Back

Back

Growth

Agile, Lean, Waterfall, Shape Up: How to Actually Choose a Product Framework

A practical guide to product development frameworks for founders: agile, scrum, kanban, lean, waterfall, and shape up, when each fits, and what small teams should actually use.

Teamwork in a modern office at night, with laptops, sticky notes, and a city view. A mix of focus, collaboration, and a casual atmosphere.

Product   19 July 2026 · 6 min read

Ask five builders which product methodology to use and you get a religious war. Ask what problem each methodology was invented to solve and you get something useful.

Here is the honest one-liner: frameworks do not build products, feedback loops do. Every methodology is just a different way of deciding what to build next and finding out if you were right. Choose by the loop you need, not the label.

What each one is actually for

Waterfall is for work where change is expensive and requirements are truly known: hardware, regulated builds, big integrations. Plan fully, execute the plan. Its weakness is the same as its strength: it assumes you were right at the start. For a startup discovering its market, that assumption is usually false.

Agile (the umbrella) was a reaction to exactly that: ship in small increments, learn, adjust. The principle matters more than any ceremony. If change is cheap in your product, exploit it.

Scrum is agile with structure: fixed sprints, defined roles, regular rituals. It shines when a team needs coordination and rhythm. For a team of one or two, most of its ceremony is overhead dressed as discipline.

Kanban is the lightest loop: a board of work, strict limits on how much is in progress, pull the next thing when you finish. No sprints, no ceremonies. Its power is focus: work-in-progress limits are the single most underrated idea in product development.

Lean startup is not a delivery method, it is a truth-finding method: build the smallest thing that tests a belief, measure a real signal, decide to persevere or pivot. It answers "should we build this at all", which every other framework quietly assumes.

Shape Up bets on fixed time and variable scope: six-week cycles, a shaped pitch, a hard stop. Its gift to small teams is the appetite question: not "how long will this take" but "how much time is this idea worth". Even if you adopt nothing else from it, adopt that.

The choice, by situation

Pre-product-market fit, alone or with a cofounder: lean startup for deciding what to build, kanban for building it. Three columns, a work-in-progress limit of one or two, and small bets you can measure. That is the whole stack.

A small team shipping a live product: kanban plus a weekly rhythm, or light Scrum if coordination is genuinely the bottleneck. Add Shape Up appetites to stop projects sprawling.

Anything where mistakes are irreversible: borrow from waterfall. Plan that part properly. Frameworks mix, and the skill is knowing which part of your work is discovery and which part is execution.

The trap to avoid

Process as procrastination. Founders adopt heavyweight ceremonies because process feels like progress and rituals feel like a real company. Every hour of ceremony is an hour not spent shipping or talking to users. At the early stage your entire methodology can fit on one line: small bets, real feedback, honest decisions, repeat.

Frequently asked questions

Which product methodology is best for a solo founder?

Lean startup thinking for what to build, kanban for how to build it. A simple board, tight work-in-progress limits, and every feature framed as a test. Scrum's ceremonies add little when there is no team to coordinate.

Is waterfall ever right for a startup?

For irreversible, well-understood chunks of work, yes: a payment integration, a compliance requirement, a migration. Plan those properly. Discovery work should stay iterative.

What is Shape Up in one paragraph?

Basecamp's method: shape an idea into a pitch, give it a fixed appetite (often six weeks), build with variable scope, and stop when time is up. Its core insight for founders is deciding how much time an idea deserves before starting, not after.

How do I know if my process is working?

One test: how quickly does a real user's feedback change what you ship? If the loop from signal to shipped change is days, your process works, whatever it is called. If it is months, no framework label will save it.

Related product: FirstFlight

From the team behind FirstFlight, the AI launch strategist that takes founders from idea to a tracked 90-day launch plan.

Explore FirstFlight