TALES OF TESTING
Testing, Leadership, Software and more...
Quality Experience(QX): Co-creating Quality Experience for everyone associated with the product.3/7/2020 I have been talking about Whole Team Quality via Whole Team Testing for a couple of years now. During my workshops, I am often asked if testing can only be extended for programmers in a team. Pretty interesting question it is and my answer is obviously "no". Though I usually explain in my workshops on how to extend testing to roles beyond programmers in a team i.e. for UX or PO roles, I realized that I have not given deep thought to it and to how exactly testing could be extended to other disciplines in a meaningful way.
I read books, discussed with my colleagues, did my research and the outcome has been what I would like to name as QX i.e. Quality Experience. If QA (read that as Quality Advocates) and UX professionals collaborate in a meaningful way, I firmly believe they can co-create a Quality Experience for everyone associated with the product. So, what is QX after all? QX stands for Quality Experience. For sake of understanding, you can call it a marriage between QA (read it as Quality Advocacy please) and UX. After some need-based discussions and interactions with my UX colleagues, I realized that we can achieve a lot more if we work closely together regularly. The key idea of QX is about facilitating the collaboration between QA and UX so that they can contribute to what I would call a "Quality Experience" of the product. This is both for the end-user and for those who build the product itself. I believe that with some process optimizations, mindset enablers for testers as well as UX designers, and following some heuristics I have created, it is possible to kickstart the QX journey if that idea interests you. But what's the need? Is QA alone not enough to cater to product quality (or UX alone not enough for better user experience)? Well, I am afraid, it is not. Not at-least when you believe that Quality is value to someone (who matters) and when multiple stakeholders matter at the same time. Let me explain. Imagine that a new design change required in a product is a revenue booster for the company but it is also likely to impact the user experience. Testers often end up with oracle problem in such situations and can not decide what their quality criteria should look like. Of course, the Product Owner can be consulted here for a final decision but that's not the point. We testers are in the information business, after all (yes, even if you follow the Modern Testing). I find it important for testers to be able to make a comprehensive information gathering and present that information to decision-makers so that they can make an informed decision based on that. Now, if testers lack the tools and mindset to figure out how to go about solving such problems and gathering information that would matter, their job would be poorly done. And if you are still not convinced, I highly recommend you to read Weinberg's latest book around System Design Heuristics. This book could not have come at any other better time for me. I started reading it while working on the QX concept and it has given me some interesting insights to sharpen my thinking around this topic. On the other hand, let's assume that a UX/Interaction designer has been given a problem to solve or has been asked to create a new solution for some product. How do they ensure they have gathered enough information to do that job right? How about historical incidents or say hidden technical challenges or simply put edge case scenarios, cross-functional dependencies, and so on? I believe that having this information at hand can greatly benefit how UX designers can approach the problem and solve that in a better way. Therefore I think that an engagement between UX and QA can help both of them to perform their job even better. And hence QX seems to be a good way to go about it. Well, how does it work? Here are some ideas that I experienced to be working quite well. See if they work for you too? 1. Cross-discipline training for QA and UX For successful collaboration, it is important that QA and UX understand each other well and that they speak and understand each other's language. More important is to understand the mindset with which they both operate. I have experienced difficulties in understanding UX's point of view sometimes and my UX colleagues made the same experience. But a conscious effort made towards understanding each others' language helped us solve those issues and facilitating the collaboration thereby. That said, we recommend that testers and UX designers start from understanding each other's roles and mindset. Attending cross-discipline training should help but if that is not possible, try doing "pairing sessions" at least. 2. Process changes or optimizations A great deal of it depends on what kind of team set up you have. Some teams may have dedicated UX designers, some organizations have UX teams as a "lateral" service provider for different teams and some teams simply don't have UX people. Their designs are usually outsourced or made by engineers themselves. Some recommendations I would like to make here are: 1. Involve testers as early as possible and make them also part of the design process/discussions. UX will thank them for lots of useful info which might result in the faulty design if missed. 