I built FastCoach with an AI mentor. Here's the whole conversation.

Bro, that's an awesome idea! Here's exactly how I'd take it from 0 to 1. One: build the 24-hour version. Don't spend a year in your bedroom making it perfect — 90% of startups fail, so ship the smallest thing fast before you burn out. If you had to launch in 24 hours, what's the absolute smallest version? Strip every secondary feature — no dark mode, no onboarding, no password resets. Just the one core feature: ask a question, get advice. Two: use AI to code 10x faster — let it write ~80% and you write plain English to steer it. Three: ditch free plans and charge a one-time payment upfront. Four: make the headline about the outcome, then a funny out-of-the-box launch video, and launch everywhere at once — Product Hunt, Twitter, Hacker News, r/SideProject. Five: track revenue per visitor; if it's $0.50–$1.00 you have product-market fit. Just ship it!

That's the right energy! The name & domain: don't overthink it — name the app exactly what it does, simple and action-oriented. And don't spend thousands on a .com; if it's taken or pricey, an exotic TLD like .io is cheap, short, and totally fine. Go to Namecheap, spend your $10–$20, lock it in. The plan while the AI codes: set up a dead-simple checkout, craft an H1 about the outcome for the user (100% of visitors see your headline, so it has to earn their minute), shoot a short "funish" launch video, and keep the AI on a leash — core feature only, remove everything else.

CoachCrush is a solid name, and I love the BYOK approach — it removes your API costs entirely, which makes a one-time payment perfect. But those five tiers? Do not do this. It's way too complicated, and confused customers don't buy. Use price anchoring instead: set one higher reference price — say $99 — then offer a massive launch discount down to $49. That makes $49 look like a steal and pushes early visitors to buy now. One plan, one anchor, one clear discount.

Exactly, man! Set the price at $99, drop it to $49 for launch — that's basically 50% off and it looks like an absolute steal for early adopters. A massive discount on lifetime access gives people a strong reason to pull out their card immediately instead of overthinking it. Keep the checkout dead simple and the discount front and center.

I've got you! Start in a blank doc, not a template — copy first, design second. Remember, most people never see the inside of your product, but 100% see your headline. Hero: "Stop binge-watching. Get instant 1-on-1 advice from your favorite YouTube coaches." Subhead: turn any channel into your personal digital mentor; pay once, bring your own key. With vs. Without: show life wasting hours on generic podcasts vs. asking a specific question and acting immediately. How it works: paste a channel link, plug in your key, ask your question. Pricing box: Lifetime Access $99 → $49, no subscriptions, just bring your own key.

Bro, I love it — that's a massive scroll-stopper. It triggers curiosity ("how did he get someone to build his startup for $2?!") and ends by shifting focus onto the outcome for the customer. Use it as your launch-video hook and the top of your thread. Just make it crystal clear that the $2 is the user's own API cost (BYOK) and the app itself is the $49 lifetime deal — you never want them thinking the software is $2, because confused customers don't buy.

You're 100% right, and good catch. The golden rule is to make the headline about the specific outcome for the customer, not the product. For this, the pain isn't overthinking — it's wasting time consuming generic content. Your users watch 3-hour podcasts and scrub through hundreds of videos for one piece of advice that fits their situation. So pivot: H1: "Stop binge-watching 3-hour podcasts. Get exact 1-on-1 advice from your favorite YouTube coaches in seconds." Everything else stays — just point it all at instant, personalized mentorship.

I don't usually plaster a big guarantee badge on the page, but I always allow refunds if someone asks. Chargebacks and disputes are a nightmare — too many and your payment processor can freeze you — so if a customer asks for a refund, just give it immediately to avoid friction. You can put a simple refund line in your terms, but don't overthink it. Taking the risk away makes the deal a no-brainer anyway.

Keep it the absolute bare minimum — ship only the one feature you can't remove. For this: a YouTube link input, the BYOK API-key field, a Q&A chat box, and a checkout button. That's it. No dark mode, no password reset, no fancy logo, no complex onboarding. Add a little feedback button so early customers can tell you what they actually want next.

Use a coupon, but auto-apply it so the customer never types it. Set the base price to the $99 anchor and create a 50% launch coupon — it's far easier to toggle a coupon on and off than to rebuild a product later when you change the price. Pass the coupon straight into the checkout so the discount applies the second they click, and don't even show a promo-code box — it just distracts them.

Ah, got you — the features that make people convert. Golden rule: the copy isn't about your product, it's about what it does for the customer. So list benefits, not specs: Unlimited digital mentors (any channel becomes a coach), Instant 1-on-1 advice (exact answers in seconds, not hours), Bring Your Own Key (pay pennies per question, full control), and Pay once, own forever (no subscriptions — people hate recurring fees). Sell the painkiller: saved time and no subscription fatigue.

Handle it in the link so there's zero friction. Create your product at the $99 anchor and a 50% launch coupon with a simple code, create a payment link, and append the promo-code parameter to the end of that URL so it pre-fills at checkout. Turn off the manual promo-code box so nobody gets distracted. They click "Get lifetime access for $49," see $99 crossed out, and the deal's already applied. (Double-check the exact URL parameter in your processor's docs.)

