TL;DR
- Running eight half-experiments at once feels like hedging. It guarantees that none of them produces a signal you can read.
- Pick one channel, the one your first ten customers already use, and commit for three months minimum, even if it turns out to be the wrong one.
- Write the kill number down before you start, one metric that maps to revenue. No judging the channel before the window closes, no running it without the number.
- Prove the channel by hand before you automate. Clay and Apollo scale a motion that already works, they cannot create one.
- Placeholder profiles everywhere, creative energy in exactly one place. Add channel two only after channel one has a baseline.
Everything is job #1
The job title says marketer, singular.
In practice you own strategy and execution across social, content, the website, email, the CRM, landing pages, and whatever the CEO read about this morning. Most people in this seat are self-taught, report to a founder who has never done marketing, and run a little LinkedIn, a little cold email, a little content, a little ads.
I have been this person. At EuroVPS, my first marketing job, I owned all of it alone and taught myself by doing a ton of dumb shit with AdWords and a blog nobody read.
Breadth feels like hedging your bets.
The honest name for that strategy is spray and pray. Velocity Engine, an agency that builds marketing systems for teams of one, describes it as running 15 things at 60% instead of 3 things at 100%.
The feeling that everything you do is random is accurate, and the cause is rarely missing knowledge. You know the options. You have no way to pick between them.
Nobody jumps the whole dial in a week. The first notch is a list of what you refuse to run, written down.
Every guide written for you splits the problem in half. The one-person-marketing-team articles are about burnout and systems, and never mention where customers come from. The first-100-customers articles hand a founder eight tactics, and never acknowledge that the reader is a team of one.
This post is the missing middle, one motion for the person doing both jobs, what to run and what to refuse to touch until you have customers.
"Where do I start" is the wrong question
A longer list of tactics will make the paralysis worse, because the paralysis was never caused by a shortage of options.
Watch what a good mentor does when a solo marketer brings this question to a call. They refuse to answer it. They keep narrowing, one channel, one persona, one scenario, until the question becomes small enough to have an answer.
That narrowing is the entire strategy.
With no team and no historical data, focus is the only way to learn anything at all.
So replace the question. Stop asking where to start, and ask what you refuse to do until customer #10.
Then write the refusals down. They are decisions too, and they are the ones that protect you at midnight when a new channel starts whispering.
Spreading out is how you learn nothing
Here is the math under the reframe. You have roughly a two-to-three-month window to make a go-or-no-go call on anything, before the runway conversation or the board update or your own morale forces one. A channel produces a readable signal only if it gets enough volume inside that window.
Split your attention and budget five ways and every one of the five moves too slowly to tell you anything. You have not de-risked the plan. You have run five experiments that are all underpowered.
one channel deep, five run shallow
Signal you can read
Same three months on both lines. Even a wrong channel crosses the dashed line and hands you a clean no. Five shallow ones never tell you anything.
Paid channels make this literal. Meta's ad system needs roughly 50 conversions a week per ad set just to exit its learning phase and start optimizing. Below that it is guessing, and so are you.
The floors hold across platforms, $250 to $500 a day of spend before a second paid channel makes sense, and LinkedIn ads rarely work standalone under about $1,000 a month because the clicks cost too much.
Split a $3,000 budget across five channels and none of them ever clears its floor. The platform never learns. Neither do you.
This is why the rule that sounds wrong is right. One wrong channel, run to depth, beats five channels run shallow, because the wrong one at least hands you a clean no and a reason to move.
CB Insights' postmortem research puts "no market need" at the top of the startup killers at 42%, and finding your first customers is the market-need test, run live. What you need from these months is a readable answer, in either direction. Coverage does not produce one.
Eight channels at 12% attention each is eight experiments you cannot read. One channel at 100% is the first one you can.
Pick the channel your first customers already use
Choosing does not require a scoring spreadsheet. Work backwards from money.
Revenue comes from customers, customers come from some pool of attention, and your job is to find the one pool you can reach, afford, and measure. There is no button that makes revenue go up, so walk the chain until you hit an input you control, then push only that one.
Ruling channels out is the half of the decision almost nobody does on purpose. If your buyers find solutions through Google searches and AI assistants, social is close to zero-value for you, and you get to delete it from the list without guilt.
A channel ruled out on paper stops nagging you. A channel merely neglected keeps generating shame in the background.
The test for the pick fits in one line. Go where your first ten customers already spend attention and money, and ignore what would look impressive in a board update.
Write the kill number down first
Commitment without measurement is stubbornness, and measurement without commitment is channel-hopping. You need both, welded together.
The window is three months minimum, twelve at the outside. Before you send the first message or spend the first dollar, write down the exact result that will prove or kill the channel, qualified replies, sign-ups, booked calls, whichever single metric maps to revenue at your stage.
The deal has two clauses. You may not judge the channel before the window closes, and you may not run it without a number that ends the argument.
The parts that end the argument
Channel: cold email to seed-stage B2B founders1
Window: 12 weeks, ends the last Friday of the quarter2
Kill number: 15 qualified replies a month by week 123
Ruled out until then: LinkedIn posts, the blog, paid ads4
Verdict meeting: one hour, last Friday. Double down or kill5
The channel, named narrowly
Narrow enough that one person can run it by hand, aimed at a named buyer. "More outreach" is a mood, this is a channel.
A window with an end date
A date you can point to. The judging is scheduled, which is what stops it from happening early, in a weak moment.
The kill number
One metric that maps to revenue, with the threshold chosen while you were still calm.
The refusals
Written down, so the other channels stop auditioning for your attention every morning.
The verdict meeting
Double down or kill, no third option. Either verdict pays you in information, which is the whole point of the quarter.
