The $7 Notebook That Replaced My Grocery Anxiety

A personal price book is just a written record of what you actually pay for staples you buy repeatedly, kept somewhere you'll actually consult. I use a $7 waterproof notebook, track 17 items across four stores, and update it every three weeks in about twelve minutes. The system fails if you track too much or make it complicated; mine works because I ignore everything except the core basket that determines 70% of my weekly spend.

Start with the 17, not the 47

My first attempt crashed in February 2026 because I tried to log every price I saw. Forty-seven items, color-coded by store, abandoned by week three. The second attempt worked: I asked what I actually buy every single month regardless of sales. That list was seventeen items: eggs (18-count), whole milk (gallon), butter (1 lb unsalted), boneless chicken thighs, dried black beans (2 lb), long-grain white rice (5 lb), all-purpose flour (5 lb), olive oil (17 oz), yellow onions (3 lb bag), carrots (2 lb bag), canned diced tomatoes (14.5 oz), pasta (16 oz), shredded cheddar (8 oz), ground beef (80/20, per lb), sandwich bread, coffee (12 oz), and peanut butter (16 oz). These determine my baseline. Everything else is optional noise.

The 1.7x rule that exposes fake sales

I learned this from the 1.7x price multiplier hiding in plain sight: if a "sale" price is more than 1.7 times my recorded lowest price, it's not a sale, it's a psychological trick. My lowest logged price for 18-count eggs is $3.49 (Aldi, July 2026). The local premium market regularly advertises eggs at $5.99 as "this week's special." That's 1.71x. I walk past. The price book only works if you trust the floor price more than the signage. Your memory will betray you; the notebook doesn't.

Four stores, one entry per trip

I shop Aldi, Lidl, Costco, and a regional chain I'll call FoodSmart. I do not track every price at every store every week. I track one: whichever store I'm in. The entry takes ninety seconds in the parking lot before I unload. Date, store, then only the items on my 17-item list that I actually bought or noticed. If eggs are $4.29 and I didn't buy them, I still log it. The goal is building density, not completeness. After six months I have 12-18 data points per item across stores. That's enough to see patterns without becoming a person who enjoys data entry.

The notebook that survived a rainstorm

I use a Rite in the Rain pocket notebook, 3.5 by 5 inches, $6.95 at a hardware store in March 2026. Waterproof paper, bound with wire that hasn't rusted. It lives in the car's center console, which means I cannot forget it without also forgetting how to drive. Previous attempts used phone notes (distracting, easy to ignore) and a spreadsheet (required opening a laptop, which I never did). The physical object creates friction in the right direction: I must handle it to shop, and handling it reminds me to check prices. It has survived one coffee spill and a thunderstorm in the parking lot at Costco.

When the system breaks: the missing item problem

The price book fails when you need something not on the list. In May 2026 I needed coriander for a recipe. No data. I stood in the spice aisle paralyzed by choice between $4.89 organic and $2.99 conventional, with no anchor for what coriander should cost. I bought the $2.99 and added coriander to a secondary list: "occasionals" I buy 3-4 times yearly. That list now has eleven items. I do not track them with the same density. I accept higher variance for lower frequency. The alternative is expanding the primary list until the system collapses under its own weight.

The weekly shop lie I stopped believing

Your weekly shop is built on a lie: the assumption that you need everything fresh, every week, from the same place. My price book revealed that I was paying 34% more for milk, eggs, and butter by buying them at FoodSmart for "convenience" rather than routing through Aldi every third week. The time cost of the second stop is eleven minutes. The savings is $8.40 per week. Annualized: $437. I now make the second stop without resentment because the notebook proved the math. The feeling of inconvenience was accurate; the feeling that it didn't matter was wrong.

Updating without obsessing: the 12-minute rule

I update the price book every three weeks, not weekly. Weekly updates made me anxious about precision I didn't need. Every three weeks catches enough price movement to matter without making the task feel constant. The update takes twelve minutes: I flip through the notebook's raw entries, transfer any new lows to the master list in the front, and cross-check against my last master update. If an item hasn't appeared in two update cycles, I flag it for attention next trip. The master list fits on one page. The raw entries fill the rest. I do not digitize. Digitization is a trap that leads to formatting and then abandonment.

What I deleted from the system

I removed unit price calculations. Every guide tells you to calculate cost per ounce, but I was doing this in the aisle with a phone calculator, blocking traffic, making errors. Now I record only the package price for the standard size I actually buy. For rice, that's always 5 pounds. I don't compare against 2-pound bags or 20-pound sacks because I don't buy them. The price book serves my actual behavior, not an optimized theoretical shopper. I also removed sale dates. Tracking when sales occur felt strategic but never changed what I paid. The lowest price matters. The calendar pattern does not.

The privacy case for paper

I considered apps. Most require accounts, location tracking, or permission to scan receipts. My grocery data is not something I want profiled, sold, or used to train models that predict my weaknesses. The notebook generates no data exhaust. It cannot be breached, monetized, or discontinued by a startup running out of funding. The tradeoff is real: I cannot sort or search my records digitally. But I also cannot be manipulated by a dashboard designed to increase my engagement with a platform rather than decrease my spending. The notebook is slower and dumber and therefore more honest.