The Agile Approach: A Different Way to Build Software

The agile approach is an iterative, incremental method of developing software and managing projects that prioritises customer feedback, adaptability, and working deliverables over upfront planning and rigid documentation. The approach emerged in the 1990s as a response to the slow, document-heavy waterfall methodology that dominated software development through that era.

Seventeen software developers met in 2001 in Snowbird, Utah, and produced the Agile Manifesto — a four-value, twelve-principle declaration that has shaped how technology companies build software for the past quarter century. The agile approach now extends well beyond software into project management, marketing, hardware development, and even some manufacturing contexts.

The fundamental shift agile represents is from "plan everything upfront, then execute" to "plan incrementally, adapt as you learn." Waterfall methodologies require defining the entire project scope, requirements, and design before any building begins. Agile methodologies break work into small iterations (typically 2-week sprints), build working software each iteration, get customer feedback, and adjust the next iteration based on what was learned.

The result is faster delivery of valuable functionality, better adaptation to changing requirements, and substantially less wasted work building features that turn out not to matter. The agile approach pairs naturally with the broader practices captured in the Agile overview.

Adopting the agile approach is more than implementing new processes. It requires culture change — empowering teams to make decisions, encouraging customer involvement throughout development, accepting that requirements will evolve, and measuring success by working software delivered rather than documentation produced. Organisations attempting agile transformation without the cultural shift often produce "fake agile" — implementing standups and sprints while keeping waterfall command-and-control underneath. Real agile transformation takes 1-3 years for most organisations and rarely succeeds without leadership buy-in, dedicated coaching, and willingness to retrain teams over extended periods.

The financial impact of waterfall failures helped fuel agile adoption. Studies through the 2000s consistently showed waterfall software projects failing at high rates — over 50 percent late, over budget, or cancelled outright. The Standish Group's CHAOS reports documented these failures for decades. Agile approaches emerged partly because the existing methodology was clearly not working at scale. Organisations that adopted agile during the 2000s-2010s often saw measurable improvement in delivery success rates, though attribution to methodology versus other changes (better tools, better engineers, smaller projects) is debated.

The Four Values of the Agile Manifesto

The Agile Manifesto's four values establish priorities, not absolutes. Each value pairs two things and indicates which the agile approach prefers when trade-offs arise. Value 1: Individuals and interactions over processes and tools. The phrasing acknowledges that processes and tools matter; the value just says when both cannot be perfect, prioritise the human side. Teams that talk to each other and work well together produce better results than teams with rigid processes and elaborate tools but poor communication. Hiring for collaboration capability and investing in team chemistry pays back more than mandating specific software platforms.

Value 2: Working software over comprehensive documentation. The value does not say documentation is bad; it says working software is more valuable. Documentation that does not produce working software is waste. Documentation that helps deliver working software is appropriate. Many waterfall projects produce thousands of pages of requirements and design documents, then build software that does not match what users actually needed. Agile flips the priority — build something working, learn from it, document just enough to support continued development.

Value 3: Customer collaboration over contract negotiation. Customers and developers working together produce better software than customers handing fixed requirements to developers and demanding adherence to contract terms. Customer involvement throughout development surfaces misunderstandings early when they are cheap to fix; rigid contract enforcement surfaces them late when they are expensive. Value 4: Responding to change over following a plan. Plans matter as starting points, but reality reveals truths that plans cannot anticipate. Teams that adapt to learning produce better outcomes than teams that grind through outdated plans pretending they still apply.

The manifesto's wording matters. It does not say processes do not matter or documentation is useless — it says when trade-offs must be made, prioritise people over processes and working software over documentation. The framing as priorities (over) rather than absolutes (only) is essential. Teams that interpret agile as anti-documentation or anti-process miss the point. The right amount of process and documentation depends on the context; agile just establishes that humans and working code take priority when the trade-off appears.

