Notes from building the Studio Toys websites: what I tried, what I decided, and why. I build these sites on the side, mostly because I think they'll help people.
Looking back at when Studio Toys actually moved forward, I noticed a pattern I'm not proud of at first glance. Most of it got built while I was between jobs.
The part that sucks
Being between jobs is stressful. There's the money, the waiting to hear back, and the quiet worry about what comes next. Nobody plans a side project around that. When I had steady work, Studio Toys mostly sat still. A row would say "coming soon" for months, sometimes years, because by the end of a workday there wasn't much left in me.
The part that was secretly a blessing
But those in-between stretches gave me something a full schedule never did: long, unbroken time to think and to make things. I could follow an idea all the way through instead of picking it up for twenty minutes and putting it down again. Building also kept me steady. On days when the job search went nowhere, I could still finish something, see it work, and feel like the day counted.
A lot of what's on the list came from those stretches: charts I'd wanted for years, a shop for a cat logo I kept drawing, a book I'd been meaning to finish.
What I'm doing differently
I don't want my projects to depend on bad luck. So now I keep things small enough to work on in short sessions: one chart, one listing, one blog post at a time. I'm also focusing on the few sites that can earn a little, so Studio Toys can someday be a cushion for the in-between times instead of just a way to fill them.
If you're between jobs right now
I'm sorry. It's hard, and it's not a reflection of what you're worth. If you can, make something small while you wait. It doesn't have to be a website. It just has to be yours, and finished. It helped me more than I expected.
Studio Toys has been around since 2020, but for most of that time it was a page of ideas more than a page of websites. Plenty of rows said "coming soon" and stayed that way. This fall I decided to change that, and now I'm relaunching it: every website on the list is either up and running or has a preview you can open.
What's on the list
Some of the sites are meant to be useful every day: Cool Charts (an encyclopedia where every topic is a chart), Mostly Money (sign-up bonuses and small ways to earn money online), Simplest Shopping (a short list of simple, useful things from Amazon) and Copycat Clothing (minimal shirts with cats). Some are things I care about: Soul Search, Diversity Corner, Amusing Mom and my book, Humble. And some are just for playing with, like Isolation House and Some Downtowns.
Why so many?
Fair question. I've been told more than once that I should pick one. The honest answer is that each one started because I wanted it to exist and couldn't find it done the way I wanted. Going forward, I'm putting most of my time into the few that can pay for themselves and keeping the rest as my playground.
What changed behind the scenes
Every site now has its own analytics, a sitemap and a proper description for Google, so people can actually find them. I also added a sign-up box on studio.toys, so you can get a short note whenever something new launches. The posts below go into how some of these came together.
Copycat Clothing started as a logo. I had a cat drawn out of the letters for ctrl + c, the shortcut for copy, and I kept thinking it would look good on a plain shirt. So I built a small shop around it.
The name and the cat
It started out as Copycat Concept. That sounded like a design studio, not a place to buy a shirt, so it became Copycat Clothing. The cat changed too. I tried a lot of versions and ended up with one squished slightly sideways, which looks friendlier and sits better on a chest. Along the way I added a second design, a small cat asking うそ? ("really?") in Japanese, placed just under the collar like a pendant.
What I decided not to do
I almost let customers design their own shirts. Then I realized plenty of sites already do that, and do it well. Copycat is better as a short list of designs I actually like. I also kept prices low. I'd rather people try a shirt than make a big margin on the few who do, and shipping is free on orders over $50.
How the shirts get made
I don't keep any stock. Each design is set up with Printify, a print-on-demand service, and when an order comes in, they print and ship it. The product photos are Printify mockups with the model cropped out, because I wanted real-looking shirts rather than computer-drawn outlines.
The payment surprise
This was the part I didn't see coming. I set up a new Stripe account for the shop, the same way I did for Cool Charts. Before a single order, Stripe closed it as high risk. I never found out exactly why. New clothing shops that sell printed designs seem to get flagged a lot. Rather than fight it, I switched the checkout to PayPal. It turned out to be much easier to set up, and the shop is now open for orders.
When I looked for a place to compare religions, or to find a congregation to visit, most of what I found came from one faith's point of view, comparing everyone else against it. Others were plain directories that list every building with a map pin. I wanted something calmer: compare the religions side by side, then find a few places worth visiting near you.
