Shoestring manual testing. When you can't afford automation, yet.
You don't always have the budget, the tools, the headcount, or the stable system under test that test automation needs. This talk is the playbook for the other case — a disciplined manual testing program built around cheap technicians, fast training, and a business case any manager will fund. The goal is not "manual forever." The goal is enough signal to ship safely while the product is stabilizing.
Abstract
Total test automation is a mirage for most real projects.
Complete test automation sounds ideal — start the test, leave it running, come back for results. In practice, most real projects hit at least one of the obstacles: the system under test changes too fast for scripts to keep up, there's no budget or time for script development, the right tool doesn't exist at a reasonable price, the team doesn't have the automation skill yet, or the critical quality risks are the ones automation is worst at — usability, localization, error handling, installation, documentation. The answer is not to give up on testing. The answer is a well-run manual testing program.
This talk walks through that program end-to-end. How to size a manual test effort using two honest rules of thumb. How to structure the team so one test manager and two or three test engineers can supervise a much larger bench of test technicians. Where to find good technicians cheaply — students, customer-support crossovers, moonlighters, data-entry staff — and what each group's failure mode is. How to train them in a week, not a quarter. How to run multiple shifts without burning people out. Which test types you should not run manually no matter what. And how to sell the whole thing to management as a flexibility and cost-coverage play, not a compromise.
It pairs naturally with the ten-step bug-reporting process in the QA Library — the one piece of craft that every manual test technician has to master first. We link out there rather than repeat the material.
“Using manual testing inappropriately can mislead people about the extent of test coverage. Using automation inappropriately usually just fails. Pick the right tool for each test.”
— Rex Black, Inc.
Outline
What the talk covers, in order.
What manual and automated testing actually are
Manual testing is developing and executing tests that rely primarily on direct and continuous human interaction — especially evaluating correctness and test status. Automated testing is developing and executing tests that can run unattended, including comparing actual to expected results and logging status. The simple field test: if the tester starts the run, walks away, and comes back for results, it's automated; if the tester enters data, observes behaviors, and actively runs the tests, it's manual. Manual testing may still use tools (for load simulation, for capturing performance stats, for creating specific failure conditions) — tools are not automation. Unattended is automation.
The mirage of total automation
Complete automation looks ideal on paper and almost never works in practice. Five obstacles come up again and again: the system under test changes frequently (and scripts break faster than they're written), no budget or calendar exists for script development, no suitable tool exists at a justifiable price, the team lacks automation skill or experience, and — the one teams most often miss — some of the critical quality risks aren't amenable to automation at all. Usability, localization, error handling, installation, documentation: all of these are cheaper and more honest manually.
Sizing the manual test effort honestly
Start from the tests, not the calendar. For each test case, know person-hours of effort, wall-clock duration, and prerequisites. For the team, know the overhead: bug reporting, status documentation, communication and email, breaks, and — the big one — blocking issues and debugging. Then apply two rules of thumb before you promise anyone a schedule. Expect six hours of actual test time out of an 8–10 hour day. Expect 75% downtime against a buggy build and 25% downtime against a good one. Testers do not test for eight hours; testers test for six hours and spend the other two fighting their environment.
- Six productive test hours out of an eight-to-ten hour day is the norm, not the floor.
- 75% downtime on a bad build means you need four days of calendar to get one day of test coverage.
- Include reporting, status, comms, breaks, and blocking issues in the budget up front — not as excuses at the end.
Team structure — one manager, engineers, many technicians
A shoestring test org has three roles. The test manager has overall responsibility for the team, the schedule, and the release recommendation. Test engineers provide technical leadership and oversight — they design tests, troubleshoot the hardest failures, and supervise a small group of technicians. Test technicians do most of the actual manual test execution. The leverage in this structure is the ratio: one or two engineers can ramp and supervise several technicians, which lets you expand coverage without tripling salary cost. It also creates a career ladder — technicians grow into engineers over time.
Where to find good test technicians cheaply
Four groups work. College and technical school students and grads — CS and engineering majors are especially eager; the tradeoff is that they may not be able to commit enough hours. Technical and customer support staff — they understand the problems customers face; the tradeoff is that they sometimes want to solve the problem, not just identify it. Moonlighters — they bring real experience from other jobs; the tradeoff is focus, and occasionally their day job pulls them out. Data-entry personnel — they're already experienced computer users; the tradeoff is that they may not be curious enough to dig into a bug. Pick the mix that fits your product and your office culture.
