Skip to content
← All articles

The 4 Stages of Software Development (and How to Tell Which One You're In Before Launching a New Product)

The 4 stages of software development for a new product — discovery, MVP design, development and launch, operation and scale — with a 5-signal self-diagnosis to tell which stage you're in.

yellow, red, green, and black abstract painting
Photo by Steve A Johnson on Unsplash

The 4 Stages of Software Development (and How to Tell Which One You're In Before Launching a New Product)

The 4 stages of software development for a new product are: (1) Discovery and definition, (2) MVP design and architecture, (3) Development and launch, and (4) Operation, measurement, and scale. Each stage has a distinct goal and a verifiable deliverable: a defined business problem, a designed MVP scope, a product in real users' hands, and finally usage data that decides whether you reinvest, pivot, or stop.

Knowing which stage you're in is a money decision, not a vocabulary exercise. A team that thinks it's in development (stage 3) when it doesn't actually have a defined problem (stage 1) pays for weeks of engineering on a product nobody asked for. This guide defines the 4 stages, gives you a 5-signal self-diagnosis to locate yourself, and shows what each stage delivers — and the mistake that usually sinks it.

How to Tell Which Stage You're In: a 5-Signal Self-Diagnosis

Place yourself in the highest stage whose signal you fully meet — if a signal is only half-true, you're still in the previous stage:

  1. Can you state your business problem in one sentence, with a number in it? (E.g., "manual order entry eats 40 minutes of staff time every day.") If not, you're in stage 1: Discovery and definition.
  2. Do you have a prioritized v1 scope, plus wireframes or a spec reviewed by someone other than you? If not, you're in stage 2: MVP design and architecture.
  3. Is a first version of the product live in the hands of real users — not your team, not your friends? If not, you're in stage 3: Development and launch.
  4. Are you measuring usage with defined metrics (activation, retention, adoption) on a regular cadence? If not, you're just entering stage 4: Operation, measurement, and scale.
  5. Does data — not enthusiasm — drive what you build next, and what you deliberately don't? If the roadmap is wish-list-driven, you're still maturing in stage 4.

The most common mistake isn't being in an early stage — it's believing you're further along than you are. Most teams that claim they're ready to build still haven't met signal 1.

Stage 1: Discovery and Definition

Stage 1 turns an idea into a defined business problem with measurable success criteria. The deliverable is short but demands honesty: a problem statement, the primary user, and the number the product has to move to be worth building.

In Kirana Labs' practice, the most common pattern at this stage is the opposite of what you'd expect: clients arrive with plenty of ideas but no clarity on what to prioritize or how to validate it. The first discovery sessions aren't technical work — they're focus work: identify what will add the most value for the primary users, where the core value of what you're building sits, and what the initial value proposition will be. For typical new products, this discovery takes a couple of sessions — not months.

Signal of progress: you can explain the problem and the success criterion without naming a single feature. Signal you're stuck: every conversation circles back to screens, not to the problem. Common mistakes: building without a defined problem, and accepting every scope idea that comes up (early scope creep), which bloats the MVP before the first line of code exists.

Stage 2: MVP Design and Architecture

Stage 2 defines the first version's scope and how it gets built: wireframes, a functional spec, and decisions on the technology stack, infrastructure, and third parties. The deliverable is a v1 scope a team can build without guessing.

This is where the real cost of the project gets decided, long before the first sprint: the MVP's scope, the number of integrations, and the team's composition move the budget more than any code-level decision. It's also where you choose the vendors the product will rely on from day one — for example, infrastructure on AWS and payments through Stripe are typical stage-2 decisions, because they shape architecture, operating costs, and delivery speed.

To choose a stack on merit (not on hype, and not on whatever the first developer you hire happens to know), see our guide on choosing the right technological stack for your project.

For typical new-product engagements, MVP design — including validations with users or stakeholders — runs about 1 to 2 weeks. Signal of progress: someone outside the project understands the scope from the wireframes alone. Signal you're stuck: every meeting redesigns the scope instead of refining it.

Stage 3: Development and Launch

Stage 3 builds the MVP and puts it in real users' hands. The deliverable is a deployed product, with minimal instrumentation to measure usage, and a first group of users genuinely using it.

