Some honest feedback: isn't the magic of a postcard, especially a physical one, linked more to the handwritten aspect of the note, and less the voice? My old firm emphasized hand written thank you's for everything because it showed time and effort into this sort of item but I feel like that get's lost with a digital postcard + audio file link. Especially when considering going back and reading notes versus having to scan a link.
Thanks! I agree that handwritten notes are special. My intent was not to replace that so much as offer another way to make a physical note feel personal, while removing some of the "burdens" of sending traditional mail.
It was also just a fun way to learn how to connect all these technologies and be able to trigger an event in the real-world from a few lines of code.
Love that idea! Was thinking it'd be neat if there were little signs at museums/landmarks/etc so people could pull up pre-made templates or upload their own photo to quickly send a card.
The opportunity for abuse with a similar product is too big. Giving someone your voice and the address of your loved ones means they can potentially mock it with AI send anything any other time, including requests for money. The volume of abusers of seniors will only get bigger.
It's definitely an attack vector we'll be seeing more of. My bank uses voiceprints as a password, for example.
My bet is some of the big "we record this conversation for quality and training purposes" databases get stolen and deanonymised for a mass leak of millions of voiceprints.
I currently have a rule with my parents that only I'm allowed remote access to their computer of any kind. And if anyone else wants access they should ask me first. Maybe one day I'll have to add some "nuclear biscuit" pre-shared code cards to that.
Love this! Impressed at the fully automated flow. A while back I was working on video/AR postcards, but more of a B2B play. We were linking music videos and behinds the scenes content to postcards, mailers, merch etc for musicians and record labels. Always wanted to try something in the DTC space. Never got it as automated as Keeps. Nice work!
Just sent one! It's a lovely idea. I think most people love getting postcards and the voice message is a nice surprise for the recipient.
Two things.
1) The photo upload gets stuck on loading from the landing page. I think this is due to the file size? I uploaded the same photo from the actual page where you can customize the postcard and I was given a size limit warning.
2) I tried using a non-US address for both the sender and recipient. Both addresses were verified, but on the preview all of the address was omitted except the street number. This could only be a preview issue but I re-added the address in the optional section just to be safe.
Edit: actually a third thing. The non-US address was recognized with the appropriate city but the country still shows United States. My order number is #2CD56A
This is a nice idea. I just checked it out (not to really send a card though), and I got this error after entering the message and trying to go to the next page:
“Failed to save your card: TypeError: Load failed”
After entering the addresses (recipient and return), the next button seemed to reload the same page. Maybe it was showing some error temporarily? I had to tap on it about three or four times to get to the next page.
Would be nice if your demo was front and center with an option to create your own after it plays.
Also, maybe add a way to email myself a reminder to send a card. Might be a good option to those that would send one for birthdays but can’t do one now. Capturing their email as a reminder is a great way to sell this. Actually, this would be a great product to sell through email
Good idea making the example more prominent. Capturing email to send a reminder for birthday/etc is smart and nice way to have a "lead magnet" actually add value. thank you
First off, congrats on the launch! It looks really polished, and I'm definitely giving it a proper try tomorrow.
How do you plan to market this? From what I can tell, you're only left with about $3.50 (~$0.90 for Lob and ~$0.50 with Stripe), unless there's some kind of upsell after completing the flow that I missed.
Congrats on the launch!
Some honest feedback: isn't the magic of a postcard, especially a physical one, linked more to the handwritten aspect of the note, and less the voice? My old firm emphasized hand written thank you's for everything because it showed time and effort into this sort of item but I feel like that get's lost with a digital postcard + audio file link. Especially when considering going back and reading notes versus having to scan a link.
Curious how you're thinking about all this.
Thanks! I agree that handwritten notes are special. My intent was not to replace that so much as offer another way to make a physical note feel personal, while removing some of the "burdens" of sending traditional mail.
It was also just a fun way to learn how to connect all these technologies and be able to trigger an event in the real-world from a few lines of code.
Appreciate the feedback.
> My intent was not to replace that so much as offer another way to make a physical note feel personal,
Your approach is great; I can see a good need for this, and it is a unique offering which amends rather than replaces postcards in my view.