- Students: eager, cheap, sometimes unavailable.
- Support crossovers: customer-oriented, sometimes too solution-focused.
- Moonlighters: experienced, sometimes distracted.
- Data-entry staff: fluent users, sometimes low-curiosity.
Training — what's universal, what's company-specific
Separate the two. Universal skills — manual test-case execution and bug reporting — deserve written docs, training, and coaching regardless of hire source. Company-specific processes — how to get a network login, badges, email, the phone system, the test repository — need documentation or a mentor but are not skills, they're rituals. Do not waste training time conflating the two. The technician who cannot reproduce and document a bug is broken in a way that does not get fixed by a second badge tour.
The ten-step bug report (cross-linked, not repeated)
Every manual test technician on a shoestring team has to be fluent in one thing above all others: writing a usable bug report. The ten-step process — structure, reproduce, isolate, generalize, compare, summarize, condense, disambiguate, neutralize, review — is the single skill whose quality most determines whether the manual program pays off. We cover the process as its own QA Library piece rather than duplicating it here; see the linked page.
- Structure, reproduce, isolate, generalize, compare — the investigation half.
- Summarize, condense, disambiguate, neutralize, review — the writing half.
- Grammar and spelling count: a poorly written report can render a technically correct one ridiculous.
Project realities — training on the job
Test technicians are almost always added late, which means training happens on the live project. Plan for it: stagger technician hires by a week or two so test engineers can ramp them one at a time; account for the productivity hit on the whole team while ramping happens; and reflect that hit in the case-progress estimate you commit to. Pretending ramp-up is free is the single most common reason manual programs miss their first milestone.
Multiple shifts — more coverage, same salary envelope
A night shift (~4 PM to 1 AM) and a graveyard shift (~midnight to 9 AM) are a realistic way to stretch a shoestring team. Reasons: schedule (more tests per day), resources (one hardware platform covers two or three eight-hour test blocks), preference (some technicians genuinely want off-hours work), politics (team coverage without burning out the day shift). Get the mechanics right: make the shift explicit at interview, pay an undesirable-hours bonus, overlap shifts for hand-offs, and invest deliberately in team atmosphere — out of sight is out of mind until the first conflict.
What not to run manually
Manual testing is the right tool for functional tests, use cases, error handling and recovery, localization, user interface, configurations and compatibility, operations and maintenance, date and time handling, installation/conversion/setup, and documentation and help screens. Manual testing is the wrong tool for monkey/random testing, load/volume/capacity, reliability and MTBF, code coverage, performance, and standards compliance — for these, any shoestring program still has to buy or build some automation, because running them manually either doesn't work or actively misleads people about coverage.
- Manual-first: anything a human judges better than a machine — usability, localization, docs, configs.
- Automation-first: anything that requires volume, repetition at speed, or machine-precise measurement — load, performance, reliability, coverage.
Managerial caveats
Four things the test manager cannot delegate. Monitor the performance of inexperienced technicians — some will not work out and need reassignment. Pay attention to evening and graveyard shifts, including personnel dynamics. Recognize that some tests require domain knowledge your technicians don't yet have and keep those on experienced staff. Provide a real career path so long-term technicians can advance into engineer roles or other parts of the company — the alternative is constant re-hiring and retraining, which kills the economics.
Selling management on the plan
The manual-testing program has to be pitched as a solution, not a concession. The business case: flexibility and responsiveness (technicians ramp fast), cost and coverage (cheaper bodies means more coverage on the same budget), schedule (test cycle time goes down), staff retention (test engineers burn out less when they have technicians to delegate to), and — worth saying explicitly — helping inexperienced people jump-start a technical career. Bring your manager a plan, not a request for more money.
Key takeaways
Four things to remember.
Total automation is a goal, not a precondition.
A well-run manual testing program is the bridge from "no testing" to "the right automation at the right time." Teams that wait to automate everything before they test anything lose the window.
Budget the overhead before the test count.
Six productive test hours per day, 75% downtime on bad builds. Reporting, comms, debugging, and blocking issues are part of the job, not excuses. Plan them in.
Leverage comes from the org shape.
