provide a platform in the field of testing,
At the beginning of my career, I carried out many projects where the tooling was a real difficulty: very complex, expensive, and time consuming to implement.
Antoine: Because of your experience, do you feel sometimes that “the history repeats itself”?
You said it!
My view is that there has been very little change in mentalities in the IT sector, especially for development teams to take quality into account.
We are still at 70% of IT projects that fail, having not yet integrated quality to its proper extent.
Quality is often an activity that is attempted at the end of the project, as if it had been forgotten, in a vain effort to improve the product at the last minute.
Antoine: I imagine that a project that did not take into account testability at the start, is like adding to change the plans of a house once the foundations have been laid?
Yes, moreover there has been a lot of literature and models on BDD and TDD.
Despite this, I have not seen any company having succeeded in integrating these practices upstream and at scale.
I think there are strong gaps between what is imagined in the models, and the reality on the ground.
For example, the examples have assumptions about a team of fifteen testers, trained and an investment in quality that is rare in reality.
Small teams with limited investment in quality represent the majority of organizations today.
The pitfall and not so good idea I see regularly, is to rely only on so-called “free solutions”.
As the organizations do not want to overload the facial cost, they try to integrate those solutions searching to “save” money in the first step.
We hope that the quality of the product will suddenly improve, without any investment.
Christian Sayegh
They take an part-time internship, ask a developer to do some unit tests, and we hope that the quality of the product will suddenly improve.
Even today I still frequently encounter this approach in teams who want to take a step forward in their quality approach.
Most of these teams are in fact unaware of the effort and investment required to initiate such a process.
It is also through these points that I feel the lack of evolution of maturity, quality is not considered as an approach as such.
Whether it’s the Merise model with 20% load allocated to tests, or the 30 to 40% for agile methods
Antoine: In line with you, I have the feeling that we have trouble communicating the value of quality, and that this contributes to reducing the perceived value of quality.
Indeed, this is a key point for me.
For the decision-makers, quality must increase turnover, profit, for more and better.
As the first function of software quality is indirect, geared towards risk reduction, it is very difficult if not impossible to factually quantify its value.
Take the example of a payment functionality, how much do we lose if its operation is degraded?
It is very difficult to answer and measure, even comparing history or performing gradual activation of functionalities.
The problem is in fact posed in the wrong order, where one asks to justify an ROI for the software quality, it is a biased equation from the start.
The value of quality cannot be justified with a traditional ROI approach.
Christian Sayegh
The second value of protecting the image of the company has the same difficulty. How to measure the negative impact on the image of the organization?
To be perceived so much at a strategic level or in a project, the quality should not pose the justification of the quality under this angle, it is for me doomed to failure.
Quality is also defined as a business support function, which positions it in checking specifications, not in adding value to the company.
It also helps to give a “simplistic” picture of quality assurance, as if it was only a question of automating existing verifications.
In addition, the teams outside of QA denigrating this function push to discredit the field.
I have an example, moreover, where a company consulted us, very motivated to initiate its quality and automation process.
Our first question was to identify who would be our operational contact for the project.
There was none and we preferred not to accept the project, at least at this level of maturity.
An organization without a clear contact for its quality initiative is probably not a real priority.
Christian Sayegh
What ends up happening is otherwise a very likely delay in the process for other internal priorities, in addition to the lack of allocated resources.
We realize that unfortunately, as quality is the last wheel of the coach, it is always what is shifted, sacrificed.
It reminds me of an anecdote when I presented fixed price projects, with a workload allocated to product documentation.
When I was told that 30 days “is a lot for documentation”, we ended up doing best effort with the remaining possible workload.
This is exactly what I see happening to quality in projects.
Antoine: In this complexity, do you have any practices or actions that you have found effective over the years?
I can evoke cases of companies which have succeeded in their approach.
First, the presence of an internal sponsor in the company, who will boost quality as a strategic axis for the organization is necessary.
It is thanks to it that a deployment could for example be postponed for lack of sufficient quality.
These are strong messages that are needed to reverse the historical trend of the organization.
He can then complement the changes in culture by delegating part of the process to influential people in the projects.
Antoine: Is it as if a quality approach was a real subject of organizational culture?
Yes exactly, and the second necessary building block is real resources with regard to the objectives set.
We see several scenarios, companies motivated but without means, others with the means but no internal skills to use them.
To be successful there really has to be a symbiosis, I have seen few organizations go up to a maturity level in CI/CD and automation.
I am rather morose about the evolution of mentalities and culture on the importance of software quality.
This importance becomes obvious only when companies have taken the wall.
Antoine: As if the organization no longer had other solutions?
In part, but it is not from this angle that it happens.
At first the organization is faced with a crisis, hits the wall, and is subject to heavy losses for the company.
It is only after this point that the company begins to consider quality as a structural solution.
Organizations actually need to see what’s going on without integrating quality.
When we are called upon, we are not directly shared with the context that led to the process.
But subsequently, by sharing and exchanging, we often obtain elements relating to a crisis that the company has had to go through.
Organizations have to experience what happens without quality to react.
Christian Sayegh
Even at this level, it remains necessary to validate the real allocation of a budget.
The response “A budget? Yes, we’ll see that in a second step ”is for me an alert which can lead me not to carry out the project.
This ties in with the lack of consideration for software quality.
The CEO or CFO wants to know how much the process will save, not what it avoids losing or hypothetically allows to gain.
Hence unfortunately the need to have a crisis to change mentalities.
Antoine: Have you over the years, in search of quality, content that has helped you?
I don’t have a specific content in mind, there is a lot of existing content through conferences, ISTQB, meetup groups.
Personally I learn a lot from experience, by being attentive in the execution and by doing retrospectives.
The objective is to improve my practices with each project by taking a step back.
This quote from Oscar Wilde is an inspiration to me.
Experience is the name man gives to his mistakes.
Oscar Wilde
Experience is often gained with mistakes, things that don’t work as expected, that’s what keeps you going.
Today, luckily with the structure that we created, we have a sharing through the various projects that we lead.
In addition, exchanging with other people and peers facing similar problems is also a real benefit.
As in the last round table in which we participated, we realize that this is a problem for everyone, without having a magic solution to the subject.
Be curious, interested, go to forums, events, there is no lack of content.
I have nothing against books and theory, it allows sharing and exchanging, practice remains for me the key to progress!
The key points of our discussion
- The mentality of the teams must evolve to integrate quality earlier
- Models are useful, but the reality on the ground remains different
- Quality is struggling to be valued and recognized as a real process
- Lack of human resources and financial is a recurring problem
- A sponsor is a necessary pillar to initiate a quality approach
- The need for organizations to experience a crisis to react
- The importance of questioning and listening to understand the context