I could even see you having real post cards that have a QR code to the voice version that you sell in shops.
Love that idea! Was thinking it'd be neat if there were little signs at museums/landmarks/etc so people could pull up pre-made templates or upload their own photo to quickly send a card.
The opportunity for abuse with a similar product is too big. Giving someone your voice and the address of your loved ones means they can potentially mock it with AI send anything any other time, including requests for money. The volume of abusers of seniors will only get bigger.
It's definitely an attack vector we'll be seeing more of. My bank uses voiceprints as a password, for example.
My bet is some of the big "we record this conversation for quality and training purposes" databases get stolen and deanonymised for a mass leak of millions of voiceprints.
I currently have a rule with my parents that only I'm allowed remote access to their computer of any kind. And if anyone else wants access they should ask me first. Maybe one day I'll have to add some "nuclear biscuit" pre-shared code cards to that.
That's a valid concern. There's definitely no voice cloning or information being sold from this app.
Until you get sold.
I thought (from the title) it was going to be a service where the audio was downloaded into a physical sound chip like a musical birthday card.
Love this! Impressed at the fully automated flow. A while back I was working on video/AR postcards, but more of a B2B play. We were linking music videos and behinds the scenes content to postcards, mailers, merch etc for musicians and record labels. Always wanted to try something in the DTC space. Never got it as automated as Keeps. Nice work!
Congrats on launching! Really neat idea, excited to see my kids' reactions when it arrives.
Notes - It wouldn't let me type a message until I uploaded a photo.
- The card preview did not render the Sender or Recipient values, until I went to the next page.
- UI validation (of name fields) should trigger after each key stroke. It still showed an error until I submitted.
- Would be nice to create an account and view/manage all my cards.
Just sent one! It's a lovely idea. I think most people love getting postcards and the voice message is a nice surprise for the recipient.
Two things.
1) The photo upload gets stuck on loading from the landing page. I think this is due to the file size? I uploaded the same photo from the actual page where you can customize the postcard and I was given a size limit warning.
2) I tried using a non-US address for both the sender and recipient. Both addresses were verified, but on the preview all of the address was omitted except the street number. This could only be a preview issue but I re-added the address in the optional section just to be safe.
Edit: actually a third thing. The non-US address was recognized with the appropriate city but the country still shows United States. My order number is #2CD56A
Looks like it was just a display issue. Your card successfully saved both addresses and is queued for printing!
Good to know! Thanks!
The order detail page still shows Sending To as United States. Is this also a display issue?
Thanks! Will look into these potential issues
Fixed the upload size bug on homepage
This is a nice idea. I just checked it out (not to really send a card though), and I got this error after entering the message and trying to go to the next page:
“Failed to save your card: TypeError: Load failed”
After entering the addresses (recipient and return), the next button seemed to reload the same page. Maybe it was showing some error temporarily? I had to tap on it about three or four times to get to the next page.
Sorry you hit some bugs, I haven't seen these yet. Will look at the logs and try to see what's going on.
Would be nice if your demo was front and center with an option to create your own after it plays. Also, maybe add a way to email myself a reminder to send a card. Might be a good option to those that would send one for birthdays but can’t do one now. Capturing their email as a reminder is a great way to sell this. Actually, this would be a great product to sell through email
Good idea making the example more prominent. Capturing email to send a reminder for birthday/etc is smart and nice way to have a "lead magnet" actually add value. thank you
First off, congrats on the launch! It looks really polished, and I'm definitely giving it a proper try tomorrow.
How do you plan to market this? From what I can tell, you're only left with about $3.50 (~$0.90 for Lob and ~$0.50 with Stripe), unless there's some kind of upsell after completing the flow that I missed.
Thanks! Yeah your margin calculations are roughly correct.
This is a fun side project built for myself, that I thought others might enjoy -- so maximizing revenue hasn't been a huge part of planning.
I plan to market this on traditional channels (Reddit, X, etc) and maybe test out some paid ads if the CAC is <$3.50 -- open to ideas!
This is awesome. curious why you opted not to use 11 labs?