KindlyChecked Launch Kit
🚀 Friends-test playbook

Get real users to real feedback.

Everything you need to put KindlyChecked in front of 50–200 friends and learn what's working before you take it wider. Pre-launch checklist, three voice-tested launch messages, install instructions, feedback form questions, and the criteria for "v1 is done."

9 sections
·
~15 min to set up
·
Use everything in order
1
Before you send anything

Pre-launch checklist

A 10-minute pass before you send the link to anyone. Click each item to mark it done.

  • App is deployed and live on a real URL (not localhost)
  • You've installed it on your own phone (iPhone and Android if possible)
  • You've scanned at least 5 real products at a grocery store and they all worked
  • You've tried scanning something obscure to confirm the "not found" screen looks good
  • You enabled at least one allergen and confirmed the warning banner appears on a matching product
  • You've shared a result via the share button — confirms sharing works
  • You turned off Wi-Fi and reopened the app — confirms offline cache works
  • Feedback form is ready (Section 4 below)
  • You've decided who's in the first wave (5–10 close friends) before going wide
⚠
The first-wave step is critical. You'll find at least 3 embarrassing bugs that you only catch when other people use it. Find them with friends who'll forgive you, fix them in a day, then go wide.
2
Voice-tested copy

The launch messages

Three versions, picked by audience and channel. Edit the voice to sound like you — these are starting points, not scripts.

📱 Short version
For stories / quick DMs
Made a thing 🌱 [link] KindlyChecked: scan any food, cosmetic, or cleaning product to see what's actually inside. Free, no account. Would love your feedback.
💬 Medium version
For group chats
Hey — I've been building something and I'd love your help testing it. KindlyChecked scans the barcode on any food, cosmetic, or cleaning product and gives you a 1–100 health score, tells you why, and suggests cleaner alternatives. Like Yuka but expanded. Open it on your phone, give it 5 minutes, scan a few things in your pantry or bathroom. If you hit anything broken or weird, please tell me — there's a feedback form at the bottom of the profile screen. [link] Works on iPhone and Android, no account, free. To install it like an app: open in Safari/Chrome → "Add to Home Screen."
✉️ Longer version
For the people whose feedback matters most
Hi [name] — I've been working on a side project and you're someone whose taste I trust, so I want your honest read. It's called KindlyChecked. Scan the barcode on anything in your kitchen, bathroom, or under the sink, and it tells you in plain language what's in it, scores it, and recommends better swaps. Built on top of Open Food Facts, so it has 3 million+ products in its database. What I'd love from you, in order of usefulness: 1. Just use it for a week. Scan stuff before you buy it, scan stuff already in your house. See if it changes anything. 2. Tell me when it breaks. Wrong score, weird UI, anything that confused you for even 2 seconds. 3. Tell me what's missing. What would make you actually keep using it? Link: [link] Open it in Safari (iPhone) or Chrome (Android), tap "Add to Home Screen" so it lives on your home screen like a real app. Free, no account, no ads, no tracking. Thanks for being a guinea pig 🥬
3
Reduce confusion

How friends install it

Most people won't know how to install a PWA. Send this with the link, or pin it in your group chat.

🍎
iPhone
Safari only
  1. Open the link in Safari
  2. Tap the Share icon at the bottom (square with arrow up)
  3. Scroll down → "Add to Home Screen"
  4. Tap "Add"
  5. The KindlyChecked icon appears — tap it to open like an app
🤖
Android
Chrome / Edge
  1. Open the link in Chrome
  2. Tap the install banner that appears, OR
  3. Three-dot menu → "Install app"
  4. Tap "Install"
  5. Icon appears in your app drawer
💻
Desktop
Any modern browser
  1. Just open the link
  2. No camera scanning, but you can paste barcodes manually
  3. Useful for testing, less useful for real shopping
📸
Want to be extra-helpful? Screenshot each step on both platforms and post them as a 5-image carousel on your story or a Notion page you link from your message. Takes 15 minutes, pays back tenfold in fewer "how do I install this?" replies.
4
Capture signal

The feedback form

Use Tally.so (recommended — beautiful, free, mobile-friendly) or Google Forms. Setup time: 5 minutes.

Link it from inside the app (Profile → Help & feedback) and from your launch message.

