I use intentionally loose test planning, what the agile manifesto calls “responding to change over following a plan“. It’s a challenge, though, in a large transition project where the project work breakdown structure is 300 lines long and lasts for 6+ months – and testing again is on the critical path. but I’ve been there before. As part of the program I have test leads referring to me in the project organization. We have a list of 25 test cases to proceed over the next week or so. The chat we recently had around how to plan the test over the next days reveals experience and core values:
- The inexperienced test lead would just plow along with no plan.
- The skilled test lead would calculate, triangulate and split every atom
- The experienced test lead would make an intentionally loose plan.
The seemingly professional way to go about it, would be a target of five test cases a day. Yet with experience and data I know that, it would never happen that way. At best the progress will be similar to an S-curve:
Progress will initially be slow, then steep, and then slow. If your progress follows the trends, you are fine. But if it flattens (red dotted line), it’s time to get things in motion. One thing that could be blocking you is the timely resolution of bugs and issues found during testing.
https://www.o2sn.dk/ltabook/
If we really really have to assign a number of test cases performed pr. day – let’s at least make it similar to an S-curve. At best we are surprised by faster progress. We can never know what we encounter during a testing activity. We might have intuition and trends, but no one know the future for sure. It always takes longer than you expect, even when you take into account Hofstadter’s law. The world around us might be in chaos due to organizational turmoil or international hybrid wars. The system under test might have changed since last. The people doing the testing might have changed or the tooling disrupted.
The surprising fact from the book “How big things gets done” is that no less than 92% of megaprojects come in over budget or over schedule, or both. IF they are rarely succesful with all their time, budget and quality parameters – how could you ever? As project lessons and experience grows, the lesson is really that we can only plan to a certain degree. As with the IT systems we test, we need to aware that we cannot predict everything. The rest we need to leave for discovery during “operation” and react to with proper instrumentation. Instrumentation and observability tools in the system of delivery – not in the system being delivered. See the difference?
Every organisation is perfectly designed to get the results it gets.
Jonathan Smart, BVSSH
Command and control is rarely the best move.
I often say to myself, that “command and control” leadership is a sword be wielded carefully. I can always dial up control and require deliveries – and escalate if not sufficiently on time. But I see again and again that even professional IT people deliver what they are motivated to do, and my requests are one one request in the inbox. I need to work with them again and again to want to deliver my requests. The curse and freedom of being a staff-level tester is that I daily have to deal with intrinsic motivation. Intrinsic motivation and politics. Oh, and and knowing when to apply pressure. So intrinsic motivation, politics and knowing when to apply pressure.
Would you work with someone who always escalated every minor misstep? What would that tell you about the mentality of the organization? What does it tell you about the leadership principles if every step is not based on trust or being safe? Are you shooting the messengers or are failures shared as organizational learnings?
Which reminds me of another framing of the meme above, below by Timo Koola:
The inexperienced delivery lead just ships what the customer asks. The intermediate delivery responsible stresses on figuring our and debating what the customer want. But the guru delivery lead focus on what the customer asks. It pays of in the long run. This angle is probably not universally applicable. I know my integrity to never kiss the ring, but rather flow around big blockers. In my recent projects I see more repeat business, more political power and recognition in respectfully giving what the customer reasonably asks. I have tried to go into “legal mode” to many times to see that there is rarely any long term value in grand standing.
But don’t take my loose planning as a lack of planning – it’s intentionally loose.