Tuesday, September 18, 2012

Quality Insurance


How does a change in the software affect your business process?
What's the risk of having an issue in production? What's the occurrence of -  and damage on failure?


Most business fear lies in signing analysis documentation to send it off to development. They suddenly become responsible for knowing what they want and how they want it, before ever trying it out. Making mistakes is human, not being able to see the future is human as well. So why would you sign a document, knowing mistakes are most probably inside?
What else could we do, except for agile product delivery to deliver the clients explicit and implicit changing needs? And even there, how can we prevent a business going broke as a result of a software defect that was not found during development?

For example:
http://www.reuters.com/article/2012/08/02/knightcapital-loss-idUSL2E8J27QE20120802

The answer might be.. QUALITY INSURANCE (or QI)


Why the term QI?

Quality insurance is a term used to oppose the often wrongly used term 'Quality Assurance'

How would that conceptually work?


We would make a difference between soft- and hardware insurance.
By default, this insurance system would not focus on the underlying infrastructure. Neither is it explicitly excluded from the insured scope.

A software insurance policy could be taken after the Go/No-go decision was positive and would start at the end of supplier warranty.
For starters, the business processes in scope of the insurance policy would be described by a team of experienced analysts and testers.
An intake is done on the delivered product in a technical go live in a production-like environment.
This intake would consist of an exploratory test.

Possible product issues and risks are reported. Per identified risk, an insurance fee is mapped to an insured amount.
The interesting risks to insure are the ones with a very low probability of occurrence, but having a very high impact on failure.

The company can then decide to reduce risks (and insurance costs) by fixing defects if the fix cost is lower than the fee or decide to insure themselves for this risk.

   
What would be the Strengths?
  • Increase the business attention on product risks. Issues can be resolved before go-live.
  • The business can go live with a product they are not sure about resulting an overall decreased project failure.
  • Direct linking of risks and issues with costs. This confronting exercise increases awareness.

 What would be the Weaknesses?

  • Resolving issues late in the development cycle is still more expensive. The direct cost of software delivery is not reduced but rather increased due to the additional insurance policy.
What would be the Opportunities?
  • Level out the current overrated importance of certification  - a tester will need to prove what they're worth. Also development quality will stand more in the picture. 
  • The real value of a good tester becomes visible. The tester's function would get more appreciation and direct value. 
  • The value of the Exploratory Testing method can be proven. Only by exploring, focusing and de-focusing, the dangerous underlying issues, inconsistencies and process gaps can be found... and insured.
  • Increase software quality by increasing the awareness of the cost on failure.
What would be the Threats?
  • Decrease of software quality and user satisfaction due to competitive insurance policies. 
  • Decrease of insurance quality due to competitive insurance policies

Tuesday, September 4, 2012

10 Signs that you're not cut to be a tester

I had quite some fun reading trough signs indicating why you would be not cut for IT, not cut to be a developer or an IT consultant on the Tech republic blog:
http://www.techrepublic.com/blog/10things/10-signs-that-you-arent-cut-out-for-it/3072

I don't mean to say that it made a whole bunch of sense to me, but I did notice that one particular blog post was still outstanding... so I decided to suggest this one myself:

These are my top 10 signs that you're not cut to be a tester:

1. You get nervous from buggy software

The software you're facing on a day to day base is going to be full of bugs and will look like a poor Picasso imitation. You'll need to find workarounds. And when a bug has been resolved, you'll be facing the next problem. You'll have to go trough again.

2. You're unaware of business expectations

If you're not aware of business expectation, you'll look over half of the defects that are in front of you. You'll read over inconsistencies and you won't question the rationale behind an architectural solution that has been proposed. An overview of what business wants and why they need an application, gives you structure and focus on finding the defects that potentially are of high impact.

3. You get tired from explaining defects occurrences

Not everyone knows the system like you do. Not everyone understands the reason why a bug is a bug. You'll need to explain it... again and again. Even when it works on their machine...

4. You don't read blogs, books or attend conferences about software testing

A tester keeps up to date on the latest evolutions on tools, techniques, methods and can learn from practical experience of other testers. It's no use to invent what already has been invented for you. It is  even less useful not to learn from others mistakes.

