The demand-maturity ladder: a four-stage method for early-stage growth

Stage 1: no proven demand

Stage 1 means no proven demand. The product exists, or is nearly built, but nobody outside the founder's circle genuinely wants it yet. The bottleneck is not distribution. It is whether real people with this problem exist, and whether the founder has actually talked to them. The job at Stage 1 is to manually recruit and interview target users, not to run launches.

Diagnosis signals

• Zero retained users.
• The founder cannot state a sharp ideal customer profile (ICP) in one sentence.
• Fewer than about 10 real problem conversations with target users who are not friends.
• The founder describes the product as "just built" and asks about launching.

The canonical move

The canonical move is one of the following:

• Run 15 problem interviews with target users who are not friends.
• Run a manual recruitment sprint: identify the 3 to 5 specific places where the exact users already gather (named communities, subreddits, job titles, search terms) and reach out to 25 to 30 individuals who visibly show the problem.
• Do concierge onboarding: manually onboard 3 to 5 users one to one, doing the work for them by hand, to test whether the problem is real enough that they engage.
• Sharpen the one-liner: rewrite the value proposition and test it cold on 10 strangers. They should understand who it is for within five seconds.

The quality bar

A real conversation means the person has the problem, is not a friend, was asked about their current behavior and pain rather than pitched the product, and gave a specific, quotable answer. Five of those count. "I posted in a Slack and three people reacted" does not.

Exit criterion

The founder has had about 15 real conversations, can state the sharp ICP in one sentence, and has at least 3 people who confirmed the problem is painful enough that they would try a solution. Ideally, at least 3 people are already using a manually onboarded version of the product.

A 7-day grade is honest at this stage: whether the founder had 5 real conversations, and what 3 of those people said, is gradable within a week. The slow-feedback problem only appears at Stage 3.

Stage 2: users, no retention signal

Stage 2 means users exist but there is no retention signal. Some people signed up, nobody clearly sticks, and the founder cannot say why anyone stays. The bottleneck is activation and retention. Pouring traffic in now wastes it: it pours water into a leaky bucket. The job at Stage 2 is to talk to every user and every churned user, and to find and measure the activation moment.

Diagnosis signals

• Between 5 and 40 users exist.
• Retention is flat or unmeasured.
• The founder cannot name why anyone stays.
• No paying users, or payment is not sticky.

The canonical move

The canonical move is one of the following:

• Interview every churned user from the last 30 days.
• Define and instrument the activation moment, then measure the share of new users who reach it.
• Hold one-to-one onboarding calls with every new signup this week.
• Run a simple Sean Ellis style survey with existing users: how would they feel if they could no longer use the product?

The quality bar

The founder has spoken with every user who signed up in the last two weeks, can describe the activation moment concretely along with the percentage of new users who reach it, and has contacted churned users and recorded their reasons in their own words.

Exit criterion

The founder can name at least 5 users who genuinely stuck, through repeated use over weeks or a payment that renewed, and can articulate why those users found the product and why they stayed.

Stage 3: faint pull

Stage 3 means faint pull. A retained core exists, and the founder can articulate where sticky users came from and why they stayed. Only now is distribution the real question: the founder has earned the right to distribute.

Diagnosis signals

• A retained core of users exists.
• The founder can tell the why-found and why-stayed story.
• There is a crude but repeatable pattern of acquisition.

The canonical move

The canonical move is to pick one channel, chosen from evidence of where sticky users actually came from and where the ICP already spends attention, and commit to it for 4 to 8 weeks. Not spray-and-pray. The weekly move is one concrete unit of work in that channel, for example:

• Build in public: ship 4 posts this week on the platform where sticky users already are, with a defined structure.
• Founder-led content: write one deep post answering the exact search query the ICP uses.
• Targeted community participation: be genuinely helpful 5 times in two named communities, with no pitching.
• Cold outreach: send 40 personalized messages to one tight ICP segment.

