Days 1-3 · Sharpen the problem
- 1
Write the one-sentence problem
‘[Who] struggles with [what] because [why existing options fail].’ If you can't fill the blanks, interview first.
- 2
Define the smallest proof
Decide what evidence would change your mind - e.g. 10 people give an email, or 3 pre-pay. No proof point = no validation.
- 3
List 20 target people
Communities, LinkedIn, your own network. Specific roles beat ‘people who like productivity’.
- 1
Days 4-9 · Talk before you build
- 1
Run 10 problem interviews
Ask about their last time hitting the problem - stories, not hypotheticals. Never pitch; listen.
- 2
Score pain with money
‘What do you do about it today?’ If the honest answer is ‘nothing’, the pain is weak regardless of what they say.
- 3
Test willingness to pay
Offer a paid pilot or waitlist with a price on it. Interest is free; commitment costs something.
- 1
Days 10-14 · Fake door, real signal
- 1
Ship a one-page offer
Landing page stating the outcome, price and a single CTA. Use any free page builder or SlashAI's landing copy generator.
- 2
Send 100 targeted visits
Post where your 20 targets already hang out. Track clicks → signups honestly.
- 3
Make the call
Hit your proof point? Build. Close but promising? Iterate the offer. Nothing? Next idea - cheap exit is a win.
- 1
Founder Roadmaps
20 visual paths for the journeys every builder repeats - pick one, follow the phases.
Week 1 · Ruthless scope
- 1
Write the core job
One sentence: user does X, gets Y. Everything that doesn't serve it is v2.
- 2
Pick boring tech
Stack you already know > stack trending this week. Managed auth + managed DB + one framework.
- 3
Sketch the 3 screens
Most MVPs are list, detail, form. Paper first - it's faster than Figma at this stage.
- 1
Weeks 2-3 · Build the spine
- 1
Walk the happy path daily
Every evening, use your own app end-to-end. Broken today means fix tomorrow morning.
- 2
Hardcode what you can
Fake the data, skip the admin panel, email from your own address. Real infrastructure comes after demand.
- 3
Instrument as you go
One analytics event per key action (signup, core action, upgrade) so week 4 isn't guesswork.
- 1
Week 4 · Polish just enough
- 1
Fix only embarrassing flaws
Broken flows and confusing empty states. Not dark mode, not micro-animations.
- 2
Add a feedback loop
Visible ‘talk to us’ link plus an in-app way to report issues - every early user is a co-designer.
- 3
Deploy and tell 10 people
Ship to your interview group first. Their confusion is your launch-prep checklist.
- 1
Weeks 1-2 · Pick two channels, ignore the rest
- 1
Go where they already are
Two communities max - subreddits, Discord servers, industry Slack groups. Read the rules; be a member before being a marketer.
- 2
Lead with usefulness
Answer questions genuinely, mention your product only when it truly solves the asked problem.
- 3
DM your interview list
Your validation contacts already trusted you with their time. Convert them first.
- 1
Weeks 3-4 · Make onboarding foolproof
- 1
Time-to-value under 5 minutes
Users should feel the core benefit in one session. Delete every step between signup and the wow moment.
- 2
Do things that don't scale
Onboard users personally over video call. You'll learn more than any dashboard can teach.
- 3
Watch sessions
Recordings of 5 real users beat 500 survey responses. Note where they hesitate, then fix exactly that.
- 1
Weeks 5-6 · Turn users into a flywheel
- 1
Find your retention floor
Week-2 return rate tells you if you have a product. Below ~20% for weekly-use products, fix retention before growth.
- 2
Ask for one referral each
Happy users introduce you to one peer. That's 2x without a marketing budget.
- 3
Document what worked
Write down which messages and channels produced activated users. That's your growth playbook for the next 1,000.
- 1
Weeks 1-2 · Prompting fluency
- 1
Master one provider playground
Spend hours, not minutes: system prompts, temperature, stop sequences. Feel how each dial changes output.
- 2
Learn token economics
Count tokens, estimate costs, understand context windows. Every architecture decision downstream depends on this.