8 questions, ~2 minutes to fill out

1
What's your name?
SHORT · OPTIONAL
Lets you follow up with people who give great feedback.
2
How many products did you scan?
MULTIPLE CHOICE
  • ○ Just 1 or 2 (poked around)
  • ○ 3–10 (gave it a real try)
  • ○ 10+ (became a habit)
3
What's one thing you loved?
LONG ANSWER
Even a sentence helps.
4
What's one thing that confused or annoyed you?
LONG ANSWER
The honest answer is the useful one. Be brutal.
5
Did anything break?
LONG · OPTIONAL
Bugs, wrong scores, missing products. Paste a screenshot if you can.
6
On a scale of 1–10, how likely are you to keep using this?
LINEAR SCALE
This is your North Star metric. Track the average over time.
7
What would make you a 10?
CONDITIONAL · LONG
Only show if answer to #6 is < 10. The most useful question on the form.
8
Can I follow up if I have questions?
YES/NO + EMAIL
Build a list of engaged early users for v2.
📊
After 2 weeks, do this: Read every response in one sitting. Sort by Q6. Words that show up 3+ times in Q4 = your real product problems. The 9s and 10s = your beachhead users. Ignore the 1–4s for now — you can't be for everyone in v1.
5
Lightweight analytics

What to track

Set up Plausible ($9/mo, privacy-friendly) or Vercel Analytics (free if you deploy on Vercel). 5-minute setup.

Five metrics for the first month:

Signal of reach
Total visitors
How many people opened the link
North Star
Returning visitors
If people come back, you have something
Engagement
Sessions per user
3+ = actually using it, not looking once
Funnel health
% reaching result page
Friction in the scan flow shows here
Product signal
Top scanned products
Tells you what category to expand next
✋
Don't install Mixpanel. Don't install GA4. Time spent setting up complex analytics is time not spent fixing bugs. Five metrics is enough for v1.
6
1 week after launch

The follow-up

Send to everyone who got the original link, even if they haven't responded.

⏰ Day 7 follow-up
Soft, no pressure
Hey — quick check-in on KindlyChecked. Have you had a chance to try it? If yes: what's one thing I should fix? Even small stuff. [feedback form link] If no: totally fine, I get it. If you want to give it a quick scan now, link's still live: [link] Either way, thanks 🌱

You'll get more responses from this than the original launch — the launch message gets buried fast in group chats.

7
Triage incoming reports

What to do when bugs come in

You will get bug reports. Most fall into these categories:

  • "It said XYZ scored badly but I think it should be fine" The score is a function of the data and the algorithm. Decide if the algorithm needs tuning or if the user just disagrees. Both are useful info — they need different fixes.
  • "It couldn't find [obscure product]" Working as designed. The not-found screen exists for this. Confirm they saw it. If they didn't, there's a real bug.
  • "It crashed when I [X]" Real bug. Get exact steps to reproduce. Don't try to fix it from one report — wait for two reports of the same thing before assuming it's not a one-off.
  • "It's slow" Probably the API. Check if Open Food Facts is having issues. If consistent, the v2 conversation is about caching/CDN.
  • "The camera doesn't work" 80% of the time: they're on iPhone Chrome (Apple restriction — PWAs can't access camera there). Tell them to use Safari. 20%: they denied permissions; tell them to enable in browser settings.
📝
Keep a running doc of bug reports. Group them by frequency. Fix the top 3 every weekend until friends-test wraps. Don't try to fix all of them — diminishing returns.
8
Resist forever-iteration

Knowing when v1 is done

You'll feel pressure to keep iterating forever. Resist it. v1 is done when you hit all four of these:

v1 done criteria

  • You've gotten 20+ feedback form responses
  • The average rating on Q6 (likelihood to keep using) is 7+
  • You've fixed every bug reported by 2+ people
  • The top 3 feature requests are clear from the data

Then you stop iterating on v1, write a v2 spec, and decide whether to:

  • → Add accounts + cross-device sync
  • → Submit to app stores via Capacitor
  • → Launch publicly (Product Hunt, etc.)
  • → Pivot the focus based on what friends actually loved
  • → Do nothing for a month and let the dust settle
💡
Whatever you decide, decide it deliberately. Most side projects die from indefinite iteration on v1, not from failed v2s.
9
Future-you will thank you

The story you'll tell later

Save these. You'll want them for whatever comes next — talking to investors, talking to press, telling your origin story, or just remembering.

📸
Launch screenshot
A screenshot of the launch message you sent
📊
Raw responses
Tally responses, exported as CSV every month
😱
Surprise scans
Screenshots of scans that surprised you
💌
First stranger
The first message from someone you don't know
🌱
"Native version?"
First time someone asks if there's an iOS/Android app — that's signal

This is the early data that becomes a story. Future-you will thank present-you.

Now, go.

You've got an app, a plan, the messages, the form, and a clear bar for "done." Stop reading guides. Start sending links. Good luck. 🌱

Send friends the User Guide alongside the install link — it's the difference between "I scanned one thing" and "I use this every week."