Home/Learn/How to market a product

Win paying customers beyond your friends

To market a product, test several channels at the same time and read their results each week. Begin with people who know you, check the path to payment and set a cap and stop rule for paid tests. Move effort toward the work that brings customers as the evidence develops.

Updated November 2026

Friends have tried the product, and some early users may be paying. Reaching the next buyer is less familiar. You could write search pages, post on LinkedIn, approach podcasts or try ads and email. Each asks for time, and the channel lists you have read give little help with the order.

We began with side projects and used a coding agent to run our marketing work. That experience showed which decisions came back each week: which channels to test together, what needs fixing before visitors arrive, and where to put the next round of effort.

To market a product, run several channel tests

A small product can test several routes to buyers during the same period. If outreach produces little interest, search work can still be progressing while the outreach is reviewed. Running the attempts one after another would delay that search work until the first channel had been judged.

The reason to test is that the successful channel is uncertain. We count finding a useful channel among several attempts as a good outcome. Allow for attempts that do not work when deciding how much effort to put into the first choice.

The mix depends on the buyer and the product. Founder posts on LinkedIn, podcast appearances, paid ads and outbound email reach people in different circumstances. The question for each test is why that particular buyer would encounter the product there and have a reason to consider it.

A small test needs enough definition to be read later. A search page answers a particular need. An outreach attempt addresses a selected group of potential buyers. A paid test has a limit and a stopping condition. These give the eventual review something more concrete than a record that marketing happened.

The aim is to keep several possible routes open while learning which ones merit more work. It does not require launching every channel available. A founder still needs a set of tests that someone can run, inspect and change. Adding another channel without capacity to read it would leave another unfinished decision.

Choose the initial mix with that capacity in mind. Search pages and outreach produce different kinds of work, and both may depend on changes to the site. Keeping those dependencies visible makes the combined effort manageable: the next task can be a page repair as readily as another campaign.

Ask people you know before writing to strangers

Start the outreach with contacts who fit the buyer you want and already have a reason to read your message. Knowing someone does not make them a customer, but it gives you a relevant place to begin the conversation.

In one recorded attempt, cold LinkedIn messages to 20-30 founders received no replies. The effort then moved to requests for warm introductions. The useful lesson is the change in the route to a conversation. More messages through the same route would have continued the same unproductive attempt.

Cold outreach has a place among the possible tests. It can reach people outside the founder's network, and the company may eventually need that reach. But when messages are inexpensive to send, the recipient has many competing requests for attention. A larger recipient list does not remove that difficulty.

A warm list still needs judgment. Rank the people who know you by how closely they match the buyer the product serves. A contact who faces the relevant problem belongs ahead of someone whose only connection is that they once met the founder. The list should reflect buyer fit as well as familiarity.

Among those contacts may be people who know a suitable buyer. Ask for an introduction tied to the problem the product solves. The founder can then follow the conversation and understand the response in the context of that person's need, before deciding how far to extend the outreach.

The builder can start with those contacts while other channel tests run. The warm-contact work can develop conversations as search pages and other production continue. Later outreach choices can draw on what those conversations reveal about the need, without treating a reply as proof that a person will buy.

Fix the path to paying before sending traffic

Before a campaign sends someone to the product, the route from the first page to payment needs to work. A buyer who cannot take the next step leaves the team with a weak response that may have little to do with the channel.

On one of our own products, the upgrade page had a cancel button and no upgrade button. That was a concrete obstacle to paying. A new post or another message would still have brought interested users to the same incomplete page.

The check needs to follow the intended action through to its destination. A homepage can open correctly while the next page fails to offer the action the visitor expects. Reading only the marketing copy would miss the missing upgrade control. Walking through the route makes the fault visible.

Search has its own entry requirement. In another site review, no page was in Google because the sitemap and canonical tags referred to an old domain. The site needed attention to those references before its search pages could be judged on the buyer interest they attracted.

The missing upgrade control stops an action after arrival. The old domain references concern the earlier question of whether a search engine finds the pages. A channel review needs to account for both sorts of fault. Otherwise the builder may judge a campaign from a route that visitors could not complete.

Checking that route also makes the weekly discussion more specific. If a control is missing, the next task is to repair it. If the route works but the test brings no qualified interest, attention can return to the audience, message or channel. Each finding calls for its own action.

Each week, decide which channel gets more effort

The weekly read should decide where to put the next round of effort across the channels already running. A report describes the attempts; someone has to use it to choose the next work.

That choice can move effort substantially. In a recorded client discussion, work started on LinkedIn and moved toward organic search a week later. Search appeared to offer more potential, and the founder was more interested in pursuing it. The plan changed when the conversation exposed a stronger opportunity.

A founder can keep the weekly decisions in their own hands by reading the numbers and assigning the next tasks. That includes time to understand each report, consider it alongside the other channels and make the changes. The arrangement is workable when that recurring job fits the builder's week.

An agent can run channel tasks and prepare a brief for the founder. That takes on production and reporting while leaving the founder to interpret the findings and choose a change. The useful question is whether the founder has room to make those decisions and direct the next tasks.

People can also choose the experiments, direct the agents and review the findings with the founder each week. They connect the reports to the work that follows and review what will reach buyers. The founder contributes knowledge of the product and helps decide the direction.

When comparing these options, establish who reads the results and who decides to move effort. Then ask how many channels that arrangement can keep running. A useful report must reach someone able to change the following week's work; otherwise a finding can remain in the brief while the same tasks continue.

Do it yourself on a side project

A builder who enjoys marketing a side project can run the tests and read the reports themselves. The time spent learning which channels fit can be part of the appeal, alongside building and improving the product.

The open-source system can run on the builder's own machine with their keys and data. For someone comfortable directing a coding agent, that provides a self-run route to the work described here. The builder remains responsible for deciding which tasks are useful and examining what the agent produces.

A company with clients creates a different demand on the founder's time. We have heard from founders who have plenty of growth ideas but no time to carry them out. Product work and client commitments keep taking the hours those ideas would need.

Look at an ordinary week of serving clients and building the product. Is there time to inspect the channel reports and direct several tests? Occasional interest in marketing may not cover those recurring jobs. The arrangement needs to work during the weeks when product and client work take most of the founder's attention.

Some founder involvement remains useful whichever route runs production. Decisions about the buyer, the product and important public work require their judgment. People who run the research, campaigns and follow-through take on the production tasks, leaving the founder to focus on those decisions.

On a side project, operating the tests yourself can be part of learning how to market the product. At a company serving clients, assign time for that work explicitly. The tests need someone to run them and someone to decide what follows from them, week after week.

Rank your contacts, then start three small tests

  • Rank the warm contacts

    Write a list of contacts who know your work and rank them by their fit with the intended buyer.

  • Walk through the payment path

    Follow the route from the first page to payment, and inspect the search references that point visitors toward it.

  • Choose the channel tests

    Choose three channels to test for this product and decide who will run the attempts and read their responses.

  • Set paid-test limits

    Give each paid attempt a cap and a clear stop rule before it begins, then use that rule during review.

  • Decide the next week's work

    Read the channel results each week and move effort toward the attempts that show a route to paying customers.

Find the channels that bring your first customers

Discuss your product with Tin and the growth experiments that could reach its next buyers. We can talk through where to start among your contacts, which channels to test and how to decide where effort goes each week.

Book a call