- 3
Build 5 prompt patterns
Few-shot, chain-of-thought, role prompting, structured JSON output, self-critique. Keep them in a personal library.
- 1
Weeks 3-4 · Retrieval (RAG)
- 1
Chunk a real document set
Take docs you know well. Experiment with chunk sizes until retrieval finds what YOU know is there.
- 2
Build search + generation
Embeddings into a vector store, top-k retrieval, grounded answer with citations.
- 3
Break it deliberately
Feed tricky questions. Learn failure modes: wrong chunks, missing context, confident nonsense.
- 1
Weeks 5-6 · Tools and agents
- 1
Function calling end-to-end
Let the model call 2-3 real functions - weather, calculator, DB lookup. Validate every argument yourself.
- 2
Build one reliable agent loop
Think → act → observe → repeat, with a step cap. Reliability beats autonomy; ship suggest-mode first.
- 3
Add guardrails
Input filtering, output schema checks, spend limits. Assume the model will misbehave eventually.
- 1
Weeks 7-8 · Ship something real
- 1
One feature, real users
Add an AI feature to an existing project. Streaming responses, error states, fallbacks when the API fails.
- 2
Set up evals
20 golden examples, scored automatically. Now you can change prompts without fear.
- 3
Track cost and latency
Log both per request. Optimize the expensive slow path - usually by caching or routing to smaller models.
- 1
Weeks 1-2 · Know your numbers cold
- 1
Build the simple model
Users, growth rate, revenue, burn, runway. One sheet, defensible assumptions, no hockey sticks without reasons.
- 2
Nail the story in 3 sentences
Who it's for, what changes with your product, why now. If a friend can't repeat it, simplify again.
- 3
Collect proof
Retention curves, revenue screenshots, letters of intent, waitlist growth. Evidence per claim, organized in one folder.
- 1
Weeks 3-4 · Materials that get meetings
- 1
Draft the 10-slide deck
Problem → solution → market → traction → business model → team → ask. Use the Pitch Deck generator for structure, then make every number yours.
- 2
Write the one-pager
Many investors read the memo before the deck. Same story, denser, sendable as PDF.
- 3
Rehearse the hard questions
Why you? Why now? What if BigCo ships this? Honest answers build more trust than smooth dodges.
- 1
Weeks 5-6 · Run a tight process
- 1
Build the target list
30 investors who actually fund your stage and sector. Warm intros through founders they backed beat cold emails 10-to-1.
- 2
Batch the meetings
Compress into 2-3 weeks so momentum and urgency work for you, not against you.
- 3
Drive to a clear close
Know your minimum raise, acceptable terms and walk-away line before term sheets arrive.
- 1
T-minus 7 days · Freeze and polish
- 1
Code freeze
Bug fixes only. New features during launch week are how launches break.
- 2
Prepare assets
Screenshots, demo GIF/video under 60s, logo pack, one-line description in 5 lengths (tweet, post, directory, press, bio).
- 3
Line up the first wave
DM 20 supporters asking them to upvote/comment on launch morning. Authentic comments, not copy-paste spam.
- 1
T-minus 1 day · Dry run
- 1
Full dress rehearsal
Fresh browser, incognito, mobile too. Signup, core action, share, payment - click everything.
- 2
Check status and support paths
Status page or pinned note, error monitoring live, inbox ready to answer fast.
- 3
Schedule posts
Draft your launch post and tweets tonight. Launch day should be conversations, not writing.
- 1
Launch day · Be relentlessly present
- 1
Reply to everything fast
Every comment, question and bug report within the hour. Speed signals care; lurkers notice.
- 2
Fix critical bugs immediately
Ship fixes and say so publicly - ‘fixed 20 min after report’ wins trust.
- 3
Log everything
Objections, praise, confusion, feature requests. Tomorrow's roadmap writes itself from today's replies.
- 1
Week 1 · Validate & design
- 1
Validate the idea in 3 days
Post in 3 communities (Reddit, Twitter, IndieHackers). Get 10 people to say 'I'd pay for this'. If <5 interested, pivot the angle.
