A Solo Developer's Cooking Journey
How trying to save on delivery fees turned into an actual hobby
My Delivery Bill Was 340,000 Won a Month
Late last year I checked my credit card statement and found 342,000 won in delivery app charges. Take out the delivery tips and disposable container fees, and the actual food was about 260,000 won. The remaining 80,000 won was pure service fees. Paying 80,000 won a month for convenience suddenly felt like a waste.
But honestly, I couldn't cook. Ramen and fried eggs were the entire repertoire. (And even then, I'd burn the fried eggs.)
It Started with Three Rice Bowls
If you search "beginner cooking for one" on YouTube, about a million rice bowl recipes come up. Tuna mayo bowl, pork bowl, egg bowl. Started with these three. The common thread is that you just put stuff on top of rice, so the failure rate is low.
My first tuna mayo bowl had way too much mayo -- it looked like porridge. By the second attempt I learned to control the portions. Took three tries to figure out that one can of tuna plus one and a half tablespoons of mayo is the sweet spot.
Ingredients cost about 3,000 won per meal. Delivery would be 10,000 won. That's 7,000 won saved. Even just 20 home-cooked meals a month is 140,000 won in savings.
Failure Story: The Doenjang Jjigae Incident
Riding high on confidence, I attempted doenjang jjigae (soybean paste stew). The recipe looks simple enough -- tofu, zucchini, onion, doenjang paste. Just boil it, right?
The problem was the paste quantity. The recipe said "1 tablespoon of doenjang," but I had no calibrated sense for tablespoons. I put in three spoonfuls and the saltiness exploded. Tried to dilute it with more water, but then there was too much broth. Cranked the heat to reduce it, and the bottom burned.
The result was a salty, burnt-tasting soybean soup. Ate half, threw the rest away. Opened the delivery app again that night.
Cooking and Coding Have Things in Common
After a few months of cooking, I noticed parallels with coding. Recipes are algorithms, ingredients are inputs, and the dish is the output. Follow the recipe precisely and you'll usually get something edible. Wing it by feel and there's a high probability of disaster.
The key is reproducibility. Not a one-off lucky dish, but consistently hitting a similar flavor each time. Once I started recording exact seasoning ratios and cook times instead of vaguely adding "a bit" -- fewer failures. I built a recipe database in Notion with seasoning ratios and cooking times.
Current Repertoire
About six months in, I can make 12 dishes. Three rice bowls, two fried rice variants, three pasta dishes, two stews, two side dishes. If I plan the week's menu, I can get through weekday dinners without delivery.
Delivery app spending dropped from 340,000 to 110,000 won. That's 230,000 won saved per month. Groceries run about 150,000 won per month, so the net savings are around 80,000 won. (Exactly the delivery fee surcharge I was paying.)
But if you calculate time-efficiency, it's honestly questionable. Grocery shopping 30 minutes, cooking 30 minutes, dishes 20 minutes. That's 1 hour and 20 minutes per meal. You could argue I'd make more money coding during that time.
Why I Keep Going Anyway
The 80,000 won savings isn't really the point. When you've been staring at code all day, doing something with your hands is a reset. When I'm stuck debugging, I go to the kitchen and chop onions. Chopping onions requires zero design patterns or time complexity analysis. You just cut. That's what I like about it.
Next goal is kimchi jjigae, but apparently the flavor changes depending on how fermented the kimchi is, so I'm a bit nervous.