Key Agile Practices: How Teams Actually Work

Sprints are time-boxed iterations, typically 2 weeks long, during which a team commits to delivering a specific set of features. Sprints establish rhythm and predictability; everyone knows when reviews happen, when work commits, and when learning becomes the next sprint's input. Daily standups (15-minute meetings, typically same time each day) keep the team aligned on progress and obstacles. Each team member shares what they did yesterday, what they will do today, and what blocks them. The standup is for coordination, not detailed problem-solving — that happens after the standup with the relevant subset of people.

Sprint planning at the start of each sprint reviews the prioritised backlog and selects work for the sprint. The team commits to specific deliverables. Sprint review at the end of the sprint demonstrates the completed work to stakeholders. Stakeholders provide feedback that shapes the next sprint's priorities. Sprint retrospective (separate from review) examines the team's process — what worked, what did not, what to improve. The retrospective is critical for continuous improvement; teams that skip retrospectives stop improving.

Backlog grooming (or refinement) maintains the prioritised list of work the team will pull from in future sprints. The Product Owner (in Scrum) maintains the backlog with input from stakeholders; the team participates in refining items that are large or unclear. Continuous integration builds and tests code changes automatically as developers commit them, surfacing integration issues immediately rather than at the end of long development cycles. Continuous deployment extends this — code that passes tests deploys to production automatically, sometimes many times daily in mature teams.

The Definition of Done establishes what "complete" means for the team. Common Definition of Done elements include: code is written, code passes automated tests, code is peer reviewed, documentation is updated, code is integrated into the main branch, code is deployable to production. The team agrees on the Definition of Done and applies it consistently. Without an explicit Definition of Done, teams ship work that they later discover was not actually complete — missing tests, missing review, missing integration. Establishing this definition upfront prevents recurring rework.

Roles in Agile Teams

Scrum, the most common agile framework, defines three roles. The Product Owner owns the product vision and the backlog of work. They prioritise features, define acceptance criteria, and decide what gets built when. They represent customer interests to the development team and represent the team's progress and capacity to stakeholders. The Product Owner is typically full-time with one team; multiple Product Owners across multiple teams produces conflicting priorities and team disruption.

The Scrum Master facilitates the agile process. They run the ceremonies, coach the team on agile practices, remove obstacles that block progress, and protect the team from external interruptions. The Scrum Master is not a project manager — they do not assign work or manage individual performance. They facilitate the team's self-organisation. Effective Scrum Masters often coach 2-3 teams; less effective ones get pulled into doing-the-work mode and lose the facilitation focus.

The Development Team builds the software. In Scrum, the team is cross-functional (includes all skills needed to deliver — engineering, design, testing, dev-ops as appropriate), self-organising (decides how to do the work), and small (typically 5-9 people). Larger teams scale by splitting into multiple teams that coordinate at higher levels. The team commits to sprint goals collectively and is collectively accountable for delivery, not just individual contribution to assigned tasks.

The Product Owner role is often underestimated. Effective Product Owners need product strategy capability, customer empathy, ability to say no to features that do not fit the product vision, and skill in negotiating priorities among competing stakeholder demands. Many agile transformations fail because the Product Owner role is assigned to someone without these skills. Investing in Product Owner training and selection is critical for agile success. The team's velocity matters less than whether they are building the right things, which depends on Product Owner judgement.

Benefits of the Agile Approach

Faster delivery of value is the most-cited benefit. Agile teams deliver working software in 2-week increments versus waterfall's months-to-years cycles. Customers see and use software earlier, which produces revenue earlier and feedback earlier. Better adaptation to change handles the reality that requirements evolve during development. Agile teams welcome late-stage changes; waterfall teams resist them as scope creep. The adaptation matters substantially in markets where customer needs shift, regulations change, or competitive pressure forces pivots.

Higher quality through continuous testing and integration catches defects early when they are cheap to fix rather than late when they are expensive. Test-driven development, automated testing, and continuous integration practices keep quality high throughout development. Increased customer satisfaction through frequent delivery and feedback alignment produces software that actually solves customer problems versus software that technically meets specifications but missed the actual need. Transparency through visible work, regular demonstrations, and open process reduces the surprises and miscommunications that plague long projects.

