Quality as an attitude means not locating software quality in an isolated test department, but living it as a mindset throughout the entire development process. This can only be achieved through interdisciplinary collaboration, communication with different stakeholders and the conscious use of new technologies such as AI where they solve specific problems.
Key Takeaways
- After one year, the podcast broadcasts on over 20 platforms, has achieved more than 22,000 downloads and has over 1,200 subscribers.
- 15 partners, including the three national testing boards, several specialist publishers and conferences, actively contribute to the podcast’s reach and networking.
- Quality as an attitude runs through the entire software development process and is no longer limited to an isolated testing department.
- AI in testing requires two things: actively trying out available tools and at the same time honestly asking what specific problem is to be solved with them.
How a blog plan became a podcast
Some formats start with a completely different idea. The podcast Software Testing was originally intended as a blog, with interviews in text form on the topic of Future Testing: How will we test in the future?
The turning point came with a guest. Thilo Linz, founder of Imbus, didn’t have time to answer written questions and suggested a video call instead. The planned transcription effort turned into a real conversation. It was precisely this interview that provided the impetus to completely change the format.
The lesson from this is practical: a format must suit the guests, not the other way around. If you want to attract exciting interviewees, you should keep the effort on your side and choose the medium in which people like to talk anyway.
One year of podcasts in figures
After a year, there are measurable brands. The first episode went online on April 18, 2023. On the first birthday, there are 67 to 68 episodes, depending on how you count, over 22,000 downloads and more than 1,200 subscribers.
The podcast is distributed on over 20 platforms, including Apple, Spotify and Amazon. Since the turn of the year, a YouTube channel has been added, which also shows each episode as a video.
Listeners can enjoy the episodes while cooking, on the way to work or while jogging. Knowledge in small bites, integrated into everyday life, is what the format is made for.
Why partners support a podcast project
Reach and contacts play a part in determining whether a specialist podcast takes off. The Software Testing podcast therefore started with partners from the community from the very beginning, not as a solo attempt based on the principle of “I’ll give it a try”.
The three national testing boards are involved: the Austrian Testing Board, the German Testing Board and the Swiss Testing Board, all of which represent the ISTQB in German-speaking countries. In addition, there are publishers such as dpunkt and Hanser, which have published books on the subject, as well as the dpunkt section around the German Testing Magazine.
Partners bring two things to the table that a new format can hardly achieve on its own: Access to good discussion partners and distribution in existing networks. One visible result of this is the book raffles that are created with the publishers.
Conferences and community organizations are also part of this, including the QS-Tag, the Software Quality Days, the German Testing Day and the OOP, as well as groups such as the ASQF and the TAV specialist group of the German Informatics Society. Two new partners were recently added: the ASTQB international podcast network and heise developer.
On-site recordings hit the studio
Live episodes at conferences have their own character that no studio recording can provide. In the exhibitor area, interviews with speakers are produced directly on location, with all the trappings: a plate falls, someone walks past, there’s a clatter.
It is precisely this friction that creates the energy. The setting resembles a sports presenter speaking in the middle of the action, while a lot is happening all around.
Since the pandemic, the value of such live events has become clearer again. Those who work in testing and quality take away networking and inspiration from conferences, which are difficult to replace digitally.
What AI means for testing, and what it doesn’t mean
AI affects testing from two directions, which should be kept apart. The first question: How do we test AI ourselves? This is where the concept of quality shifts, because AI systems do not function deterministically and require a different perspective. The book “Basiswissen KI-Testen” by Runze and Röttger provides the basics.
The second question: How do we use AI in testing? Many tools now carry the AI label prominently, but often had similar functions before, just without the name. A lot is possible today, from large language models such as Mistral, ChatGPT and Gemini to recognition algorithms and pattern recognition.
A twofold approach helps when dealing with this. On the one hand, it’s worth trying out the tools and experimenting. On the other hand, you should take a step back and ask yourself: What problem do I actually want to solve?
Concrete weak points in testing provide the direction for this. The media discontinuity in the transition from the technical side to the technical side is one such point. Test data in heterogeneous system landscapes is another. This is where AI can provide real benefits instead of remaining a nice gimmick.
Quality starts at the beginning and runs all the way through
Today, quality is no longer the task of a compartmentalized test department. In the past, code was thrown over the fence, tested there and sent back via a ticket system. Those days are over.
Projects are interdisciplinary and everyone contributes their strengths. This creates friction that has to be absorbed somewhere. At the same time, quality has taken on a different significance: It starts at the very beginning and runs through the entire software process.
This results in an attitude requirement. Quality as a mindset is not created when a department bluntly knocks down test cases, but through exchange, discussion and integration throughout the entire process.
Quality has simply taken on a different significance because it starts at the very beginning and runs right through the entire development process. Richard Seidl
Why social skills are becoming more important in testing
Testing and quality live in a field of tension between different stakeholders. Some want quality, others want quick completion and still others want to see proof. If you want to be effective here, you have to communicate and interact.
After around 20 years of agile projects, it is clear that this topic is still growing. Soft skills and social skills are not an afterthought, but the prerequisite for interdisciplinary teams to achieve common quality in the first place.
The target group is therefore broader than the term testing suggests. Business analysts, requirements engineers, project managers, developers, test automation specialists, test analysts and test managers are all equally interested in quality across the entire software process.
How a listener community is part of the format
A specialist podcast grows stronger when listeners become a community. There is a program for this via the Steady platform, comparable to Patreon in the US, with three levels: Bug Hunter Brigade, Testcase Titans and Quality Wizards.
There is an open point behind the contribution: a podcast like this costs a lot of time and money, and a sustainable format needs a balance between what one person can achieve and external support.
In return, the community offers insights and exchange. Members see the editorial schedule for the coming weeks, give feedback, receive backstage information from conferences and meet up at listener meetings, on site and online.
Topic suggestions are expressly encouraged. Anyone who knows a speaker or is missing a topic can contact podcast@software-testing.fm directly. Episodes are planned on topics such as AI, accessibility, security, documentation, test automation and test data.
