The Value of Probing Questions in Development

Why they really need us even if they think they don’t.

“Software developers keep asking irritating questions. They want to talk to people outside their department. They keep asking what we’re trying to achieve instead of just doing what we ask. Let’s just use AI.”

We really frustrate people. Granted, many of us are just frustrating people (https://lnkd.in/g2zvq8Mv). But we also need to perform our role.

To be blunt, if you don’t ask probing questions at all levels of your assignment to understand what the software you’re building needs to achieve for people, and instead just build what you’re told, then the detractors are right—you are replaceable. Worse, if you then insist on executing without regard for cost or the end user, you’re doing harm and you should be replaced.

We are useful because we channel our experience toward a purpose. We translate partially articulated, partially thought-out, and sometimes contradictory requirements into achievable implementations. And even if we trust our leaders and stakeholders, we need to understand what they want.

We use this understanding of purpose to deliver outcomes instead of output. We take our knowledge of established patterns and techniques. We apply our experience with what tradeoffs can and should not be made in any specific context. We use our dread of how it can all go horribly wrong to prevent horrible things from happening. And we do this even when our bosses are actively telling us not to.

This matters most urgently when systems fail, when revenue opportunities drive changes that conflict with user expectations, when a viral trend threatens to expose people’s sensitive information, when a framework exploit threatens the system, or when usage scales beyond the original design.

But it matters just as profoundly day to day because there’s never enough time or money to waste. Those irritating, obnoxious questions are often the clearest path to success—building what people actually need, can maintain, and will use instead of what they say they want in the absence of our partnership.

If we do this, the AI tools will change many aspects of our work and we will become less of a resource constraint, but when an opportunity is important and challenging enough, we will continue to be necessary.