2. Early and frequent communication between UX and QA would help. Try "brainstorming" sessions for early design stages. Ask tester for hidden scenarios or technical hacks to go around things. Ask them for user complaints or known production issues surrounding the design under discussion. 3. Testers may perform focused UX testing and consult UX from time to time for their design-related findings. A "pair testing" session with UX expert can greatly benefit testers for more test ideas surrounding usability and human-computer interaction under different contexts. 4. Instead of creating a misinformed bug report surrounding usability and UX, testers can always consult UX colleagues first for feedback around their findings and use them as their oracles. Most of the time, UX people have information and insights (from their interaction with real users, test sessions they perform, qualitative and quantitive data analysis they do, etc.) that explain why that design is made a certain way. 5. Not every design change goes via the standard UX lead design process. Engineers sometimes have to make decisions that may result in a change in product design or impact certain product behaviors. Such changes can always be sent to UX designers for their feedback. The tester can play a big role in making this activity happen on behalf of the engineering team. 3. UX testing heuristics for testers Mere exchange with UX colleagues without having proper knowledge of how to test for better user experience can be futile. Which is why I recommend testers to get good at UX testing too. By that, I don't mean typical usability testing or accessibility testing alone. After giving a deep thought to all possibilities involved, I have come up with the following heuristic for testers. Keep in mind that it can also be used as a means to facilitate collaboration and have a better discussion with UX colleagues. Not everything might be applicable for all the contexts. Choose those that fit best in your context. Here you go: Problem -To come up with relevant test ideas, Testers and UX must be on the same page in terms of understanding the problem they want to solve with the proposed design. UX often get first-hand information from the PO about the problems, which testers sometimes are lacking. Trying to understand what problem UX wants to solve with their solution, can open up lots of possibilities for testers and confine them to think in the right direction. It would also spare them from coming up with the right questions and perform better impact analysis of the change.
User Needs - This is highly subjective and might vary from project to project. The key idea is to understand what User Needs are getting addressed through the design and if you as a tester can foresee any challenges/impact of that. This can also be very well made as a part of 'understanding the problem' part. It does not matter how you do it, but what matters most is that you know what aspects of user needs you are dealing with.
User vs Business Needs
Finding Balance - Solving the Oracle Problem When you as a tester are unable to decide if the proposed solution is good for the product but bad for users and vice versa, use some of the methods below to make your decision-making more concrete and practical. Based on the configurations of teams, there may be no dedicated UX expert to cater to these needs. It is therefore advisable for testers that they wear UX hats and find the balance by asking the right questions before it is too late. Plenty of efforts could be saved if testers too are involved in design/UX decisions early where dedicated UX expert is absent or designs have been made by third party services etc. What Must Not Change? Whatever we ultimately do, what are the things you don’t want to be changed? (from System Design Heuristics by Jerry Weinberg) Introducing design changes in an existing product that are targeted at a specific goal by UX designers, can harm things that are not meant to be affected by that. I highly doubt if there are deliberate efforts made to analyze this regression impact on the design level itself. This is why, if a tester asks this question right in the early phase of a design change, it is likely to save lots of rework for the later. “If we don’t start that way, it’s all too easy to lose track of the unchangeable.” - Jerry Weinberg. Impact Analysis What are the visible and invisible parts of the product that are impacted by this change? And what is the impact?
Creativity Running out of test ideas to decide if the solution is perfect? Pour some creativity in.
Exactness, Intuitive and Counter-intuitive Design Most of the time, testers are more focused on the technical and functional aspects of the product so much that they unknowingly tend to ignore looking for obvious problems. A deliberate attempt needs to be made to look at the product like an unexperienced user to see if they will understand what we expect them to. To do this:
Well, I wish I could work on this and develop this idea even further. But for the benefit of time, I would stop here. Maybe I would get back to this again when I have more ideas. But in the meantime, feel free to comment below and share more ideas if you like. And do not forget to tell me how you find the QX idea so far. And, I recommend you to read System Design Heuristics by Jerry Weinberg. The book gave me lots of ideas to ponder upon. Until then...