What separates a healthy build from one that drags on forever isn't the methodology — it's cadence: short iterations, frequent demos with the product owner, and explicit exit criteria per iteration. A basic CI/CD pipeline (continuous integration and continuous delivery) keeps releases predictable and keeps the "big launch" from becoming a single point of risk. For typical new products, development up to first launch takes around 3 months.

The lessons that save the most pain at this stage are uncomfortable, and we've documented them in maximizing your software's success: lessons from the trenches: scope that grows quietly, launches without exit criteria, and teams that show progress with slides instead of working software.

Signal of progress: every demo shows working product, not slides. Signal you're stuck: you've been "two weeks from launch" for a month.

Stage 4: Operation, Measurement, and Scale

Stage 4 starts on launch day: measure real usage, decide with data, and keep the product healthy as it grows. The deliverable isn't a version 2 — it's an informed decision: reinvest, adjust course, or stop.

This is where products that scale separate from products that fade. The right questions are business questions: do users activate (find the value the first time)? Do they come back without being pushed? Does serving them leave room for margin as you grow? Alongside that, one technical discipline matters most: managing technical debt — the stage-3 shortcuts that, left unpaid, make every new feature slower and more expensive to ship.

When the product grows or needs to migrate platforms, providers, or architecture, the biggest risk isn't technical — it's operational: the user noticing. That's why we wrote how to achieve a seamless software transition.

Signal of progress: roadmap decisions cite usage metrics. Signal you're stuck: the roadmap only adds features, and no usage metric justifies any of them.

What About SDLC? How This Maps to the Classic Life Cycle

The 4 stages in this article are a product framework — where your business is on the path from idea to launched product. The SDLC (Software Development Life Cycle) is the engineering counterpart: the international standard ISO/IEC/IEEE 12207 defines software life-cycle processes, from requirements through maintenance. They don't compete: the SDLC describes how to build software well; these stages describe what needs to be resolved in your business before and while you build.

Frequently Asked Questions

What's the difference between the 4 stages and the SDLC?

The 4 stages (discovery, MVP design, development, operation) organize the business decisions of a new product: which problem, which scope, which launch, which metrics. The SDLC, defined in the ISO/IEC/IEEE 12207 standard, organizes software engineering processes — requirements, design, construction, testing, maintenance. In a real project they work together: each product stage consumes SDLC processes to execute.

How much does it cost to develop a new product?

It depends on three variables that are decided almost entirely in stages 1 and 2: the MVP's scope (how many features make the v1), the number and type of integrations (payments, messaging, existing systems), and the composition of the team building it. That's why a serious budget is never quoted against an idea — it's quoted against a designed scope; any number without that groundwork is a figure without a foundation. The right move: define the MVP (stage 2), then ask for estimates against that scope.

Can I skip discovery if my idea is already clear?

Having the idea clear is exactly what discovery puts to the test. In Kirana Labs' experience, founders typically arrive with many ideas and no clarity on prioritization or validation; discovery takes a couple of sessions, and its output is a prioritized problem with a measurable success criterion. Skipping it doesn't save those sessions — it converts them into weeks of development spent on the wrong scope.

What is an MVP, and why launch one instead of the full product?

An MVP (minimum viable product) is the smallest version of a product that lets you validate your value hypothesis with real users. Eric Ries popularized the concept in Lean Startup: the goal of an MVP isn't a small product — it's validated learning, maximum learning per dollar invested. Launching the full product first inverts the risk: you spend the entire budget before the first evidence that anyone needs it.

Do I need an in-house team or a development partner?

It depends on your stage and your ability to attract talent. An in-house team gives you control and accumulated knowledge, but hiring well takes months and a new product doesn't justify the overhead yet. A development partner gives you immediate capacity and repeated experience on the idea → MVP → scale path, with risk shared across visible milestones. The deciding question is simple: is your competitive edge building software, or running the business that software enables?

From the Stage You're In to the Next One

Kirana Labs guides new products across all 4 stages with product teams in Monterrey, Mexico that work US hours and stay accountable for roadmap milestones and a measurable release cadence — from a couple of discovery sessions to a product in operation. If you want a team to review which stage your idea is in and what it takes to move forward, start with the New Business service, and see the evidence in our case studies.

Start a conversation.

Tell us the number you want to move: a launch date, a cost, a cycle time, a cash position. We'll come back with a baseline, a plan, and what we'd be accountable for.