- 2
Map the data model
List every entity: users, projects, items. Draw the relationships on paper. Decide what's a table and what's a column.
- 3
Pick your stack
Bubble for complex UIs, Glide for mobile-like apps, Softr for Airtable-powered portals. Supabase for auth + database. Stripe for payments.
- 4
Design 3 core screens
Landing page, main dashboard, settings. Use a template and customize - do not start from scratch.
- 1
Week 2 · Build the MVP
- 1
Set up auth + database
Supabase gives you auth, database and storage in 10 minutes. Connect it to your no-code tool with the REST API.
- 2
Build the core workflow
One thing the product does. Not ten things. Ship the single action that solves the main problem.
- 3
Add Stripe payments
Lemon Squeezy or Stripe - one-time or subscription. Gate the core feature behind the paywall.
- 4
Test with 5 real users
Share the link. Watch them use it. Fix the 3 biggest friction points.
- 1
Week 3 · Launch & iterate
- 1
Polish the landing page
Headline, 3 benefits, social proof (even if fake testimonials from testers), pricing, FAQ.
- 2
Launch on Product Hunt + Twitter
Prepare your PH listing 3 days early. Schedule tweets. Ask friends to support on launch day.
- 3
Collect feedback religiously
Add a feedback widget or a simple Typeform. Every response is gold for the next iteration.
- 4
Ship 2 improvements in week 1 post-launch
Show you're listening. Public changelog or tweets about what you fixed.
- 1
Days 1-3 · Setup
- 1
Pick one service
Web design, writing, video editing - pick ONE. Specialists get hired faster than generalists.
- 2
Build a 3-project portfolio
If you have no client work, create 3 fictional projects. Show the problem, process and result.
- 3
Set your rate
Research competitors on Upwork/Fiverr. Price 20% below market to start. You'll raise after 3 testimonials.
- 1
Days 4-7 · Outreach
- 1
Find 30 potential clients
Small businesses on Instagram, local restaurants, startups on Twitter. Anyone who needs your skill but isn't great at it.
- 2
Write 5 personalized DMs per day
Not 'Hi, I'm a freelancer'. Instead: 'I noticed your website loads slowly - here's how I'd fix it'.
- 3
Follow up once after 3 days
Most people are busy, not uninterested. 'Just bumping this - happy to help whenever.' One follow-up, then move on.
- 1
Days 8-10 · Close
- 1
Offer a small paid test project
$50-200 for a defined task. Low risk for them, proof for you. Upsell from there.
- 2
Send a simple proposal
What you'll do, timeline, price, payment terms. Use the /Proposal command in SlashAI.
- 3
Deliver early, ask for a testimonial
Finish before the deadline. Ask: 'Would you recommend me to others?' Get it in writing.
- 1
Month 1 · Define & show up
- 1
Pick your topic
What do people ask you for help with? What do you know better than 90% of people? That's your brand topic.
- 2
Optimize your profiles
Photo, bio, banner - all consistent. Bio formula: 'I help [who] do [what] through [how]'. Add a link.
- 3
Post 3x per week
LinkedIn or Twitter (pick one first). Short posts with a hook, value, and CTA. Do not perfectionism yourself into silence.
- 4
Comment on 10 posts daily
Thoughtful comments on bigger accounts get you seen. This is the fastest growth hack in 2026.
- 1
Month 2 · Build depth
- 1
Write 4 long-form pieces
Blog posts, Twitter threads, or LinkedIn articles. Deep content builds credibility faster than shallow posts.
- 2
Start a newsletter
Beehiiv or Substack - weekly email with one insight. Your newsletter becomes your professional moat.
- 3
Collaborate with 2 creators
Co-write something, appear on a podcast, do a Twitter Space. Cross-pollination grows both audiences.
- 1
Month 3 · Compound & monetize
- 1
Repurpose your best content
One long post → 5 short ones → a newsletter issue → a video script. Maximum output, minimum effort.