0 Comments
Disclaimer: Purpose of this blog is not to undermine the importance of automation skills for testers but rather to highlight the lack of awareness in our industry for tester's usefulness in teams beyond typical testing and automation tasks. Recently, I was invited to conduct a workshop around "Whole Team Testing" at Test Leadership Congress in New York. Participating in this conference has been a great experience, especially because I truly believe in the need for conferences dedicated to test leadership and career path for testers in changing era. The discussions I had there with fellow testers, managers and directors of engineering/testing were interesting. Well, not only interesting but insightful and with lots of ideas, concerns, observations to ponder upon. Out of the things we discussed, what held my attention the most was the growing concern of evaluation of testers in Agile teams. The evaluation can be for hiring or for their overall performance in team. What is the problem exactly? If you analyse the trend from State of Testing report, it is evident that with increasing Agile adoption, centralised testing units are mostly getting dissolved and testers are now reporting to Dev/Team leads. With this change of structure, hiring new testers and evaluating those in teams naturally becomes the responsibility of the Dev/Team leads. And this is where things start to get interesting. How? Let's take a look:
But this situation is dragging us into bigger problem. We are giving too much importance to things that are just a part of big scheme of things that matter and contribute to software quality and teams' overall ability to ship quality software, faster and frequently. Of course automation skills and other technical skills are helpful, rather important I would say. But that has become a norm now and what we really need at this point of time are tools to see beyond this norm. If we want excellent testers in team who can really add value to software quality, it's high time that hiring managers come out of their obsession of hiring testers based on the norms that they can most comfortably evaluate. Please, it is not about what we are most comfortable with but what Agile team really needs from a tester to ship a quality software. Heavy emphasis on technical skills for testers is not my big concern, but lack of awareness around other things that testers must get evaluated against, concerns me the most. Purpose of this article is to highlight some of those areas (beyond typical technical skills) that I think managers/leads may want to consider while hiring new tester or evaluating those in their teams. Why do we need testers at first place? One may argue that in an era of AI and advanced automation where every check can be easily automated, why do we need human testers at all? That's interesting argument and before I explain why we are talking about it, I want you to have a look over the diagram below: The diagram represents a system that we as a team operate into. I created it based on my experience as a tester so far and by interviewing some programmers I have closely worked with. Looking at bunch of things mentioned there, you can figure out what all can happen (rather typically happens) when tester in team is not available. If you read the diagram carefully, you will notice that impact on finding and reporting bugs is just one segment that gets highlighted when tester is missing in team. Impact on writing new tests or automating them is another segment. But is that all you need testers for in a team? Certainly not! The purpose of having testers in team is way beyond finding bugs, writing tests or automated scripts for that matter. If utilised to their full potential, testers in team can very well serve as a mode to protect your system from running into collapse or explosion mode (please read more about that in "How Software is Built" by Jerry Weinberg). How? It's simple. By providing system related feedback to the controller which is typically team lead or test manager in typical team set-up. The better you utilise tester for their abilities the better feedback you get from them which in turn can help you command better control over your system and protect it from collapse/explosion. You need tester in team to provide you with information that you as a stakeholder can use for making better decisions. And when I say feedback, it can be anything from information about bugs, risks highlighting, asking context revealing questions, questioning user stories or decisions made, finding out historical data, highlighting third-party dependancies, sharing news about decisions made by cross functional teams, collaborating with other disciplines to create valuable assets, their observation around team dynamics, their predictions about possible failures and so on. And if you are lucky, they can also tell you some great things about quality of your product. The point is, tester in your team can add far more value beyond finding bugs, if you allow them to contribute beyond their typical role. If not, you can always empower them to do so. Here are couple of things that in my experience testers can contribute to and against which you could evaluate them (for hiring or performance reviews etc): 1. Primary analyser of production logs and alerts Production deployments are typically done by programmers who then closely follow the logs to see if there are any obvious issues created by latest build. Consider handing over this responsibility to testers. It is likely to serve multiple purpose:
