The Tool Desk
Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Some software-testing practices are useful in one team and wasteful in another. A 2023 collection of opinions from Agile Testing Days community members challenges familiar assumptions about who tests, how much automation is enough, and what quality work should look like. These are practitioner views—not proof of universal rules. Their value is in the questions they prompt: What risk are we addressing? What information will this practice produce? And what will it cost to sustain?
These are context-dependent opinions, not a consensus
Agile Testing Days’ 2023 collection, “Celebrating Unpopular Agile Software Testing Opinions,” asked community members for opinions about agile software testing that they stood by. It is a set of attributed perspectives, not a controlled comparison of testing methods or a survey showing how common each view is. It should not be read as representative of all testers or software teams.
As an Amazon Associate I earn from qualifying purchases.
Eric Proegler’s guiding caution, quoted in the collection, is: “There are no best practices or answers that apply to every context…Instructions that are context-oblivious or context-imperial are potentially harmful.” That is a useful way to read the opinions below: consider the project’s risks, users, constraints, and ability to respond to findings before adopting or rejecting a practice.
Free tools Windows power users keep installed
One-click scans. No signup required.
Testing can be shared across the team
One view in the collection is that anyone on a team can test, and that a teammate’s unwillingness to test can create a bottleneck. It also questions whether a dedicated tester must personally test every feature. The implication is not that specialist testers are unnecessary; it is that testing need not be a handoff in which everyone else treats quality as someone else’s job.
When this view is persuasive
Shared testing can help when developers, product staff, and testers can get feedback early, discuss risks directly, and investigate findings without waiting for a separate testing phase. It can also reduce queues around a single specialist.
What to make explicit
Shared responsibility does not mean vague ownership. Teams still need to decide who investigates particular risks, who can make release decisions, and when specialist skills are required—for example, to assess a complex accessibility issue or a difficult failure mode. Before changing roles, ask whether the people taking on testing have the time, knowledge, and authority to act on what they find.
More testing is not automatically more confidence
João Proença’s opinion in the collection is concise: “Sometimes a lot of testing is exactly what you don’t need.” Read in context, this is a challenge to treating test volume as a proxy for quality, not an argument that testing is generally unnecessary. A large effort can be a poor investment if it produces slow, repetitive feedback or findings the team cannot use.
Judge tests by the information they provide
For a proposed test or test suite, consider the risk it covers, the consequence of missing a defect, how quickly and clearly it reports a problem, and the effort required to build and maintain it. Also ask whether the team can investigate and fix the failures it surfaces. A test that is expensive to maintain and rarely informs a decision may deserve revision; a slower test may still be worthwhile when it covers a consequential risk that faster checks miss.
Regression work and automation need a purpose
Joanna Denni’s cited opinion questions excessive regression effort and automation when people’s work becomes centered on generating and maintaining large suites and acting as release gatekeepers. That concern is about a possible imbalance, not evidence that broad regression coverage or automation is bad in general.
Review a regression suite by asking which important failures it is meant to catch, whether its results are reliable enough to guide action, how much upkeep it requires, and whether it delays feedback in ways that matter. The answer may be to remove redundant checks, improve brittle tests, or retain costly coverage because the risk warrants it. The point is to make the trade-off visible rather than equating a bigger suite with a safer release.
Visual checks are one possible tool, not a quality strategy by themselves
For websites, captured screenshots can help teams inspect visual changes, but they do not replace other forms of testing or establish that an experience is usable or accessible. ScreenshotNeo is a website screenshot API and MCP server for developers. It can return PNG, JPEG, WebP, or PDF captures, with options such as full-page capture and element selection; teams should decide whether those capabilities answer a real testing need before adding them to a workflow. Its API and options are documented at ScreenshotNeo’s documentation. The service says it bills only clean shots; bot checks, blank pages, timeouts, failed loads, and cache hits cost nothing, with the response indicating the page verdict and billing status. It also offers an MCP server for AI agents. Sign up for ScreenshotNeo’s free plan for 1,000 screenshots a month with no card required.
Accessibility testing is a must, in one practitioner’s view
Eduarda Loureiro’s opinion in the collection is that accessibility testing is a must, not merely a nice-to-have. That is a practitioner position, rather than a legal or standards citation. Its practical challenge is to treat accessibility as part of the quality questions a team investigates, instead of assuming that a product is ready because its ordinary functional checks pass.
Teams can use the opinion to ask when accessibility risks are considered, who has the skills to assess them, and whether findings can affect implementation and release decisions. The collection alone does not establish which legal requirements apply, which evaluation method is sufficient, or how accessible a particular product is; those questions require applicable standards, law, and product-specific assessment.
Rank #4
Traditional test cases are not the only route to quality
Butch Mayhew’s opinion challenges the idea that executing traditional test cases is required to release high-quality software. This is not proof that written cases have no value. A defined case can make an expected outcome repeatable and communicable; other approaches can help expose problems that a predefined script does not anticipate.
The useful question is what the work needs to achieve. Where repeatability, auditability, or shared expectations matter, documented cases may help. Where the team needs to investigate unfamiliar behavior, relying only on a fixed list of steps may leave important questions unasked. A team can choose a mix based on the risk and the information it needs rather than treating one form of testing as the sole legitimate one.
Recommended Free Tools
TDD and exploratory testing need not be opposing camps
Lisa Crispin’s contribution argues that teams should learn test-driven development (TDD). The collection also warns that exploratory testing can be overshadowed by TDD and behavior-driven development (BDD). Together, these views prompt a question about balance: does the team get useful feedback from tests designed alongside development, and does it also leave room to investigate behavior that was not anticipated in advance?
Best Value
The collection attributes a numerical claim about bug prevention to Dave Farley, but it does not establish that claim through a primary source. It should not be treated as a verified statistic. The more defensible takeaway from the differing opinions is that teams can consider TDD, BDD, and exploratory work as approaches that may serve different purposes, then inspect what each contributes in their own context.
What one small survey result can—and cannot—tell you
An indexed abstract for an arXiv paper reports a survey of 72 practitioners from eight countries and says respondents considered test management and test automation their most challenging activities. This is a result about that study’s respondents, not a measure of the whole software industry or proof that those activities are inherently the hardest. It is consistent with treating coordination and maintenance as real costs when teams decide how to organize testing, but it does not establish why respondents found them challenging.
A practical way to evaluate a testing opinion
When a forceful claim about testing sounds right—or wrong—use it as a prompt for a local decision rather than a universal rule. Compare the practice with the alternatives on the dimensions that determine whether it will help your team:
- Risk: What failure could this catch, and what would a miss mean for users or the product?
- Feedback: How quickly and clearly will the practice reveal information the team can use?
- Cost: What setup, execution, and maintenance effort does it require?
- Coverage: Which users, platforms, and accessibility concerns does it include or leave out?
- Ownership: Who has the skills and authority to investigate findings and act on them?
Then decide what evidence would change your mind. A practice can be valuable when it addresses a meaningful risk and yields actionable information; it can be wasteful when its cost and delay outweigh what it helps the team learn. That is the common thread in the collection—not a single preferred testing recipe, but a case for examining why a team does the work it does.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