- 2
Create a lead magnet
Free template, checklist, or mini-course. Collect emails in exchange for something genuinely useful.
- 3
Pitch yourself for 3 opportunities
Speaking, guest posts, consulting. Your brand is now your proof - leverage it.
- 1
Days 1-10 · Python basics
- 1
Learn variables, loops, functions
FreeCodeCamp Python course (YouTube, 4 hours). Code along, don't just watch.
- 2
Install Jupyter + pandas
pip install jupyter pandas matplotlib. Set up in VS Code. You'll live here for the next 35 days.
- 3
Complete 20 small exercises
LeetCode Easy or HackerRank Python. 2 per day. Builds muscle memory fast.
- 1
Days 11-30 · Data skills
- 1
Pandas: filter, group, merge
Kaggle's Pandas course (free, 5 hours). Then apply it to a real CSV dataset.
- 2
Matplotlib + seaborn charts
Bar, line, scatter, heatmap. Make 10 charts from real data. Focus on labeling clearly.
- 3
SQL basics
SELECT, WHERE, JOIN, GROUP BY. Use SQLBolt (free interactive). Do all 37 exercises.
- 4
Build 3 mini projects
Sales dashboard, weather analysis, survey results. Each one: question → data → insight → chart.
- 1
Days 31-45 · Portfolio
- 1
Kaggle competition or dataset
Pick a beginner competition. Clean data, explore, model, submit. Even a simple approach counts.
- 2
Write up your projects on GitHub
README for each project: what, why, how, result. Clean notebooks with comments.
- 3
Build a simple dashboard
Streamlit or Gradio - one app that loads data and shows 3 charts. Deploy on Streamlit Cloud (free).
- 4
Share your work publicly
LinkedIn post with charts + link to GitHub. Tag data communities. This gets you noticed.
- 1
Days 1-14 · Launch
- 1
Pick your niche
The narrower the better. 'AI tools for teachers' beats 'AI news'. Specific audiences are easier to find and monetize.
- 2
Set up on Beehiiv or Substack
Free tiers are generous. Customize the landing page with a clear value prop and sample issue.
- 3
Write 3 issues before promoting
Publish 3 issues so new subscribers see a backlog. Quality > frequency. One good insight per issue.
- 4
Create a referral program
Beehiiv has built-in referrals. Offer a bonus (template, checklist) at 3 and 10 referrals.
- 1
Days 15-30 · Grow
- 1
Cross-promote in 5 communities
Reddit, Slack groups, Discord servers, Twitter threads. Share your best issue, not 'subscribe to my newsletter'.
- 2
Guest post on 2 platforms
Write for someone else's newsletter or blog. End with a CTA to yours. This is the #1 growth channel.
- 3
Consistent schedule
Same day, same time, every week. Your audience should know when to expect it.
- 4
Add a lead magnet
Free PDF, template, or resource list. Gate it behind subscription. This converts casual readers into subscribers.
- 1
Days 31-60 · Monetize
- 1
Reach 500 subscribers
This is the minimum for sponsorship interest. Focus on quality subscribers, not vanity numbers.
- 2
Apply to ad networks
Beehiiv Ad Network, Swapstack, Paved. Or cold-email brands directly with your audience stats.
- 3
Create a paid tier
One premium article per month or exclusive content. $5-10/month. Even 10 paying readers is validation.
- 4
Track your metrics
Open rate, CTR, growth rate, revenue per subscriber. These numbers help you pitch sponsors confidently.
- 1
Days 1-30 · Skill up
- 1
Pick one stack and go deep
React+Node, Python+Django, or Go+Kubernetes. breadth is the enemy. Be dangerous in one thing.
- 2
Build 3 portfolio projects
One full-stack app, one API, one open-source contribution. GitHub with clean READMEs is your resume.
- 3
Complete 1 certification
AWS Cloud Practitioner, Google Analytics, or a freeCodeCamp cert. One is enough - don't collect them all.