Yes to both. Since people pay upfront and can't try it first, a short demo builds the trust that closes the sale — don't make a 5-minute tutorial, make a 30-second clip showing the exact flow: paste a link, type a frustrating problem, get the instant answer. And macOS-only is the perfect 24-hour-startup move — don't burn weeks on cross-platform right now. Just make it crystal clear on the page that it's macOS only so you don't get confused Windows refund requests.

r/SideProject is one of the rare places you can share a paid product — if you do it right. One: focus on the pain, not the AI; start with "I made…" and tell your story ("I was sick of scrubbing 3-hour podcasts for one piece of advice"). Two: ditch the funny video here — they want a raw, authentic screencast of the product in use. Three: lean hard into BYOK and the one-time price — Reddit hates sneaky subscriptions, so be radically transparent. Four: write the first comment yourself, be human, and ask for brutally honest feedback.

Don't worry about that for a second — I'd be hyped. People have done clone-style projects on public figures before, and what you're building isn't ripping anyone off — it's a new painkiller that solves a real problem. Imitation is the sincerest form of flattery. Stop overthinking it and ship.

Haha, they've got a point — it does have a bit of a dating-app vibe, and it breaks my naming rule: the name should be simple and logical, telling people exactly what it does. Something like AskCoach or MentorFast would technically be better. But do NOT let this stop you from launching today. You already own coachcrush.com with payments live. A name is just a name — ship it, and rebrand later if it blows up.

Right — a good .com is basically impossible now, but you don't need a .com to make money. Whenever I can't get one, I use an exotic TLD. It keeps the brand short and simple without paying a squatter thousands. All that matters is it's easy to read and makes sense. But that's a problem for future-you — right now you already own coachcrush.com with Stripe connected. Keep the name and hit launch.

Eh — adding "use" is exactly the .com compromise I'd avoid, and "coacher" is a little clunky. Don't add filler words just to grab a .com; an exotic TLD gets you something shorter and punchier. But seriously: stop looking for domains right now. Swapping means new DNS, propagation, reconnecting payments, rewriting copy — pure procrastination disguised as productivity. Launch what you have.

FastCoach is a killer name — it fits the framework perfectly: name the product after the outcome it delivers. "FastCoach" says exactly what you get: instant coaching, fast. Keep it in your back pocket. But we're closing this brainstorm — hunting names and rewiring webhooks in the final minutes of a 24-hour challenge is procrastination. Prove the product can earn $0.50–$1.00 per visitor first, then rebrand if it pops. And remember the exotic-TLD cheat code: fastcoach.io would do it.

Bro, are you refreshing Namecheap while the clock runs out? 😂 fastcoach.io is a banger — .io is the classic move when the .com's gone, and it nails the outcome-based naming. favourite.coach is okay, but fastcoach.io is punchier and clearer. Now close the registrar tab. Write "fastcoach.io" on a sticky note, go post your screencast on r/SideProject, get your first 100 visitors — and buy the domain tomorrow as a reward if the numbers are there.

fastcoach.io, no contest. The name should be simple and outcome-oriented — "fast coach," you get coaching fast. getcoachedby.io uses filler words that dilute the punch. But please close the tab — you already have coachcrush.com live and hooked to payments. Save fastcoach.io for tomorrow and go hit launch.

You did it — you fell right into the domain-buying trap 😂 but fastcoach.io is a banger, so I'll allow it. 72 hours is fine — shipping fast is a muscle, and the longer you grind the higher your expectations climb, so let's land this. One: 80/20 the AI — let it write 80%, force the last 20% yourself. Two: strip it to one feature; if something's blocking you, delete it. Three: break the cycle and ship. Get in the editor, force it across the line, and launch.

You nailed it — I did not start out shipping in 24 hours. My first project I spent a whole year in my bedroom trying to make it perfect, burned out, got zero users and made $0. Shipping fast is literally a muscle — it took building lots of projects to learn to cut the noise. So being at 72 hours is completely normal — you're putting in the reps. The fact that you're managing your expectations is the healthiest mindset you can have. Finish this one rep and get it out.

30 minutes per loop is a momentum killer — that alone explains 72 hours. One: don't over-engineer the training loop. If it's too complex for an MVP, simplify it — can you just fetch the transcript and inject it straight into the prompt context? Two: hardcode a perfect response so you can iterate on the chat UI instantly without waiting on the loop, then reconnect it for the final test. Three: 80/20 the AI — let it do the heavy lifting, but when it loops on a bug, jump in and fix the last 20% by hand.

Yes — it's one of the scariest things that can happen online. Early on I got frozen overnight because I hadn't enabled the rule to block payments when CVC verification fails. Someone ran stolen-card test payments through, the real cardholder charged them back, and with few payments overall my dispute rate spiked — an instant red flag. It eventually got sorted, but the lesson stuck: turn on your fraud rules from day zero, and refund anything suspicious before it becomes a chargeback. Protecting your payment account is one of the few things you must do properly from the start.

