How a Solo Founder Should Pick Marketing Channels (and Split the Budget)

Ask a solo founder where their marketing budget goes and you usually get one of two answers. Either "a bit of everything," the budget divided evenly across four channels because that felt balanced, or "the platform I like posting on," the channel they enjoy rather than the one their buyers actually use. Both are how a small budget produces only noise: spread too thin to read on one side, pointed at the wrong audience on the other.

The job is not to be on every channel. As one operator you win by running one or two channels well, the ones that actually reach your buyer and that you can sustain. Getting there is a method: score each candidate channel on fit, model rough cost and payback for the top ones, then weight the budget by intent with a stop-rule on every line. Here is the whole thing, sized for a person, not a team.

Start with the channels your ICP is actually on

Before scoring anything, throw out every channel your ideal customer is not on. A popular channel your buyer does not use is not a candidate, no matter how cheap or trendy it is. TikTok is a fine channel and a terrible one for selling tax software to accountants. So the first cut is simple: which channels could plausibly reach this specific buyer at all? That short list is what you score. Everything else is off the table and named explicitly as off the table, so you do not quietly reconsider it every week.

Score each candidate on four axes

Rate every surviving candidate 1 to 5 on four things. This turns a vague "LinkedIn feels right" into something you can compare.

Funnel-stage match. Does the channel serve the stage you need most right now? Paid search captures decision-stage demand (someone already searching for what you sell); content and organic social build awareness at the top. If your funnel leaks at consideration, an awareness channel does not fix the leak and scores low even if it is cheap.

Business-model fit.A B2B product with a high price per customer justifies LinkedIn's expensive impressions and a long content play, because one customer pays for a lot of clicks. A low-price impulse purchase does not, and leans toward visual paid social. Score against your actual model, not against general popularity.

ICP reach. Is your buyer concentrated on this channel, and can you reach them affordably there? A channel with a huge audience that is only 1 percent your ICP scores worse than a small channel that is almost entirely your ICP. This axis can veto: a low ICP-reach score sinks a channel no matter how well it does on the others. You cannot stage-match or out-cheap your way into an audience that is not there.

Operator skill and capacity. A channel you can run well beats a theoretically better one you cannot sustain. Factor in the hours you actually have and what you are genuinely good at. A high-fit channel nobody maintains converts at zero. Two channels that score otherwise equal go to the one you can run. This is the axis that gates your realistic set, and the one founders most often ignore in favor of the channel that looks best on paper.

Sum the scores, or weight ICP reach higher since it almost always matters most, and you get a ranked shortlist. Do not over-engineer it; the score is a decision aid to overrule your bias toward the fun channel, not a verdict that thinks for you.

The shortlist is an anchor, a test, and your list

A solo-founder shortlist almost always comes out the same shape. One anchor channel: the highest-fit one you can run, which gets the most budget. One test channel: promising but unproven for your business, which gets a small, kill-thresholded experiment. And your owned email list, funded regardless of score, because it is the one channel you actually keep. Everything else is deferred and named, not quietly reconsidered.

That owned-list rule matters more than it looks. Paid channels rent you attention; the day you stop paying, the traffic stops. Your email list is the asset that compounds and that no algorithm change can take from you. For any budget, even a bootstrapped one running on hours instead of dollars, the list gets built first. Paid scales a motion that already works, it does not replace the thing you own.

Model rough CAC and payback before you commit

For each channel you plan to fund, estimate two numbers: the cost to acquire one customer, and how long it takes that customer to pay you back. If you have prior data, ground it: "search ran about $60 per customer last quarter." If you do not, state the break-even cost your unit economics allow and label it a hypothesis to be read by a set date. Month one of a new channel is buying the data, and you should say so out loud.

Then sanity-check every estimate against what a customer is worth. A channel whose realistic cost per customer sits above your customer lifetime value does not get funded to "see if it improves." It gets cut or rethought. A channel that "hits its cost target" when that target is above what the customer pays you is a loss that looks like a win, and it is the kind of loss that quietly drains a small budget for months.

Allocate by weight, with a reserve

Now the money, and the rule is the same as it was for the channels: never start from "divide by N." The anchor channel gets the majority, enough to produce a readable number of conversions (roughly 30 before its cost figure means anything). The test channel gets a real but bounded slice, sized to answer a question and then capped, not a third of the pool just to seem fair. Hold roughly 15 to 25 percent in reserve and feed it to whichever channel clears its bar first.

A concrete shape on a $1,000 month: $650 to the anchor (say paid search, the proven decision-stage channel), $200 to a paid social test, $150 in reserve, with the owned list built on your hours in parallel. Not $333 across three. The even split overfunds the weak channel, starves the strong one of the volume it needs to read, and leaves nothing to pour into the winner once one emerges.

Put a kill-threshold on every line

Every funded channel gets two numbers set before any spend, while your judgment is still clean. The cost line: pause and reallocate if cost per customer runs roughly twice your target. The minimum volume: the number of conversions (about 30, or a set share of the test budget) before that line is allowed to mean anything. Below that volume the number is noise, so you do not kill a channel on three data points. Above it, a channel that has clearly missed its line pauses and its budget moves to the one that is working.

A channel with no kill-line is a channel that quietly drains the budget while you tell yourself it just needs more time. Committing the line up front is what stops a sunk cost from making the decision for you in week three, when you are attached and reluctant to admit it is not converting.

Sequence it, do not deploy it all at once

Do not put 100 percent of the budget live on day one. Stage it: start the anchor, let it produce enough data to read, then release the test and the reserve toward whatever is clearing its threshold. A fully-deployed day-one budget has nothing left to follow the winner with, which means even when you find the channel that works you cannot press the advantage. Sequencing is how a small budget compounds instead of just being spent.

The method, packed

That is the whole method: cut to the channels your ICP is actually on, score each on stage, model, reach, and your own skill, build a shortlist of one anchor plus one test plus your owned list, model rough cost and payback against what a customer is worth, weight the budget instead of splitting it, set a kill-threshold on every line, and stage the spend so the reserve can follow the winner. It works on a thousand dollars or on zero dollars and your hours, and it is yours regardless of what you do next. Channel and budget allocation is one of 31 skills in the Full-Stack Marketer Skillpack. The pack version runs this exact method on autopilot: it scores your candidate channels, refuses the even split, models the CAC and payback, sets the kill-thresholds, and learns your business once so it stops re-asking your budget, your ICP, and your hours every time you decide where the money goes. The thinking in this article is the product. The pack is what it looks like when that thinking runs every time, in minutes, without you driving each step.