Is software testing an āindustryā? I thought it was an intellectual process profession.
I see:
a trend of increasingly less smart approaches to testing.
an increased demand for juniors due to pay constraints, or hiring seniors on low pay.
more variation in tools and programming languages required for, and used by testers; together with an increase in demand of expertise in programming
AI usage ideas in testing come and go.
after an upward trend of common ground and few large communities, there seems to be more disconnect, split communities, groups, organizations, or people not being part of any;
Has it already been 3-5 years? What has indeed happened - great question!
A more fragmented industry network in two regards: After the implosion of Twitter, people are online at more different places: smaller community slack sites, mastodon, and BlueSky. Also, there is a larger range of cheap online conferences - and a handful of smaller testing conferences.
Congrats again to the team on the turnout for TestBash UK - there is power in meeting people live.
There is a better diversity of speakers and a whole new set of keynote-givers⦠in most of the places. A broader set of topics too. Like a TestBash almost always has something about mental health. Keep experimenting with formats that bring people together when we are together.
For me, the testing work is more about operational testing and compliance requirements than functional test cases and scalability. New Fields of Testing Activities | Complexity is a Matter of Perspective. I have a hunch that LLMs will hit all the document writing and test case stepping someday soon.
To me, itās also more about the testing activity than the testers (formal roles). The Testing, not the Testers | Complexity is a Matter of Perspective. Plenty of different people are involved in doing testing in some capacity. Perhaps we can expand the scope of the MoT to be more about the testing activity, so we are not as siloed?
A last thing could be that people are stuck in previous ways of working, for various reasons. We have great up-to-date models for better deliveries but teams seem to lag enough of free processing power to notice that things can be different. Make MoT the place to experience that things can be improved.
Welcome Back!
Iāve heard nothing but good things about Rosieland, so Iām excited to have your leadership at MoT.
I think you should be aware of the changing landscape for those entering Quality Assurance as a career. I lead Women Who Code DFWās testing club and see an influx of folks graduating from QA bootcamps. I also find myself meeting folks who are interested in getting into tech, but they donāt know what job theyād like.
I think MoT can be a valuable learning tool to fill the gaps for folks who want to learn the role theyāve jumped into. I want every newbie to have a welcoming community to ask questions and receive peer support.
Software testing should be an intellectual process undertaken by professionals, but there are not many software testers worthy of being called professionals. And the intellectual process has been almost entirely eradicated in the quest to reduce testing time, de-skill it and āautomate everythingā.
Software testing is now largely controlled by development managers who know absolutely nothing about it and are not interested in learning. They also know nothing about risk, which is a key consideration for a professional tester.
This is why I abandoned a 20+ year career in exploratory testing and only do accessibility testing now. As James Bach once said, āI refuse to do bad testingā. And I wonāt work for people I donāt respect.
Iād say the focus has moved from just testing activities to a broader quality engineering remit, where we use our skills across the SDLC. Things like requirements reviews, tooling, environments and data strategies, providing a layer of governance around test activities so that even if people without tester job titles do test activities, they follow good practices.
Also lots of AI buzzwords around as well. Its always interesting to see how many people suddenly become experts in something fairly new within a short space of time
And Iāve seen a lot more focus on leadership and soft skills and less on pure technical skills, which I believe is a really positive step.
Those are the 3 areas where Iāve noted a shift.
What do you see as having changed after 3+ years out of the loop? It would be interesting to have a different viewpoint.
Your experience seems very different to mine, as I havenāt witnessed testing being controlled by dev managers where I have worked. Iāve been lucky to work alongside very talented test managers.
I think its unfair to say that many testers are not worthy of being called professionals. There are many excellent testers who I have been proud to work alongside who are very professional, and try to hone their skills. One of the issues around learning is the expectation for people to do it all in their āownā time. This is wrong - learning during work time should be mandatory, and optional during family time. For those with young families trying to juggle a difficult job and family life, they donāt have the option to spend evenings learning, and they shouldnāt be judged or disadvantaged. Maybe there are some who are ājobbingā testers (and its fine if thatās what they want to be), but others may be desperate to learn but not given the chance. Iād like to see companies do better at giving people time to learn and protecting that time.
I do agree with you on one point - automating everything is a fallacy, which is why we follow the mantra of āAutomate the right thingsā, which brings the thinking, analysing and decision making back into the role.
As always, context is important. Our functional testing work was mostly for digital agencies, most of which never had a strong (or even perceptible) testing culture and often didnāt employ any testers, but at least they used to take our advice.
