How Transistor Found 51 Paying Customers Before Launch
Short answer: Start with the part of your audience that has already demonstrated the problem, invite them individually into paid early access and record why each person buys. Attribute customers to specific relationships or content rather than crediting “the audience”. A pre-launch cohort is useful only if customers activate, remain and would have bought without a public launch discount.
An audience creates access, not demand. Thousands of followers can produce no customers when the product solves a different problem from the one that earned their attention.
The relevant asset is not reach. It is a history of useful contact with people who recognise the problem and trust you enough to try a paid solution.
Transistor's paid early access
Justin Jackson and Jon Buda began working together on Transistor in early 2018. Buda had built a podcast-hosting tool for his employer, while Jackson had years of experience publishing about products and podcasting. They signed partnership documents in February and worked on the company while maintaining other income.
Nine days before the 1 August 2018 public launch, Jackson reported 51 paying early-access customers and US$781 MRR. Transistor's fourth-anniversary account says the first month after launch added 46 paying customers and brought revenue to about US$1,400 monthly. Twelve months after launch, the business reached about US$20,000 in monthly revenue.
Jackson's 2018 review says Transistor reached nearly US$5,000 MRR by December. His later account explains that he invited beta users, wrote documentation, handled support and marketed while Buda developed the product.
The sources do not publish the size of each audience channel, the 51 customers' exact origins or early-access retention. Those gaps make attribution, rather than the headline number, the useful lesson.
My view is that founders should charge an existing audience before public launch. Free beta enthusiasm is cheap. A paid invitation reveals whether trust transfers to this particular problem without exposing an unfinished product to broad traffic.
Use the Existing-Audience Launch Attribution Map
The Existing-Audience Launch Attribution Map connects each early customer to a prior trust source and a buying action.
| Map field | Record | Why it matters | |---|---|---| | Prior relationship | Newsletter, community, podcast, client or referral | Identifies transferable trust | | Problem evidence | Recent attempt, spend or workaround | Separates audience fit from friendliness | | Invitation | Personal message, post or event | Shows the effective route | | Paid commitment | Price, date and promised outcome | Establishes demand | | Activation | Core job completed | Tests product value | | Retention or referral | Continued payment or introduction | Tests quality beyond goodwill |
Record one primary and one assisting source at most. Giving full credit to every touch produces totals larger than the customer base.
Find problem overlap inside the audience
Segment people by behaviour, not follower platform. Identify those who have launched a podcast, asked about hosting, paid for related education, run an audio programme or complained about an existing provider.
Make a list of 30 to 60 qualified people you can contact respectfully. The guide on [getting your first 10 customers](/sales-and-marketing/how-to-get-your-first-10-customers/) explains why a narrow, direct route is more useful than broad awareness at this stage.
Do not export private community data or treat access as consent to pitch. Follow platform rules and applicable marketing and data law. Requirements vary by jurisdiction, so obtain qualified local advice for email, tracking and personal-data use.
Sell a paid early-access outcome
Define what the customer can do now, what is missing, the price, support, data handling and cancellation route. A paid early-access plan should complete a narrow job. “Help us build” is not the customer outcome.
Invite people individually and ask about their current workflow before presenting the product. When the problem fits, show the shortest path to value and ask for payment. Record objections from qualified non-buyers.
Avoid a deep lifetime discount. It attracts price-led buyers and makes later retention difficult to interpret. A modest founding price or standard monthly fee with hands-on onboarding provides cleaner evidence.
Separate audience advantage from product evidence
Compare three groups where possible: people with a prior relationship, their referrals and cold qualified prospects. Measure invitation-to-trial, trial-to-paid, activation, support and 60- or 90-day retention.
If only close followers buy and referrals do not, trust may be doing more work than the product. That does not make the revenue false, but it limits what you can forecast.
Ask customers why they bought and what alternative they would have chosen. Count an answer as problem evidence only when it refers to an actual plan, spend or constraint.
Worked example: CastRail Hosting
CastRail Hosting's founder has a newsletter for independent audio producers. These figures are illustrative.
From 3,800 subscribers, 240 have clicked hosting or migration articles within 90 days. The founder selects 60 who also run an active show. Forty receive individual invitations and 20 are held as a later comparison group.
Twenty-four reply, 16 attend a demonstration and 11 pay £29 for early access. Invitation-to-paid conversion is 11 divided by 40 = 27.5%. Nine publish an episode and eight remain after 60 days.
| Source | Invited or referred | Paid | 60-day retained | |---|---:|---:|---:| | Newsletter relationship | 40 | 11 | 8 | | Customer referrals | 18 | 6 | 5 | | Cold qualified outreach | 40 | 4 | 3 |
The existing-audience cohort is strongest, but referrals also convert and retain. That suggests the offer is not supported solely by personal loyalty.
