The Solo Founder Growth Playbook

01 The Problem Every Founder Ignores

i was spending 2-3 hours a day on social media. every single day. writing tweets, commenting on LinkedIn posts, trying to find the right people to talk to, checking what worked, adjusting.

2-3 hours. on distribution. while the product sat there half-built.

every founder i talked to had the same problem. they all knew marketing mattered. they all felt guilty about not doing it. and they all had the same coping mechanism: post once a week, feel bad about it, promise to do better next Monday.

i tried posting manually. lasted a week. got 5 views. one from my wife.

tried buffer. scheduled ten posts. zero replies. nobody engages with a scheduled post from an account nobody follows.

tried an outreach tool. got my LinkedIn restricted in two weeks.

hired a VA. their comments on technical posts were so generic people stopped replying.

tried an AI tool. every reply started with “Great point!” people could smell it from a mile away.

this playbook is everything i learned building it.

02 Your Actual Options

before we get into how the system works, let's be honest about what the alternatives actually cost. not what they advertise. what they cost once you count your time.

Option A: Do it yourself

free in dollars. 2-3 hours a day in time. that's 60-90 hours a month you're not building product. at even $50/hr founder time, that's $3,000-4,500/month in opportunity cost.

most founders quit at week 2. the few who stick with it burn out by month 3.

Option B: Hire someone

Option C: Galevox

that's the pitch. the rest of this playbook is the honest version of how we got here and what the system actually does.

03 Voice: The Hard Part

this is the part that took the longest. not the infrastructure. not the scheduling. making AI sound like a specific human being.

here's what i tried first: i wrote a style guide. “be casual, use short sentences, don't be salesy.” it was terrible. every post read like LinkedIn AI with a hoodie on.

what actually works is a corpus-based approach:

1. collect your real posts that performed well. 20-30 minimum. the ones where people actually responded.
2. extract a voice fingerprint. the system analyzes sentence structure, vocabulary quirks, topic transitions, how you open and close posts. not what you say, but how you say it.
3. build an anti-pattern list. what you'd never say. no em dashes. no “delve into.” no “Great point!” openings. the list grows every week.
4. inject into every session. before Claude writes anything, it reads your voice profile and anti-patterns. the result sounds like you typed it, not like a prompt generated it.
5. update weekly. as new posts perform well, the corpus grows. the fingerprint evolves with your voice.

the breakthrough was realizing that voice isn't about word choice. it's about rhythm. the pattern of long sentence, short sentence, paragraph break. the topics you gravitate toward. the things you never talk about.

04 How the System Works

the system runs hundreds of sessions per day across X and LinkedIn. here's what it actually does:

the system talks to platforms through real browser sessions, not API-only. this matters because platform APIs are increasingly restricted. the free X API tier is write-only. you can't even search.

05 Finding the Right People

most SMM tools pick topics based on engagement. likes, retweets, impressions. that's the wrong signal.

a tweet about “AI will replace programmers” gets 500 likes. zero of those people will ever buy your product. meanwhile, a reply to some founder complaining about posting consistency gets 3 likes and a DM that turns into a customer.

so we score topics differently:

warm lead conversion at 30% weight is the key insight. the system tracks which topics eventually produce people who come back. a topic that generates one returning visitor per 50 replies is 10x more valuable than one that generates 200 likes per 50 replies. because returning visitors turn into conversations. likes don't.

the query generator runs weekly, creating new search queries for high-scoring topics and retiring low-performers. each query gets a 24-hour cooldown so we don't hammer the same conversation twice.

priority accounts get special treatment. we track 20-50 ICP-aligned founders. when they tweet, we reply early. these aren't influencers. they're potential customers who happen to be active on social media.

06 How Leads Happen

here's how it works in practice. real numbers from running the system on @malakhovdm:

the key word is “inbound.” nobody gets a cold DM. nobody gets a connection request they didn't expect. every warm lead is someone who chose to engage with our content multiple times. the system earned their attention first.

each lead gets tagged with their origin topic and platform, so we know which conversations actually produce pipeline. this data feeds back into topic scoring. the system gets better at finding the right people over time.

07 What's Real and What Isn't

here are the actual numbers. not projections. not “if trends continue.” what happened.

what we changed based on the data:

• shifted from vanity topics to ICP topics. “AI replacing jobs” gets engagement but zero leads. “solo founder posting struggles” gets fewer likes but actual conversations.
• voice fingerprint became feature #1. the first paying customer cared about voice quality above everything else. “will it sound like me?” was the first question.
• added warm lead conversion weighting. topic scorer now optimizes for people who come back, not impressions. reply quality went up, volume stayed the same.
• built the score funnel. free growth score at galevox.com/score shows founders what their X presence looks like and what galevox would do with it.

the compounding effect is real. consistency is the one thing machines are better at than humans. the system shows up every single day, rain or shine, whether i feel like marketing or not. and that consistency is slowly turning into recognition.