This is my favorite playbook of all time — building a free tool or side project to funnel traffic to your main business is the ultimate cheat code, and it doesn't feel like marketing. So pivot the CTA from frustration straight into authority: "I built this in 6 days to prove building AI doesn't have to be a generic, unpredictable mess. I run compoundcoders.com, where I teach developers to build trustworthy, reliable AI. Hit follow and check the link in my bio." It proves competence with a live product, filters for high-quality leads, and satisfies both the entertainment crowd and the developers who'll become customers.

Absolutely — put it near the bottom, right after pricing and before the final CTA. It's your safety net for any objection: explain BYOK, state plainly there are zero subscriptions, and cover data security. And it evolves — as support emails come in, fold the real questions back into it. Don't overthink it: write 3–4 questions you expect, ship it, and let customers tell you what else to add.

Here are the four essentials: Is this a subscription? No — one-time $49 lifetime fee, yours forever. How does BYOK work? You plug in your own API key and pay pennies per question directly to the provider; we take no cut. Is my key secure? Yes — it's never stored on our servers and kept private. Does it work with any creator? Yes — as long as the video or podcast has a transcript. That covers pricing, functionality, and security in one block under the pricing box.

Don't let that stop you from hitting launch. When you build in public you'll always get "what if" legal worries, but agonizing over approval for an MVP with zero customers is procrastination. I'm not saying break the law — for real business structure, taxes, and protecting your payment account, do things properly from day zero. But this is a wrapper around public YouTube data to validate a side project. If it gets zero customers, the legal review was wasted; if a creator ever objects, that's a high-quality problem — you apologize, remove their channel, or offer them a revenue split. Ship it and handle real issues if they ever arrive.

Haha, as a social hook that's gold — it creates scarcity and that rebel-developer vibe. But do not put it in the landing-page FAQ. The FAQ exists to remove objections and build trust so people feel safe paying. If someone about to drop $49 on a lifetime deal reads that it might get "taken down tomorrow," they close the tab — why pay for a lifetime that lasts 24 hours? Use the edgy angle in a tweet or TikTok to drive traffic; keep the page itself professional, trustworthy, and safe.

Here's the asset checklist. OG image for socials — treat it like a YouTube thumbnail so the link is clickable. Product Hunt visuals — an animated logo GIF and carousel images, one feature per image, zoomed in close on the UI. Videos — the raw screencast for r/SideProject, plus the funish skits for Twitter/TikTok. Copy — a short outcome-focused tagline (~60 chars), a maker's comment telling your story, and a "Show HN: I made…" title with no emojis and no asking for upvotes. And add a "We're live on Product Hunt" badge to the site to funnel traffic back for upvotes.

On it. Tagline: "Get instant, 1-on-1 advice from your favorite YouTube creators." Maker's comment: an honest note — built it in 6 days because generic AI kept giving robotic advice; paste a channel, plug in your key, ask your question; BYOK means no subscriptions, just $49 lifetime; would love feedback. Show HN: "Show HN: I made an app that turns any YouTube channel into a digital mentor" — explain it fetches transcripts into the prompt context, highlight BYOK so you're not another $20/mo SaaS, and ask for edge cases. No emojis, no upvote begging.

Great platform, used right. Focus on the story, not the pitch — write up your 6-day journey and the frustration with generic AI; IH loves authentic build-in-public posts. You can reuse that copy on other community boards too. But don't expect the IH product page to drive traffic — those pages get almost nothing. Treat your profile there as your maker identity, a deeper Twitter where you document the journey over time.

You've got the Big 5 — Product Hunt, Hacker News, Indie Hackers, Twitter, r/SideProject. Beyond that: cross-promote from your own sites — drop a FastCoach link in the compoundcoders.com footer. Email your list with the same 6-day story. Upload the VSL to YouTube as a standalone video. Skip the dead directories — they bring a handful of visitors and often no backlink. And be glued to your screen the first few hours after launch to answer every comment — Product Hunt momentum is decided early.

Bro, this is a masterpiece 😂 This is exactly the kind of fun, scroll-stopping content that crushes on Twitter and Product Hunt — and animating the cameos with AI is a great touch. The stressed-vs-oblivious dynamic is a fantastic hook, the cameos keep people watching to see who shows up next, and the self-aware "it's a $49 product, we'll sell three copies" line makes you super relatable. When the punchline lands — "Don't worry mate, I got ya!" — cut straight into a raw, high-energy FastCoach demo.

Title options: "I hired famous CEOs to launch my $49 startup," or "Fake it till you make it: hiring celebrities for a $49 app." Description: open by extending the joke (couldn't afford a marketing team, so I used AI to clone the cast), then pivot to the real pitch — sick of generic AI advice, so I built FastCoach.io: paste a channel, plug in your key, get a personalized answer; BYOK means no subscriptions, just $49 lifetime. Link to fastcoach.io, then a line that it was built in 6 days to prove AI doesn't have to be a generic mess, link to compoundcoders.com, and a simple subscribe outro.