The doc takes ten minutes to write. It is the difference between a system and three months of vibes.
The window exists to kill the launch-high loop, the emotional spike at the start of a new channel, the crash when reality shows up, the jump to a fresh channel to feel the spike again.
Placeholder profiles stay live everywhere so the brand looks alive, and they get zero creative energy until the primary channel throws a signal.
Tracking can be one free-text box at signup asking where they heard about you. PostHog calls it the highest-signal piece of marketing data you will collect, and it costs nothing. Do not build an attribution model to avoid doing outreach.
How-to guide
The one-channel system
The whole post as five steps. Mentors teach this sequence as channel stacking, and it is the difference between a system and a buffet.
Pick one channel
Work backwards from revenue to the one pool of attention your first ten customers already sit in. Write down the channels you are ruling out, next to the reasons.
Write the kill number
One metric that maps to revenue, and the threshold that proves or kills the channel. This happens before the first message goes out, not after.
Run it by hand for three months
You do the outreach, you write the posts, you answer every reply. Placeholder profiles everywhere else, zero energy spent on them.
Read the signal and decide
When the window closes, the number decides. Double down or kill it. Either outcome is a win, because now you know something.
Stack the second channel
Only after channel one has a baseline do you add another, and channel one keeps running while you do. Stacking, never scattering.
A window with checkpoints is easier to hold than a window run on willpower.
This is what that looks like in my own GrowthMentor dashboard, one call to pressure-test the pick before anything spends, one on the kill number and the tracking, one mid-window teardown of the message itself. Membership is unlimited calls, so the third checkpoint costs exactly what the first one did, nothing.
By hand before you automate
Your first ten customers will come from the least scalable motion available, direct messages, calls, hunting down the person whose hair is on fire today.
Automation multiplies whatever you feed it. Feed it a message that has never converted by hand and you get the same silence, delivered faster, to more people.
There is a paradox here worth naming, because someone will use it against you in a planning meeting. Gartner found that 61% of B2B buyers prefer a rep-free buying experience, a number that climbed to 67% by early 2026.
The market as a whole is going self-serve, and your first ten customers will still come from conversations. Both are true, they just live at different scales. Automate at customer 100. Customer #1 is a human who has to be found and spoken to.
So the gate is simple. Before you open Clay or Apollo or an ads manager, one manual version of the motion has to clear your kill number.
Tools scale a working motion. They have never once created one.
Pressure-test your channel pick before you commit the quarter
Book a 1:1 call with a mentor who has run acquisition as a team of one. Bring your channel shortlist and your kill number, leave with both sharper. One membership, unlimited calls.
Guard the one motion
Focus dies of Tuesdays more often than of bad strategy, the sales deck someone needs by Friday, the event booth, the CEO's LinkedIn, the website refresh.
Social, content, website, CRM, email, landing pages, and tech procurement is several full-time roles, and when it all lives in one person, every ad-hoc request is a withdrawal from the only account that produces customers.
Two defenses work. Quantify the cost out loud, "this deck costs us three days of outreach this week," so the trade becomes a decision your founder makes instead of a favor you absorb. And under-promise on everything outside the one motion, so the motion itself can over-deliver.
And name the real problem, you are alone in the role. No peer to sanity-check the channel pick, no senior marketer to tell you the month-two numbers are normal, nobody to say the plan is sound, keep going.
I started GrowthMentor because of that exact feeling, back when I was hiring freelancers off Upwork partly just to have someone to talk through problems with.
The isolation reads like a personal failing and it is a structural fact of the seat. You are one person assigned a department's workload, which is exactly why the answer is subtraction. And if it is the role itself you are trying to figure out, we wrote a separate guide to being the first marketer in an early-stage startup.
The fix does not have to be a hire.
Raising a hand is a feature where I work. You post the decision you are stuck on, and operators who have run it apply to help. Here is what that looks like for exactly the decision this post is about, three hands up inside a day.
Name the channel and the number, today
Here is the whole post as a next move. Open a doc and write two lines, the one channel you will run for the next three months, and the number that will prove or kill it.
It takes ten minutes, and most people in this seat never do it, because committing to one thing means admitting the other seven were noise.
The hardest part is having nobody to pressure-test those two lines with. That one is fixable.
The mentors below have run acquisition as the only marketer in the building, they have made this exact pick with real budgets on the line, and one call is usually enough to tell you whether you picked the right horse or the comfortable one.
Suggested mentors
Operators who have been the whole marketing department, and will look at your channel pick with cold eyes:
Hannah Parvaz
Founder and award-winning marketer. Early-stage go-to-market from scratch.
Ekaterina Gamsriegler
Head of Growth & Marketing @ Mimo. Growth advisor. Product 50's Top Growth
Kate Busby
Fractional CMO. Social media, content, and SEO across real founder accounts.
Nilay Jayswal
Fractional GTM & Performance Marketing Consultant
Solo marketer, first customers FAQ
Vetted mentors, every one included
Three months, one channel.
Pick it with someone who has run the play solo.
Browse vetted growth and marketing mentors and book a 1:1 call on-demand. Bring your channel shortlist and your kill number. Membership is unlimited calls, every mentor included.
Talk to a mentorKeep reading
More from the GrowthMentor blog
Growth Strategy · Apr 03, 2026
How to Get Your First Customers After Launching a Startup
Foti Panagiotakopoulos
Marketing · Jan 03, 2023
How to navigate being the first marketer in an early-stage startup
Lucy Heskins
Growth Strategy · Apr 26, 2020
Your First 90 Days as a Growth Marketer: What You Need to Know
Daniel Johnson
Growth Strategy · Jun 17, 2022
When Should You Make the First Marketing Hire For Your Startup?
Sarah Wisbey