One test manager, a couple of engineers, many technicians. The ratio is what makes the economics work. Flat teams of senior people will run out of money before they run out of tests.
Pick manual for what humans judge best.
Usability, localization, error handling, installation, documentation — these stay manual regardless of budget. Load, performance, reliability, coverage — these earn automation even in a small program.
Worked examples
One bug. Eight drafts.
Embedded-product case study.
A reference point from a real embedded-product engagement. The specific device category is less important than the shape of the team, the cycle cadence, and the person-hour math — which still scans for most hardware-plus-software programs today.
Six test technicians.
Two test engineers supervising the technicians.
One test manager overseeing the team.
Toolsmiths creating and operating load-simulation tools.
Full pass: ~2 weeks, ~300 manual test cases, ~400 person-hours.
Smoke test: 2–3 days at each build candidate.
Validation test: ~1 week at release candidate.
400 person-hours / 6 productive hours per day = ~67 technician-days per pass.
Six technicians working a pass in parallel finish in ~11 working days — which matches the 2-week calendar.
At 75% downtime on a bad build, the pass slips to ~4 weeks. Build your schedule with that worst case in view, not the ideal.
Keep reading
Related pieces.
More for this audience
Articles, guides, and case studies tagged for the same readers.
- Whitepaper
Evaluation Before Shipping: How to Test an AI Application Before It Hits Production
The release-gate playbook for AI features. Covers the five evaluation dimensions, how to build a lean golden set, where LLM-as-judge is trustworthy and where it lies, rollout mechanics with named exit criteria, and the regression suite that keeps a shipped AI feature from quietly rotting in production.
Read → - Whitepaper
Choosing the Right Model (and Knowing When to Switch)
A practical framework for matching LLM model tier to task. Covers the four axes (capability, latency, cost, reliability), cascade routing patterns that cut cost 60 to 80 percent without measurable quality loss, switching costs you did not plan for, and the worked economics at 10K, 100K, and 1M decisions per day.
Read → - Whitepaper
Beyond ISTQB: A Multi-Domain Certification Roadmap for Technical L&D
Most engineering L&D programs over-index on a single certification family, usually ISTQB on the QA side, AWS on the infrastructure side, and under-invest across the rest of the technical domains the org actually needs. This paper covers a multi-domain certification roadmap (QA, AI, cloud, data, security, project management, software engineering) with sequencing logic for each level of the engineering ladder, plus the maintenance discipline that keeps the roadmap relevant as the technology shifts underneath it.
Read → - Guide
The ISTQB Advanced Level path, mapped
The Advanced Level landscape keeps changing — CTAL-TA v4.0 shipped May 2025, CTAL-TM is on v3.0, CTAL-TAE is on v2.0. This guide maps all four core modules, prerequisites, exam formats, sunset dates, and which module a given role should take first. Links directly to the authoritative istqb.org syllabi.
Read → - Whitepaper
Bug Triage: A Cross-Functional Framework for Deciding Which Defects to Fix
Bug triage is the cross-functional decision process that converts raw defect reports into prioritized action. Done well, it optimizes limited engineering capacity against risk; done poorly, it becomes a backlog-management ritual that neither fixes the important defects nor drops the unimportant ones. This whitepaper covers the triage process, the participants, the six action outcomes, the four decision factors, and the governance disciplines that keep triage effective in continuous-delivery environments.
Read → - Whitepaper
Building Quality In: What Engineering Organizations Do from Day One
Testing at the end builds confidence, but the most efficient quality assurance is building the system the right way from day one. This whitepaper covers the upstream disciplines — requirements clarity, lifecycle selection, per-unit programmer practices, and continuous integration — that make system-level testing cheap and fast rather than the only thing holding a release together.
Read →
Where this leads
- Service · Quality engineering
Software Quality & Security
Independent test programs, security testing, and quality engineering for systems where defects cost real money.
Learn more → - Solution
Risk Reduction & Clear Decisions
Quality programs and decision frameworks that shift risk discussions from anecdote to evidence.
Learn more → - Solution
Reliable Software at Scale
Quality engineering programs for organizations whose software is now operationally critical.
Learn more →
Want this talk delivered in-house?
Rex Black, Inc. delivers every talk on this site as a live workshop, a keynote, or a conference session. Tailored to your stack, your team, and your timeline.