2. Enhancers of your product's coverage for quality It has very little value if tester in your team just sticks with usual acceptance criteria for the ticket and automates the stuff for you to top it. The real value a skilled tester can add is when they question the very product coverage you have in place and help your team see elements of product that matter from quality aspects. A lot of elements matter when it comes to your product's quality. Functional acceptance criteria is just tiny part of it. A skilled tester would know about these elements and they would educate the team about same. Check out SFDIPOT from HTSM by James Bach or simply have a look on mindmap below (thanks to Albert Gareev) 3. Advocates for Testability Testability is one of the key area I would expect skilled tester in my team to help programmers and designers get it right. A skilled tester would know what makes product testable, how to evaluate it for testability and also advocate the team as in where and how they can improve it. Nothing beats the pleasure of testing better testable product. Here is a heuristic for testability if that makes you curious. Out of various types of testability mentioned there, I would expect the tester to at-least know Intrinsic Testability and how to help their team improve it. 4. Better allies for UX peeps Ever wondered what would happen if System Thinking meets Design Thinking? I firmly believe that challenges faced by testers and UX professionals are more or less same when it comes to ensuring better quality and better user experience. A regular and close exchange between these two disciplines has tremendous potential to create a better user experience with enhanced product quality which I would call as "Quality Experience". If UX designer comes with one best solution for some product problem, a skilled tester with their insights, product knowledge, awareness around cross-functional dependancies can point out variety of ways in which it may fail. This does not mean a tester has to criticise UX solutions as such but early collaboration and exchange between two can help avoid lots of unnecessary research and rework. On other hand, testers can borrow realistic information from UX's research which can help them design tests that matter and prevent themselves from straying into unwanted scope. Testers can borrow statistical data or interaction/persona based information from UX that can help them shape better scope for their tests. Well, this is indeed interesting topic and deserves deep dive. I will stop here on this one for now. A skilled tester would make this collaboration happen and would bring out the best from both the worlds. 5. Friend in need for marketing and user care This is another area of collaboration where skilled testers can add great value to solve team's problems beyond finding bugs. I have listed some of the possibilities where testers can help user care and get benefited in return and I believe same applies to their collaboration with marketing teams: 6. Alert mechanism when system is on the verge of collapse
Well, this is bit tricky but I would mention it anyway. By nature of their work, skilled testers can use their sharp observation skills to observe and understand people, situations and events around them and can draw inferences that can help identify potential risks, before it is too late. Retrospective meeting is a great place for skilled tester to raise flags for potential issues surrounding team dynamics or people issues they observed, in constructive way of course which is where their communication skills can come handy. Testers are usually in touch with fellow testers in organisation from where they get information around what other teams are working on, their future plans, blocking issues etc and this information can be very well used to analyse the impact on their respective team's road-map , work in progress or work items that share dependancies with other teams. A tester who is skilled enough to foster network and relations across teams, can certainly help bypass blockages when team needs it most. So, those are some aspects in my opinion and experience, a skilled tester can contribute value to project team beyond their traditional tasks. If you are to a hire new tester or would like to fairly evaluate/mentor tester in your team, I would suggest give these ideas some consideration. It's high time that we look at tester as someone beyond bug finder. They can do wonders for your team and can help to accelerate shipping of quality product, if allowed to use their full potential. I shared what I think can be helpful. Feel free to chip in your ideas .... |
AuthorA passionate & thinking tester. Trainer & student of the craft of testing. Father, Foodie and dog lover. Chief Editor and Co-founder of Tea-time with Testers magazine. Categories
All
Archives
January 2021
FREE subscription
|