Reduced risk is an underappreciated benefit. Waterfall projects accumulate risk through the long development phase before the customer sees anything — the risk that requirements were wrong, that the design does not work, that the technology choice was poor, all compound until late discovery. Agile distributes risk across many iterations. Bad assumptions get caught early through customer feedback on working software. The early discovery means smaller, cheaper corrections rather than late, expensive rebuilds. Risk management is structural in agile, not bolted on through risk registers.

Challenges and Where Agile Doesn't Fit

Agile requires culture change that some organisations cannot or will not make. Command-and-control management styles conflict with agile's emphasis on self-organising teams. Risk-averse contracting models built around fixed-scope deliverables conflict with agile's iterative discovery approach. Heavy regulatory environments (FDA-regulated medical devices, FAA-regulated aerospace, financial systems with audit requirements) sometimes need substantial documentation that agile minimises. Hybrid approaches combining agile development practices with regulatory documentation overlay can work, but pure agile rarely fits these contexts cleanly.

Estimation of large projects is genuinely hard in agile because the approach intentionally avoids upfront detailed planning. Customers wanting fixed-cost-and-scope contracts struggle with agile's adaptive approach. Agile typically works on time-and-materials or fixed-team-cost contracts where scope adjusts within budget. Geographically distributed teams face coordination challenges that co-located teams do not. Modern collaboration tools have improved this substantially since 2020, but face-to-face conversation remains the most efficient communication mode for many agile interactions. Agile Certification programs help teams develop the skills to handle these challenges, particularly at scale.

The contracting model question matters substantially for consulting firms and software vendors using agile. Traditional fixed-price contracts based on detailed specifications conflict with agile's adaptive approach. Time-and-materials, fixed-team-cost, or capped-cost-with-flexible-scope contracts work better with agile. Customers accustomed to fixed-price contracts sometimes resist the new model because it shifts risk from vendor to customer. Negotiating the right contract structure is often a prerequisite for vendor-customer agile engagements to succeed.

When the Agile Approach Works Best

Agile works best in environments with uncertainty about requirements, evolving customer needs, opportunity for incremental delivery, and small to medium team sizes. Software product development is the classic fit because user needs evolve with the software, customers can use partial functionality, and feedback loops produce better products than upfront specification. Marketing campaigns, internal IT projects, research and development, and consulting engagements often fit agile patterns well. The common thread is uncertainty — agile excels when nobody knows exactly what the right answer is and learning matters more than execution to a fixed plan.

Hardware development has had mixed results with agile. Software iterates cheaply because changes are bits; hardware iterations involve physical fabrication that takes longer and costs more. Hardware-only teams sometimes struggle with pure agile. Hybrid software-plus-hardware products often use agile for the software side and waterfall-like phase gates for hardware milestones. The hybrid acknowledges that different parts of the product have different change economics and the methodology should match.

Measuring Agile Success

Velocity (story points completed per sprint) is the most common agile metric but the most often misused. Velocity should track the team's capacity over time to support planning future sprints, not be used to compare teams or push individual performance. Comparing team velocities is meaningless because story points are team-specific estimates that reflect each team's particular work patterns. Better metrics include cycle time (time from work start to delivery), lead time (time from request to delivery), customer satisfaction scores, defect rates, and team health surveys. The right combination depends on what the organisation actually wants to improve.

Outcome metrics ultimately matter more than process metrics. Did the customer get value? Did the product solve the problem? Did the team learn and improve? Process metrics (story points, velocity, ceremony attendance) measure activity but not outcome. Focusing exclusively on process metrics produces busy teams that may not be delivering customer value. Balanced metrics including both process and outcome measures give a fuller picture of agile effectiveness. The right balance varies by organisation; experimentation reveals what produces useful insight.