5. You're ashamed of your role in the project

A tester could also be an analyst, a developer, a designer or an architect, but chooses to be a tester. A tester is not a developer or analyst, coming close to retirement. A tester is proud of their job and aims on improving software quality. They aim on challenging development, analysis and architecture on their correctness, completeness and coherence. A tester is proud of and dedicated to find the human flaws in a program before it is released to production.


6. You know how to check but you don't know how to explore

Checking against expectations is a given. This needs to be done. After this, the software needs exploration. How else are we going to find those undefined features and prove that if you click on the left mouse button 6 times, then do a page refresh, while keeping your right mouse button clicked and then releasing the right mouse button after which you disconnect the keyboard, your money has gone from your account, but has not been transfered?...


7. You are not keeping up to date on the technical aspects of IT

IT terms are not just hype words for a tester. A tester has to have an understanding of OO, has to know what a web service is or a message queue, what the function is of an isolation layer or why server and client side validations can be implemented while building a website. A tester needs to constantly improve and understand what they are technically dealing with as every technicality has it's weakness. That's where most probably the defects will be.

8. You don't like to communicate

Communication, communication, communication. Testing is about finding bugs and inconsistencies and communicating about them.

9. You are not aware of the application development life cycle

A tester needs to know about architecture, design, analysis, development, infrastructure, release management, ... and also how they fit together. A testers could even carefully address flaws in the application life cycle, since they might result in software bugs later on.

10. You can't get over defects you found and don't get fixed

Face it. You'll find those defects that are obvious and you would never want to see in your software. However, you're not paying for every single defect to be fixed. Decisions will be made that might not make you happy. Your job is to point to the risks and issues of not fixing a defect. Your job is not to get them resolved.

Saturday, September 1, 2012

Becoming an Inspiring Test Coordinator

Many starting testers would like to become test coordinators in stead of better testers. However, many good test coordinators are good all-around testers with extra coordination skills. I don't know for myself, how good of a tester I would make but I know I always liked testing and I keep improving my testing skills, even though officially, I'm not testing anymore for the last 3 years.

Seeing Test Coordinators, and working with and for them, I realized there are only few inspiring leaders among them. The biggest pitfalls a test coordinator has in front of them, that keep them from becoming inspiring leaders are the following:

"The common enemy" or "The common goal"?

As a coordinator, you'll need to run test activities like clockwork. The best way to do so is by making your individual testers work as a team.  Bringing people together can be done the easy way, by finding the common enemy, or the hard way, by identifying the common goal. The common goal sounds a bit naive maybe. Facing a goal that you might not reach also requires some sort of bravery. Defining achievable targets towards the common goal, is the key to move on and increase the team spirit on success and even on failure.


"Driving the heard" or "Leading by example"?

Telling people what to do, often results in the opposite effect. The best way to get respect,  is to take part in the test activities. So you better aim to be a good tester, if you ever want to become a good coordinator. Coaching is part of the job. If a tester doesn't know how to get started, you will have to come up with something that works. If your testers miss out on something, it's your job to find out about it and make them aware. 

"Responsibilization" or "Taking the problems away"?

Your testers will face issues constantly. Unstable environments, data restore issues, changing user expectations, defect discussions... Issues will overcome them on a daily base. You have 2 options. Try to let your testers solve the issues by making them responsible or taking their problems away. Taking problems away is often a hard job and might not be very rewarding. You'll need to make your team aware of how you manage their problems, you can even ask them for advice.


"Push pressure further" or "Push pressure back"?

Project managers like to hear that all is going well and tend to become a pain in the ass when testing is not proceeding as planned. Mostly that is because all the buffers they placed in the project are used and your poor test activities are getting squeezed in timelines. Your job is to make sure that testing continues like clockwork and keep management changing ideas, controls and unnecessary reports as far away from your testers as possible. 

"Mushroom coordination" or "Transaparant coordination"?


Knowledge is power. As a test coordinator, you know more than your team members. You take part in steering and coordination meetings. You can use this knowledge either to outsmart your team members and stay one step ahead of them, or you can use it to inform your team members for which they will reward you with trust.