- 1
Days 31-60 · Apply
- 1
Optimize LinkedIn + resume
Headline: role you want. Summary: what you've built. Links: GitHub, portfolio, live projects.
- 2
Apply to 5 jobs per day
Quality applications only. Tailor your cover letter using the /ColdEmail command. Track everything in a spreadsheet.
- 3
Network intentionally
Comment on hiring managers' posts. DM engineers at target companies. Ask for a 15-min chat, not a job.
- 4
Practice coding challenges
LeetCode medium, 2 per day. Focus on arrays, strings, trees, graphs. Time yourself at 30 min.
- 1
Days 61-90 · Interview
- 1
Mock interviews weekly
Pramp (free), or practice with a friend. Behavioral + technical. Record yourself and review.
- 2
Prepare your story
STAR method for every project. 'Situation, Task, Action, Result' - practice until it's natural.
- 3
Follow up after every interview
Thank-you email within 24 hours. Reference something specific from the conversation.
- 4
Negotiate your offer
Always negotiate. Use levels.fyi for benchmarks. Even 5% more compounds over your career.
- 1
Week 1-2 · Find the idea
- 1
Solve your own problem
The best micro-SaaS ideas come from personal frustration. What tool do you wish existed?
- 2
Validate with 20 conversations
DM people in your target audience. 'Would you pay $X/month for this?' Get real commitments, not polite nods.
- 3
Scope ruthlessly
One feature that solves one problem for one audience. You can add more later. Ship the smallest thing possible.
- 1
Week 3-8 · Build
- 1
Choose your stack
Next.js + Supabase + Vercel + Stripe. Or Laravel + SQLite. Speed of shipping > architectural elegance.
- 2
Build the core loop
Sign up → use the feature → get value → pay. This is the entire product. Perfect this first.
- 3
Add authentication + payments
Clerk or Supabase Auth for login. Stripe or Lemon Squeezy for payments. Gate features behind the paywall.
- 4
Ship to beta testers
Give 5-10 people free access. Watch them use it. Fix what confuses them. Listen more than you code.
- 1
Week 9-12 · Launch & grow
- 1
Launch publicly
Product Hunt, Hacker News, Twitter, Reddit. Launch in 3 places minimum on the same day.
- 2
SEO landing pages
One page per use case. 'Best [tool] for [audience]' - this compounds over months.
- 3
Reach first $1k MRR
At $29/month you need ~35 customers. At $9/month you need ~112. Price for value, not cheapness.
- 4
Automate support
FAQ page, chatbot, email templates. Your time should go to building, not answering the same 5 questions.
- 1
Week 1 · Pick your product
- 1
Choose a format
Templates (Notion, Figma, spreadsheets) sell fastest. eBooks take longer but command higher prices. Courses take the longest but scale infinitely.
- 2
Research what's selling
Gumroad Discover, Etsy digital, Product Hunt. What are the top 10 sellers in your niche?
- 3
Scope the first product
A Notion template pack, a Figma UI kit, a 30-page guide. Small enough to finish in a week, valuable enough to charge for.
- 1
Week 2 · Create
- 1
Build the product
Set a 5-day deadline. Perfectionism kills more products than competition. Ship good, not perfect.
- 2
Create a sales page
Gumroad or Lemon Squeezy handles this. Include: what it is, who it's for, preview images, testimonials, price.
- 3
Set up payment + delivery
Instant digital delivery. Gumroad, Lemon Squeezy, or Payhip - they handle everything.
- 1
Week 3-4 · Sell
- 1
Launch with a discount
20% off for the first 50 buyers. Creates urgency and gets your first reviews.
- 2
Post about it everywhere
Twitter thread, LinkedIn post, Reddit communities, IndieHackers. Show the creation process.
- 3
Collect testimonials
DM your first 10 buyers: 'How did this help you?' Screenshot their replies for social proof.
- 4
Iterate based on feedback
Version 2 should include the top 3 requests. Announce updates publicly.
- 1
Month 1 · Foundation
- 1
Audit your SEO basics
Google Search Console setup, sitemap, meta titles on every page. These free basics matter more than people think.