Picking places by how welcoming they are
Soul Search doesn't try to list everything. I pick places, starting in the San Francisco Bay Area, and one thing I look at is their website. A congregation that keeps a clear, friendly website usually takes good care of newcomers too. Each listing shows four photos in the same order: the outside of the building, the inside, a group of people, and a special event. That gives you a feel for a place before you walk in.
For everyone, including people just looking
I chose the name soulsearch.help instead of something like .faith on purpose. I want someone who isn't religious and is just curious to feel welcome looking around. The listings cover churches, mosques, synagogues and large Hindu and Buddhist temples, and I try to give every city at least a few choices.
Small decisions that mattered
I took Google star ratings off the cards. A rating says little about whether a community is right for you. I also stopped linking to photos on other websites, because they kept disappearing. Every photo is now saved with the site, with credit where it came from.
I loved the old game Theme Hospital, where you lay out rooms from a tilted, isometric view and drop furniture in place. Isolation House is that idea pointed at real life: lay out your own apartment and see how it feels before you move a single couch.
The feature I wanted most
Room planners already exist, and most of them work fine. What none of them gave me was a sense of scale. A floor plan can say a room is 10 by 12 feet, but that doesn't tell you how it feels to stand in it. So Isolation House lets you drop in a person and set their height. It's a simple wooden mannequin, like the ones artists use. Put them on the rug next to the bed, and suddenly you know whether the room is cozy or cramped.
Choices along the way
The grid uses 6-inch tiles, small enough for real furniture to fit properly. You drag furniture around right on the isometric view instead of on a flat plan, because that's the fun part. The furniture is matte and softly rounded, and the default colors are a calm beach palette with coral beds. The name came easily: "isolation" echoes "isometric," and an apartment is your own little island.
Where it stands
It's in beta right now. It works, and I'm adding the features that will bring it to a 1.0. If you try it, I'd love to hear what you build.
A website nobody can find might as well not exist. As part of the relaunch, I went through every site and did the boring work that helps search engines find it. Here's what that involved, in case you run a small site too.
The basics every site needed
A sitemap. This is a small file that lists every page you want found, so Google doesn't have to guess. Some Downtowns alone has over fifty city pages. A robots.txt file that points to the sitemap. A description and share tags on each home page, which control the text under your link in search results and the preview when someone shares it. Several of my sites had these left blank.
Proving the sites are mine
Google Search Console, the free tool that shows how your site does in search, first makes you prove you own the domain. That usually means adding a line to your DNS settings. My hosting's DNS editor turned out to be picky: it wanted the full domain name with a dot at the end, which took a while to figure out. Bing Webmaster Tools can import everything from Search Console, so you only do it once.
The surprise: my biggest site was the hardest to find
Cool Charts builds each page in the browser. To Google, every page first looks the same, with the same title and description. Some of my best charts also sit inside embedded Google Sheets, and that content counts for Google's page, not mine. So the next jobs are giving every chart page its own title and description from the server, and turning embedded sheets into real pages.
Small things that add up
Pick one version of your address, with or without "www," and stick to it everywhere. Make sure phones can see something useful. Cool Charts now lets phone visitors send the page to themselves to open later on a bigger screen. And be patient: it can take weeks for new pages to show up in search at all.
Cool Charts started as a simple idea: an encyclopedia where every topic is a chart you can take in at a glance, instead of a wall of text. For a long time it was a plain set of HTML files with a donate button. The donate button never really brought anything in, and I didn't expect it to. So this fall I rebuilt the site so it has a way to earn money, while keeping the free side as useful as it already was.
What stayed free
Everything that was already on the site stayed free: fallacies, rhetorical devices, the personality charts, Compare Religions, Nutrients, the money tools and more. I didn't want to take things away from people who already use it. Even Parenting, which I had planned to lock at one point, stayed free.
What changed
Free accounts. Anyone can make a Basic account. It saves your bookmarks and your test results to your account, so they follow you to any computer instead of living in one browser.
Full Access for $5, once. One payment through Stripe unlocks every premium chart, including new ones as I add them. No subscription. Stripe hosts the card form, so card numbers never touch my server.
A real server. The site went from static files to a small Node server with a database. That's what makes accounts and the paywall possible. I wrote about that move in the third post.
Hiding premium charts instead of teasing them
The first version showed locked previews of every premium chart: a title, a teaser and an Unlock button. I changed my mind. Visitors should be focused on the free charts, not on a wall of things they can't open. Now premium charts appear greyed out in the sidebar, the same way my "coming soon" topics do, and the Full Access pitch stays general without a chart count.
