In short: Fundraising, community and customers are one system, not three projects, and it runs in a fixed order: name the buyer, write one truth, collect evidence, raise against it, give the community a job, make first customers design partners, distribute, measure, then hand over. Every failure in the series came from doing a right thing at the wrong time.
Key points
- Running a deck, a Discord and a customer list as three projects stalls the round and leaves the community idle.
- Raise against what a non-founder has done with the product, not the narrative.
- Distribute through partners, referrals, creators, owned media and answer engines, with paid media last.
- Each vertical changes who buys, what proof counts, which channels are open and what to measure.
- A self-score under 30 suggests spending on distribution with product truth and evidence unanswered.
One system for fundraising, community and customers. Forty parts, one page.
Most early teams run three projects: a deck for investors, a Discord for the community, a list for customers. Three stories, three calendars, three people who don't talk. The round stalls because investors ask about customers. The community goes quiet because it has nothing to do. The customers don't convert because the story they heard was written for someone else.
It is one system. This is the order it runs in.
The spine
| # | Step | The rule |
|---|---|---|
| 1 | One system | Name the buyer, not the market. One calendar, one owner. |
| 2 | One truth | One product truth; one belief per audience; a claims register for what you can prove. |
| 3 | Evidence | Raise against what a non-founder has done with the product, not the narrative. |
| 4 | The round | A researched list, warm routes, a three-week window, a pipeline with pass reasons. |
| 5 | The job | A community earns its keep by doing work you'd otherwise pay for. Measure the job. |
| 6 | The hundred | Recruit the first hundred by name. Ask for judgment, give them a job, close the loop. |
| 7 | Design partners | The first ten customers: a success test, a named owner, a price after the pilot. |
| 8 | Distribution | Partners, referrals, creators, owned media, answer engines. Paid last. |
| 9 | Measure | Activation, retention, revenue, efficiency, source. Never reach. |
| 10 | Handover | Accounts, records, runbooks and reasons in the company's name. It runs when you stop. |
What changes by market
Each vertical changes four things and leaves the rest alone.
| Who buys | What proof counts | Which channels are open | What to measure | |
|---|---|---|---|---|
| AI | The team that owns the workflow, then the budget holder | Output quality on the buyer's own data; a security pack that passed review | Marketplaces, developer communities, answer engines | Retained paid teams, gross margin, cost per task |
| Crypto | Users with funds at stake, then integrators, allocators and holders | Retained usage without incentives; named, dated audits; a token with a job | Issuers, integrations, curators, ecosystems | Protocol revenue, retained deposits, incentive cost per sticky dollar |
| Fintech | The person trusting you with money, and the compliance owner behind them | Licences, the custodian, loss ratios, funded accounts that stay | Partners who hold the customer, practitioners, comparison sites | Funded accounts, cost per funded account, 90-day retention |
| Media | Readers first, then sponsors, licensees and data buyers | Owned audience, paid conversion, renewal, sponsor results | Owned lists, membership, being the cited source, events | Paid conversion, churn, revenue per subscriber, platform independence |
| Emerging Tech | The cost owner, reached through integrators, OEMs and consortia | Field deployments that passed a success test; risk retired | Integrators, OEMs, platform programmes, standards bodies | Converted pilots, contracted annual value, margin including service |
The order matters more than any tactic
Product truth before the pitch. Evidence before the raise. Design partners before the channel. Direct before partners. Pilots before contracts. Measurement before reporting.
Every failure in these forty parts came from doing a right thing at the wrong time.
Where are you?
Ten questions, scored 0–5. Under 30 and you are probably spending on step eight with steps two and three unanswered. Take the Part 10 self-score.
The series
The spine
- You don't have three problems
- One truth, three audiences
- Raise against evidence, not narrative
- Run the round like a sales process
- Community is a job, not a number
- The first hundred
- Your first ten customers are design partners
- Distribution when the platforms are closed
- Stop reporting reach
- Build it to run without you
AI
- The model is not the product
- The pilot that survives security review
- Developers, marketplaces and the answer engines
- Raising for an AI company
- Seats, usage or outcomes
Crypto
- The user story before the token story
- Token or no token
- Community that survives the airdrop
- Issuers, integrations and curators
- Protocol revenue: what the protocol keeps and what to report
Fintech
- Trust is the product
- Raising in a regulated category
- Distribute through whoever already holds the customer
- The people who vouch for you
- From funded account to lifetime value
Media
- An audience is not a business
- What investors fund in media now
- Membership is the job
- Your first ten sponsors are design partners
- From reach to recurring
Emerging tech
- Selling what nobody has a budget line for
- Raising for long cycles
- First customers in a market that doesn't exist yet
- Integrators, OEMs and channel partners
- Revenue models for long cycles
Situations
- "The story isn't landing"
- "The round has stalled"
- "40,000 members and nobody does anything"
- "We shipped everything and the number went down"
- "The agency spent $50K and I have nothing to point at"
Bring us the commercial problem.
Common questions
Should fundraising, community and customer acquisition be run separately?
No. Run them as one system with one calendar and one owner, because investors ask about customers, communities need a job, and customers need a story written for them.
What order should an early-stage company do go-to-market in?
Product truth before the pitch, evidence before the raise, design partners before the channel, direct before partners, pilots before contracts, and measurement before reporting.
What should a crypto company measure instead of reach?
Protocol revenue, retained deposits and incentive cost per sticky dollar.