In software, we rarely have fully stable requirements. Even when we do, the only measure of success that matters is whether our solution solves the customer's shifting idea of what their problem is. Agile, conceived in the late nineties and formalised at the 2001 Snowbird meeting, is often described as a methodology, when really it's a set of values and principles: prioritising the satisfaction of the customer through early and continuous delivery, welcoming changing requirements, delivering working software frequently, and building projects around motivated individuals trusted to get the job done.
Is agile implementation agile enough?
Concept is invariably easier than delivery, and translating these principles into concrete practice has proved challenging. It's been suggested that agile has drifted from its original concept, with organisations reaching for measurable, controllable process because that's more comfortable than genuine change, even though welcoming changing requirements is one of the most important agile principles in the first place. Real agile organisations are the ones constantly changing, with a suitable environment that lets trusted individuals thrive, not just at the team level but across budgeting and governance too.
Agile is fundamentally about managing uncertainty through learning and discovery rather than prediction, a shift away from a top-down model like Waterfall towards one where uncertainty is embraced and managed through experimentation.
Some common risks in implementation
There's an argument that once you've defined rigid processes like Scrum in detail, you've already limited your agility; implemented properly, Scrum can and does work as a methodology, but blind implementation of it isn't automatically agile. Agile in its true form is a genuine mindset shift, and teams aren't always given the time or training to take it fully on board.
Another common risk is a hybrid Waterfall-Scrum in disguise, where agile iteration happens within analysis, development and testing, but a top-down Waterfall process, budget proposal, estimates, design sign-off, still governs everything around it, often meaning users only discover it wasn't quite what they wanted after a long process has run its course. Delayed delivery is another risk, where important parts of a product wait for less important parts to be batched together, slowing response times and adding cost. The agile solution is small batches, which shortens feedback loops and helps build quality in continuously.
Small is beautiful
The agile solution to most of this is a lean approach: minimise waste, take small steps, and loop. Understand where you are, take a small step towards where you want to be, evaluate what happened, and repeat, choosing paths that are easily reversible wherever a choice arises. For this to work, people need genuine autonomy to find their own way to the next step, a real departure from a Waterfall culture where work is broken down and assigned, then reassembled at the end.
If you want to employ agile values properly, set goals that are easily measurable, take an experimental approach to reaching them, and keep trying to improve the process itself as you go.