- 2
Map 50 keywords
Use Ubersuggest free tier or Google autocomplete. Find questions your audience is Googling. Long-tail = less competition.
- 3
Create a content calendar
4 pieces per week: 1 blog post, 2 social posts, 1 newsletter. Batch create to stay ahead.
- 4
Write your 10 cornerstone articles
The definitive guides for your niche. 2000+ words, original data or examples, genuinely useful.
- 1
Month 2 · Distribution
- 1
Repurpose every piece
Blog → Twitter thread → LinkedIn post → Newsletter section → Short video. 5x the reach from 1x the work.
- 2
Guest post on 3 sites
Medium, Dev.to, Hashnode, or niche blogs. Include a link back to your product.
- 3
Engage in communities daily
Reddit, Twitter, Discord. Answer questions genuinely. Link your content when it's relevant, never spam.
- 1
Month 3 · Optimize
- 1
Analyze what's working
Google Analytics shows which posts drive traffic. Double down on what converts.
- 2
Build email capture
ConvertKit or Beehiiv. Every blog post should have a CTA to subscribe. Email converts 10x better than social.
- 3
Create lead magnets
Templates, checklists, calculators. Gate behind email. This builds your owned audience.
- 4
Set up automated sequences
Welcome email, 5-email nurture, product pitch. Set it once, let it run.
- 1
Day 1-2 · Idea & design
- 1
Find a specific AI use case
Legal docs for landlords, meal plans for diabetics, SEO audits for Shopify stores. Niche wins.
- 2
Design the UI
One input, one output, one page. Keep it dead simple. Use v0.dev for quick prototyping.
- 3
Set up the project
Next.js on Vercel. Supabase for auth + storage. One API route for the AI call.
- 1
Day 3-5 · Build
- 1
Implement the AI prompt
Write a system prompt that gets great results. Test with 10 different inputs before shipping.
- 2
Add auth + rate limiting
Clerk or Supabase Auth. Track usage per user. Free tier: 5 uses/day. Paid: unlimited.
- 3
Build the frontend
Input form → loading state → output display → copy/download. Add Stripe for paid tier.
- 4
Add error handling
API errors, rate limits, empty responses. Graceful degradation, not crashes.
- 1
Day 6-7 · Launch
- 1
Write a landing page
Problem → solution → demo → pricing → CTA. Use the /LandingCopy command from SlashAI.
- 2
Launch on Twitter + Product Hunt
Show the build process. 'I built this in 7 days' threads get massive engagement.
- 3
Iterate based on first users
The first 20 users will show you what's broken and what's missing. Ship fixes fast.
- 1
Week 1 · Research
- 1
List 10 direct competitors
Google your product idea + 'alternatives'. Check G2, Capterra, Product Hunt. Include both big players and small startups.
- 2
Map competitor features
Spreadsheet: rows = features, columns = competitors. Color-code: strong/weak/missing.
- 3
Read their reviews
G2 reviews, App Store reviews, Reddit threads. The complaints are your opportunities.
- 4
Analyze pricing
Price points, tiers, free trials, freemium. What's the market willing to pay?
- 1
Week 2 · Position
- 1
Find your positioning
You can be cheaper, simpler, more niche, better designed, or more honest. Pick ONE angle.
- 2
Create a competitive matrix
Visual chart showing you vs competitors on 2 key axes. Use this on your landing page.
- 3
Write your differentiation statement
'Unlike [competitor], we [unique advantage] for [specific audience]' - one sentence, crystal clear.
- 4
Test with 5 potential users
Show them your positioning + 2 competitors. Ask which they'd choose and why.
- 1
Week 1 · Foundations
- 1
Learn the core techniques
Zero-shot, few-shot, chain-of-thought, role prompting. Anthropic's guide (free) is the best starting point.
- 2
Master system prompts
Write 5 system prompts for different personas. Test how changing tone and constraints changes output.
- 3
Understand context windows
How much fits in Claude vs GPT-4o vs Gemini. Token counting matters for complex prompts.