There's a real tradeoff here. People are more likely to pay for something they can see. I chose a calmer site over a pushier one, and I'll let the numbers tell me if that was right.
Making the paywall hold up
A paywall is only as good as its weakest file. While reorganizing, I found that a few premium charts were still sitting in the main folder, where anyone could open them with the right address. I moved every chart into a charts/ folder, with premium ones in charts/premium/. The server now checks every request for a premium file, however the address is written. The one gap left is the handful of charts that still live in Google Sheets, since anyone with a sheet's link can open it. Those are on my list to turn into regular pages.
Filling Full Access with charts worth paying for
My rule for Full Access is that everything listed is actually there. I only add a chart where a good one is possible, not just to fill a slot in the sidebar. To build them faster, I had Claude make a small generator that turns a data file into a finished chart page in the Cool Charts style. That let the premium side grow to around ninety charts, covering things like pet care, first aid, cosmology, home repairs, music theory, ethics and conflict resolution.
The charts I'm proudest of aren't glossaries. They put a whole topic side by side so you can compare everything at once, and that's what I want more of.
Tests are the part nobody else does
The personality and politics tests turned out to be the most interesting thing on the site. An encyclopedia tells you what something is. A test tells you where you fit. So I added a test engine that can attach a test to any chart: "which one are you" tests, "which one fits you" quizzes and "test yourself" knowledge checks. Your result gets highlighted right on the chart and saved to your account. I think this is what can make Cool Charts different.
Two decisions that limit me on purpose
Desktop only. Wide comparison charts don't work on a phone screen, so phones get a polite "use a tablet or computer" message with a button to send yourself the link. I'd rather do the big screen well than do both badly.
No ads for my other sites. Cool Charts doesn't advertise my clothing shop or anything else, especially not on the front page. It should feel like a reference, not a billboard.
Will anyone pay?
Honestly, I don't know yet. After Stripe's fee, a $5 sale is about $4.55, once per person. Earning even $100 a month would take about 22 new members every month, and that depends far more on how many people find the site than on the paywall itself. So the next job is traffic: making each chart easy to find on Google. The paywall is built and switched on. Now I find out whether the charts are worth it.
Some Downtowns shows when a downtown is usually busy or quiet, hour by hour, on a flat, colorful map of the whole downtown. It started as one mockup of downtown Carmel-by-the-Sea. This post is everything I asked for along the way, in order, because the path from "neat mockup" to "real website" was more interesting than I expected. I built it with Claude, so most of these steps started as a single request.
1. Make the Carmel mockup a real website
My Carmel mockup was a square grid of buildings, each with a little icon for what it is, colored by how busy it gets, using foot traffic estimates from BestTime. My first ask was to turn that into a live public site. I also decided right away that I wouldn't charge for it if I could run it for free.
2. Not just Carmel: every Bay Area downtown
One town is a toy. I wanted every downtown in every major Bay Area city, with busyness refreshed about once a week. That turned it into its own site, so I named it Some Downtowns and got somedowntowns.com.
3. Start with three towns
Carmel is a neat grid, and most downtowns aren't. Before going big, I asked to start with three cities to see whether the flat grid style could handle messier street layouts. The first towns ended up being Carmel, Los Gatos, Palo Alto and Santa Cruz, drawn by hand.
4. Throw out the old way the site was built
Back in 2022 I had built an early version of this idea using one page that loaded everything else inside a frame. I asked to get rid of that completely. Now every page is a normal page with its own header and footer, which loads faster and is easier for Google to read.
5. A clickable map of the Bay Area
In 2022 I made a pixel-style map of the Bay Area that I always liked. I asked for the home page to use that idea: a square-grid map of the region where you click a town to go to its downtown. Each downtown keeps the same flat grid look. I specifically didn't want it to look like Google Maps.
6. Real streets from OpenStreetMap
Drawing every town by hand wasn't going to scale, so the site now has a builder that pulls real streets, buildings, parks, water and businesses from OpenStreetMap and redraws them in the flat style. The first test was Oakland, and I liked how it came out. My only note was that the map icons were too big, so they got smaller.
7. Places worth visiting
A busyness map tells you when to go. I also wanted it to tell you where to go. So each downtown got a hand-picked "Worth a visit" list of notable businesses, like Wakamatsu in Hayward, with a button to show each one on the map.