In the last few years they started to either employ junior testers with perhaps 0 to 2 yearsā experience or they used one of the ridiculously cheap outsource testing companies whose staff have pretty must the same lack of experience. People like that bring so little to the party that either the dev manager or project manager would have to control the testing, which they seem happy to do.
I have always said you could fire 75% of all testers and it would have no effect on the quality of software (as long as you fire the right ones). These are the people not worthy of being called professionals, and weāve all worked with them. Of course there are some very good ones, but as a hiring manager I can tell you itās very difficult to find them.
It may be entirely different in large software development organisations with internal testing teams. I donāt know because Iāve never worked for one. But as a software user, I am inclined to say that the outcome isnāt noticeably different.
I agree about learning during work time. We used to give new testers 15 days off in their first year for training and paid for their AST and RST courses. We bought loads of online training courses for testers to do between projects and took everyone to at least 4 or 5 testing conferences each year. I am sure some other companies did stuff like that, but itās pretty rare.
Welcome back Rosie!
Lovely to hear your voice on the live chat today 27/10/23
I have 2 thoughts which Iāll put in 2 separate responses:
With the continued move towards Agile teams that include QA doing testing - these days very often lone testers are part of agile delivery teams iteratively delivering chunks of product value, which I think is great. It means testers are focussed on the effort of building the product & its delivery, rather than focussing on their testing as a serviceā¦
But this has, I think, had the effect of leaving testers quite isolated in terms of extending their craft/skillset.
It is also a difficult environment for a QA-newbie to start in, for similar reasons - learning/mentorship/confidence/honing critical thinking skills
So MoT and its community has a significant role /opportunity to provide an external supportive community to those lone testers , as well as encouraging these folks to build communities or crafts within their orgs
At TestBashUK in Liverpool, I noted there a continuation of a recent theme of how do Testers/QA progress beyond Senior - so the āindividual contributorā type of pathā¦. And new roles have emerged such as Quality Engineer/ Quality Coach/Quality Advocateā¦.I am excited about these types of roles but concerned at the same time: I think it will take time to see how senior stakeholders across companies feel these roles deliver in terms of company & product value and whether these roles survive/mutate into something else in the longer term
Welcome back! Iāve not been involved in the testing community much the past couple of years and wasnāt too involved to start with. I donāt know if I can to talk to anything wider in the testing world other than my experience but Iād say my role has gotten considerably more technical and also more general/broad at the same time.
Iāve started doing automation which has opened the doors to the technical side a lot more but Iām also looking at devsā code more, as well as starting to get deeper into the world of unit and integration tests. My work is shifting left and Iām starting to talk about testing at a much earlier point in the process than I was just a few years ago. It could just be the company Iām at now and not a reflection of anything wider, though, and in fact they suggested the industry as a whole is starting to embrace waterfall again - no idea if thatās true or not.
It seems like my job is much more geared toward actual quality assurance than quality control like it has been for years until now. Iām not sure if thatās a reflection of the industry though or if itās just where my path has gone so far
Just a few things or trends that I have seen for the last 5 years.
The rise and fall of microservices. It was a hot topic back in 2017, but now it has a mature toolset and a lot of examples where microservices are more of a burden than a benefit.
The same is true for testing approaches for microservices - now it is way more established.
The overhyped blockchain and NFT. Starting around 2016 there was massive hype in the industry on everything that had ā-blockchainā in its name. A lot of projects were created - from really technically advanced to very trivial. The trivial ones got even more money :). But now - the hype is over now, nobody wants to buy an NFT image for a big pile of money. But the good thing is that developers in a blockchain world are more concentrated on building really valuable and ground-breaking technologies with a lot of research, rather than doing simple stuff.
But the testing and automation in the blockchain is still at the beginning of the evolution. Each blockchain has its own toolset, but there is a lack of universal tools or even approaches. The reason - the technology is quite complex and there is a lack of people who talk about complex things in easy terms and educate testers in its field. (I am trying to do it - thatās my mission and passion).
AI and Machine learning (including LLMs like ChatGPT). Thatās hype like a blockchain was a few years ago
Complex technology, very versatile applicability - but not easy to test.
Lack of information to explain things in simple terms (gradually). But there are a lot of courses on AI in general
But we have more and more testers in the field - so the situation might be better in the near future.
The movement to SDETās and shifting left. Not a totally new thing - but at the times of massive layoffs, the companies started to āraise the barā for test engineers. Now companies often need very technically-aware testers (SDETs), who can test, automate, read code, configure simple CICD flows, and do a lot of stuff. So testers should be prepared: to learn programming language, computer science basics, algorithms (to pass coding interviews), system design (to understand the architecture of the systems under test), and a lot more. Despite all general testerās skills like critical thinking or communication.