- 1
Week 2-3 · Intermediate
- 1
Structured output prompts
JSON, markdown, tables, code. Practice forcing specific formats consistently across 20 tries.
- 2
Chain prompting
Break complex tasks into 3+ sequential prompts. Each step feeds the next. Build a 5-step research workflow.
- 3
Build a prompt library
Organize by category. Include: prompt, model, version, success rate, notes. This becomes your professional asset.
- 1
Week 4 · Advanced
- 1
Meta-prompting
Write prompts that evaluate and improve other prompts. Use one LLM to grade another's output.
- 2
Multi-modal prompting
Images + text in Claude/Gemini. Chart analysis, document parsing, visual Q&A.
- 3
Create 3 production prompts
Real-world use case with production-quality output. Document in a portfolio. Share publicly.
- 4
Write a blog post or thread
Teach what you learned. Teaching cements knowledge and builds your reputation.
- 1
Week 1-2 · Foundation
- 1
Clean up your GitHub
Pin 6 best repos. Write READMEs for each: what it does, how to run it, what you learned. Green squares matter less than clean code.
- 2
Build a personal website
Simple Next.js or Astro site. Hero, projects, about, contact. Deploy on Vercel. It's your digital business card.
- 3
Write a README.md for yourself
Who you are, what you know, what you're looking for. Put it at the top of your GitHub.
- 1
Week 3-4 · Projects
- 1
Build a flagship project
Full-stack, production-ready, with tests and CI/CD. Something you can talk about for 30 minutes in an interview.
- 2
Contribute to open source
Fix a bug, add a feature, improve docs. Even small PRs show you can work in a team codebase.
- 3
Write technical blog posts
2 posts about what you built and why. Dev.to or Hashnode. Shows communication skills.
- 1
Week 5-6 · Visibility
- 1
Optimize LinkedIn
Headline: 'Software Engineer | React, Node.js, PostgreSQL'. Featured section: your best projects.
- 2
Share your work on Twitter
Screenshot your dashboard, tweet about a bug you solved. Developers hire people they see building.
- 3
Apply to 50 jobs
Tailored applications only. Use the /Resume command, then /CoverLetter. Track in a spreadsheet.
- 4
Practice system design
Read 'System Design Interview' by Alex Xu. Practice whiteboarding. This separates junior from senior.
- 1
Month 1 · Foundation
- 1
Pick a narrow niche
'Build in public tutorials' beats 'tech videos'. The narrower, the faster you grow. You can broaden later.
- 2
Set up your channel
Channel art, profile pic, About section, links. Consistent branding across YouTube, Twitter, and newsletter.
- 3
Plan 12 videos (3 months)
4 per month: 2 searchable (tutorials), 1 trending (news/reaction), 1 personality (story/vlog).
- 4
Film your first 4 videos
Screen recordings are easiest to start. OBS + your voice is enough. Don't wait for a fancy setup.
- 1
Month 2-3 · Optimize
- 1
Master YouTube SEO
TubeBuddy or vidIQ (free tier). Keyword research in title, description, tags. Searchable content is the growth engine.
- 2
Improve thumbnails
CTR > everything. High contrast, 3 words max, readable at phone size. Study what works in your niche.
- 3
Study retention graphs
YouTube Studio shows where viewers drop off. Fix the boring parts. Front-load value in the first 30 seconds.
- 4
Post consistently
Same day, same time, every week. The algorithm rewards consistency. Don't disappear for 2 weeks.
- 1
Month 4-6 · Scale
- 1
Collaborate with similar channels
2 channels at your level = mutual growth. Joint videos, shoutouts, podcast appearances.
- 2
Build an email list
Link in every video description. Your email list is algorithm-proof. Beehiiv or ConvertKit (free tier).
- 3
Repurpose content
YouTube → Shorts → TikTok → Twitter clips → Newsletter. 1 video = 5 pieces of content.
- 4
Analyze what works
YouTube Analytics: best-performing topics, audience demographics, watch time. Double down on winners.
- 1