Growth strategies for startups: what to do at each stage, from product-market fit to scaling with AI
From Sean Ellis's 40% test to the North Star Metric: a practical four-stage framework to pick the right growth strategy and avoid automating too early.
By Rubén García · October 3, 2026 · 3 min
There's no universal growth strategy. There's the right one for your stage. Many startups fail not for lack of tactics but because they apply scale tactics (ads, automation, sales hires) before the product has proven people need it. This is the four-stage framework I use to decide what to do at each moment.
Stage 1: before product-market fit
Don't scale anything yet. Use Sean Ellis's test: ask active users how they'd feel if your product disappeared tomorrow. If more than 40% answer "very disappointed", you have a solid fit signal. If not, iterate the product, not the marketing.
Source: Sean Ellis, PMF Survey
Stage 2: find your metric
Define a North Star Metric following Amplitude's framework: one metric that reflects the value customers get and predicts future revenue. It must measure valuable usage, not shallow activity.
- Booking platform: completed bookings, not visits.
- Design tool: designs exported or shared.
- B2B SaaS: teams using the key feature every week.
Your take, in one click
What stage is your startup at?
Around the North Star you define input metrics (activation, frequency, depth of use) the team can move directly. We gather benchmarks for each in growth metrics benchmarks.
Stage 3: build the engine
Pick one main loop, viral, content, paid or sales, and optimize it before adding channels. If your product can be tried unaided, consider product-led growth: according to ProductLed (2025), 58% of B2B SaaS companies already have a PLG motion and 91% plan to invest more. We explain it in product-led growth with AI.
40%
Sean Ellis test threshold
58%
B2B SaaS with a PLG motion (ProductLed, 2025)
91%
plan to invest more in PLG (ProductLed, 2025)
Stage 4: scale with AI
Once the engine works, AI multiplies it: more useful content for the search and AI-assistant loop, more parallel experiments and per-segment personalization. Applied too early, it only speeds up the chaos: more ads for a product that doesn't retain is more money lost.
Common mistakes by stage
- Stage 1: spending on ads to "validate" a product users don't come back to.
- Stage 2: picking a vanity metric (signups, visits) as North Star.
- Stage 3: opening five channels at once without mastering any.
- Stage 4: automating processes nobody has documented.
Summary: validate, measure, build a loop and only then automate.
Frequently asked questions
What is product-market fit?
The point where a product meets a real need of a specific market, so users use it, recommend it and would miss it if it disappeared.
What is Sean Ellis's 40% test?
A survey asking users how they'd feel if the product disappeared. If over 40% answer "very disappointed", there's a product-market fit signal.
What is a North Star Metric?
A company's main metric: it reflects the value customers get and predicts revenue, e.g. completed bookings instead of visits.