8. The domain, SSL and a logo
I set up somedowntowns.com on Namecheap. The secure "https" certificate was the tricky part: my Namecheap control panel doesn't have the SSL page most guides tell you to use, so it goes through Namecheap's own SSL plugin (or their support) instead. For the logo, I combined the red and blue building icon from the 2022 version with the current "SOME DOWNTOWNS" wordmark, so the old project lives on in the new one.
9. Real data, and what it costs
This is where a free side project meets the real world. I want busyness refreshed every week, and ideally near-live data too. Live data costs money each time it's fetched, so the plan is to fetch it only when someone actually looks at a town, with a hard cap so it never goes over about $50 a month.
For the bigger data pulls, I went with Apify, a service that runs ready-made tools to collect public data from sites like Google Maps and Zillow. I subscribed to their $19-a-month Starter plan and set myself a budget of about $60 for the first full pull across every downtown, starting with a test pull for Carmel. Apartment rents come from Zillow's free rent index, credited next to every figure.
10. Where it's going
My goal is "Google Maps, but better" for downtowns: the useful parts of Google Maps, plus colored buildings, a clear legend and a better busyness feature. On the list after that: which storefronts are vacant, a "start a business here" link to each chamber of commerce, apartment and hotel prices around each downtown, and booking a table or a room right from a place's profile.
Some Downtowns is in beta. Some towns still show sample busyness data until their real data is loaded, and each map says so at the top. If your downtown is missing or something looks off, the feedback page is the fastest way to tell me.
For years, putting a website online meant uploading files to my Namecheap hosting. That works fine for plain pages. But once Cool Charts needed accounts and a paywall, it needed a server and a database, and file uploads weren't enough anymore. Here's the setup I landed on, written for anyone who's comfortable with HTML but has never done this part.
The three pieces
GitHub holds a copy of the site's code. Mine is a private repository called cool.charts, so only I can see it.
GitHub Desktop is a free app that sends my changes to GitHub with a couple of clicks, without having to use the command line.
Railway runs the site. It watches my GitHub repository, and every time I push a change, it rebuilds and restarts the site on its own.
Setting it up
Put the site folder into a private GitHub repository using GitHub Desktop.
In Railway, create a project from that GitHub repository.
Add a PostgreSQL database to the same Railway project. The server creates its own tables the first time it starts, so there was no database setup by hand.
Add the settings the server needs as Railway variables: the database address, a long random secret that keeps sign-ins secure, the site's address, the Stripe keys for payments, and the email settings for password resets.
Point coolcharts.org at Railway instead of my old host.
The most important rule: passwords and keys never go into GitHub. They live only in Railway's variables. A file called .gitignore makes sure the local settings file and the installed packages never get uploaded, even by accident.
My everyday routine now
Edit the site in VS Code.
Preview it on my own computer with the Live Server extension. I added a small switcher in the corner that lets me see the site the way a premium member, a basic member or a signed-out visitor would. It only runs on my computer and does nothing on the real site.
Open GitHub Desktop, look over the list of changed files, write a short note and click Commit.
Click Push. A minute or two later, Railway has the new version live.
That's the whole loop. A nice side effect: the "Updated" bar on the Cool Charts home page fills itself in. When I add new chart files and push, the server notices them on startup and records the date, so I never edit that by hand.
Things I learned
GitHub Desktop doubles as a safety net. Every commit is a snapshot I can go back to, and seeing exactly which files changed before I push is a good last check.
Test with fake money first. Stripe gives you test keys and a test card number, so you can run through the whole purchase without spending anything before switching to live keys.
Order matters on the server. The check that stops non-members from opening premium charts has to come before the code that serves the public files. If it ever got moved below that line, premium charts would be wide open. There's a comment in the code warning future me not to move it.
One address, not two. Visitors to www.coolcharts.org get sent to coolcharts.org, because sign-ins belong to one address and Stripe sends buyers back to one address.
Watch for a sleeping server. Railway can put a service to sleep when nobody is using it, to save money. The first visitor after a quiet spell then waits while it wakes up. If your site is sometimes slow for no clear reason, check that setting first.
Was it worth it?
Yes. It's more moving parts than uploading files, but once it's set up, publishing is just Commit and Push. And it's all regular code I can take to another host if I ever need to. My other sites are still plain files, and that's fine. A server is only worth it when a site actually needs one.