CastRail spends 33 founder hours on demonstrations and onboarding at £32, costing 33 x £32 = £1,056. First-month revenue from 21 customers is 21 x £29 = £609, so the cohort is not contribution-positive initially.
Support logs reveal repeated migration work. CastRail turns that into a documented import process, reducing onboarding from 94 minutes to 38 minutes. It sets a public-launch gate: 30 retained customers, at least 40% originating beyond the founder's direct audience, and onboarding below 45 minutes. The audience finances evidence, not a claim of effortless acquisition.
Choose a public-launch gate
A public launch brings attention and support load at the same time. Require the core workflow to be reliable, pricing to be understood and early customers to remain through at least one meaningful use cycle.
Set a customer number, activation threshold, support capacity and source-diversity condition. Transistor's 51 paid customers demonstrate pre-launch commitment, but your number should follow the buying value and service burden.
A high-value B2B product may be ready after five committed accounts. A low-price self-serve product may need 50 or more to expose onboarding problems.
Do not manufacture an audience before solving the overlap
Publishing can build trust, but starting a general founder newsletter solely to sell unrelated software is a slow and uncertain acquisition plan. Focus content on the customer's real work and make it useful without a sale.
If you have no audience, borrow distribution ethically through relevant partners, communities, events and direct outreach. The article on [how a new business should market itself](/sales-and-marketing/how-should-a-new-business-market-itself/) can help choose a route appropriate to the decision cycle.
Run the pre-launch in 30 days
In week one, map audience members with recent problem behaviour and prepare a precise paid offer. Invite 20 qualified people in week two and conduct no more than ten demonstrations.
During week three, onboard buyers, measure the first outcome and document repeated support work. In week four, invite a referral cohort and a small cold comparison group. Decide whether to launch publicly from retained payment, source diversity and support capacity. If the audience buys but referrals do not, keep testing privately rather than forecasting broad demand.
Related guides
Frequently asked questions
How large does my audience need to be?
Large enough to identify a useful number of people with the specific problem, which may be only dozens. A 500-person specialist list can outperform 50,000 general followers. Work backwards from the cohort you need. If you want ten paid pilots and expect 20% of qualified invitations to convert, contact 50 genuinely qualified people. That conversion assumption is illustrative and should be replaced after the first batch. Do not buy reach merely to enlarge the denominator. Relevance, trust and current problem behaviour matter more.
Should early-access customers pay the full launch price?
Usually charge the intended price or a modest, clearly defined founding rate. Full price gives clean willingness-to-pay evidence, while a discount can compensate for rough onboarding or limited scope. Avoid lifetime discounts and unclear promises of every future feature. Tell customers what changes at public launch and how cancellation works. Pricing, subscription and consumer rules vary by jurisdiction, so obtain qualified local advice for terms and renewals. The useful signal is continued payment after the novelty and personal access fade.
Can friends count as early customers?
They count as revenue but provide weaker evidence when the relationship influences the purchase. Mark the relationship and compare activation, use and retention with unrelated customers. Ask what they would have done without knowing you. Do not remove them from the data or pretend they are cold demand. A friend who has the problem, pays the normal price, completes the outcome and refers an unrelated buyer can still be useful. Source diversity should increase before a broad launch or hiring decision.
What if the audience likes my content but not the product?
Treat that as evidence that trust did not transfer to the product's problem, timing or offer. Interview qualified non-buyers about their current workflow and spending. Do not ask whether they support the idea. Check whether the audience was built around the same job. You may need a different segment, narrower outcome or acquisition channel. Continuing to publish can still be worthwhile, but audience growth will not repair weak problem overlap. Test a revised paid offer with a small batch before investing in more product.
How do I attribute customers who saw several channels?
Record the relationship that created trust and the invitation that caused action separately. For example, a customer may know you through a podcast but buy after a direct email. Assign one primary source and one assisting source, using the same rule across customers. Ask at purchase and retain tagged links where lawful. Attribution will remain imperfect, so use it for directional decisions rather than invented precision. Data and marketing rules vary by country; collect only what you need and explain its use.
When should I move from individual invitations to a public launch?
Move when paid customers reliably complete the core job, a meaningful share remains through one full use cycle, support is within capacity and some demand comes beyond personal loyalty. Set the gate before launch. The required count depends on price and complexity, but the cohort should expose repeated onboarding failures. Make sure billing, cancellation, backups and customer communication work at a larger volume. A public launch magnifies the system you have. It does not repair weak activation or unclear positioning.
Comments (0)