Hey y’all, I wrote about my 6-year tenure as a CTO across two fintech startups, specifically around securing banking systems.
I was particularly proud of the GraphQL/Redis solution I came up with which enabled the part "This means there were zero database IDs exposed in any UI or API response"
As others have pointed out, make sure to mark files that are not supposed to be exposed to the client by prefixing them with a underscore, otherwise you'll have a bit of a security issue. Otherwise, you're good to go.
> When you built this, did you validate the idea at all with people in your own network (or extended network) wether they wanted something like this?
Having worked in 'websites' for 9+ years, I spotted the pain points people had with various aspects of updating their website - paying an agency thousands to create/update a simple brochureware site, not being able to navigate various CMS's people put in place, etc. I spoke to multiple people in/out of my network and in my opinion, validated the idea.
> What (in terms of feedback) have you received from people you judge to be your target market?
Good question. I've had less feedback from my target market, vs. my network (designers, developers, etc). I think this is where I didn't do particularly well. I validated the idea with the target market, but in terms of ongoing feedback of the product, I didn't acquire that.
> When you say people don't convert, do you mean they don't even test the product or that they don't convert to a paying subscription?
They reach various phases of the sign up process. It goes: Email/password -> Auth with FB -> Pick FB Page. Then you go through to your Dashboard, where you can view your site online, or pick a new theme, etc. People got to various stages of this, with most people going right through to the Dashboard (a created site), but rarely deciding to pay.
When you say "validate", what does that actually mean? Was it "I tell people the idea, they say it sounds cool"? Presumably it wasn't "People give me money and are sad when I tell them it doesn't exist yet"?
I tend to think validation is a terrible word for the stuff we do when building a product. As well as a product needing to offer something of value, it needs to offer that to people who are willing and able to commit money to acquire that value. If businesses are setting up Facebook pages, that puts them in the "free site" category of customer at a first approximation. Your business is actually an upsell, and I think that's tough.
I just noticed you're in the UK. If you're interested in a coffee/chat at some point, hit me up - I think it's a lovely looking thing, but I doubt you'd be able to sell it to someone given that it has demonstrated it doesn't generate traction so far. (I also think you've put too much work into it - but I'm guessing you already know that).
Hey y'all. I built this last week to scratch a very personal itch.
It looks back through all the playlists you've created, finds overlapping tracks, orders them by occurrence and pushes the top 30 into a new "Top Spot" playlist!
I also recorded the whole sad sorry development as a screencast, which I'll be splitting up into a dozen or so videos (with post-production commentary to provide context) and posting to YouTube so people can see the very messy, copy/pasting flow of creating a React web app in just over 4 hours.
For sure, and the I fully appreciate the irony of me promoting this on Twitter (if everyone used this, no-one would see my tweet). I feel like this can just aid a certain way of using Twitter, ie. by being broadcast-only (maybe posting from a 3rd party app), and just being selective on what you reply to (based on the email Disconnect Today sends you).
You'd be wise to think that, but I actually use Heroku's "Scheduler" add-on, so they just fire up a one-off dyno once a day and run my script. This script just checks the database for the user's it needs to check (paying && valid email) and fires off Twitter API requests for them. If there's @mentions/DMs that fit in with the frequency the user has selected, the email is fired off through AWS SES. Honestly, the server costs for this will be quite low for quite some time. If it takes off big time, sure, I'll re-evaluate the pricing model.
I was particularly proud of the GraphQL/Redis solution I came up with which enabled the part "This means there were zero database IDs exposed in any UI or API response"