The quality bar

Dual-horizon grading begins at this stage. The 7-day grade checks leading indicators: the committed channel work shipped at the committed volume and quality, posts went out, they earned qualified engagement, outreach got real replies. The lagging grade, revisited at 4 to 8 weeks, checks whether qualified signups or revenue moved.

Exit criterion

One channel reliably produces qualified signups at a predictable cadence.

Stage 4: a repeatable channel

Stage 4 means one channel reliably produces qualified signups. The job is no longer to find a channel. It is to make the working channel dependable, and only then to add a second one.

Diagnosis signals

• One channel reliably produces qualified signups at a predictable cadence.

The canonical move

The canonical move, in order:

• Optimize the working channel: conversion, cost, and cadence.
• Turn the channel into a documented, repeatable weekly process.
• Only after the first channel is systematized, test a second channel.

The quality bar

Dual-horizon grading, as in Stage 3: a 7-day grade on execution quality and a lagging grade at 4 to 8 weeks on qualified signups and revenue.

Exit criterion

The first channel is systematized and a second channel is being deliberately tested. At that point the project has largely outgrown this early-stage method.

How is the stage diagnosed?

The earliest-unmet-stage rule places a founder at the earliest stage whose exit criterion is not met. The diagnostic rules are evaluated in order, and the first match wins.

1. The founder cannot state a sharp one-sentence ICP, or has had fewer than about 10 real problem conversations with non-friend target users, or has zero retained users: Stage 1.
2. Otherwise, the founder cannot name at least 5 genuinely retained users and why they stayed, or retention is unmeasured or flat: Stage 2.
3. Otherwise, the project has a retained core and a why-found and why-stayed story, but no channel reliably producing qualified signups: Stage 3.
4. Otherwise, one channel reliably produces qualified signups: Stage 4.

Vanity volume does not move a project up the ladder. Followers, upvotes, waitlist length, and raw signup counts are not exit criteria; real conversations, retained users, and qualified signups are. A project with 200 signups and zero user conversations is still a Stage 1 project.

If the available evidence is not enough to place the stage confidently, the diagnosis stops and asks for the specific missing facts instead of guessing.

Why is skipping stages not allowed?

The no-skip rule means a project never receives a move from a later stage than the one it is diagnosed in. A Stage-1 project does not get a Stage-3 channel move, no matter how the question is phrased.

It will not hand a Stage-1 project a marketing channel.

If a Stage-1 founder insists on being told which marketing channel to pick, the answer explains why that question is premature, names what must be cleared first, and gives the Stage-1 move anyway. This refusal is the integrity mechanism of the method: every growth blog post will answer the channel question, and almost nothing tells a builder they are not ready to ask it.

What is dual-horizon grading?

Dual-horizon grading scores every move on two clocks: a leading grade after 7 days, and a lagging grade revisited at 4 to 8 weeks. A single 7-day window is the wrong yardstick for real distribution work.

Grading on two horizons removes two failure modes at once: the bias toward shallow tactics that look good within a week, and the opposite failure of writing off slow-burning distribution work too early as inconclusive.

What happens when the evidence says stop?

The hard-truth path is the rule that the method must be able to recommend a reframe, a pivot, or a stop. Honesty is a feature, not a tone.

If a project stays in Stage 1 across several weeks and the accumulated conversation evidence consistently shows no real pain, for example 8 or more honest conversations with nobody confirming that the problem matters, the move changes. It is no longer to do more conversations. It becomes a recommendation to reframe the problem or the ICP, or to pause the project.

A coach that only ever says keep going is worthless. The willingness to say stop is what earns trust and makes the graded verdicts worth believing.

Where is this method used?

The demand-maturity ladder is the methodology behind the Awiser AI Growth Copilot, which runs it as a weekly loop: check-in, stage diagnosis, one move, report-back, graded verdict. Your first moves are free.