> ## Content Index > Fetch the complete content index at: https://fern.happychases.com/llms.txt > Use this file to discover other available public pages before exploring further. # A year of gardening by spreadsheet: what the data told me - URL: https://fern.happychases.com/a-year-of-gardening-by-spreadsheet/ - Published: 2026-08-25T10:00:00.000Z - Updated: 2026-11-03T21:30:10.000Z - Description: One analyst, one terrace and one very plain spreadsheet. A guest post on what a season of tracking sowings and harvests revealed, from earlier-isn’t-better tomatoes to the courgette problem. - Author: Tomás Reyes - Tags: Garden, Small software I should say up front that I’m a spreadsheet person. By day I analyse numbers for a living, and at weekends I grow vegetables on a small terrace plot. It was only a matter of time before the two collided. Last year I tracked every sowing, every transplant and every harvest in a single spreadsheet, and I learned things about my garden I would never have noticed otherwise. Ines kindly invited me to share what I found. I’ll warn you now that this is a small sample from one terrace in one climate, so please don’t treat any of it as gospel. But I think there are some ideas here that are useful whether or not you ever open a spreadsheet. ## What I tracked I started with far too many columns. By March I had fourteen, including soil temperature and the phase of the moon, which I abandoned within a month. What survived was a short list: - **Crop and variety.** Because “tomato” isn’t one thing. - **Sow date, germination date, transplant date.** Three dates that together tell the story of a plant’s early life. - **First harvest and last harvest.** To see how long each crop actually produced. - **Weight harvested, in grams.** I put a kitchen scale by the back door and weighed everything. - **A free-text note.** Often a single sentence like “slugs ate half”. That was it. The lesson, learned the hard way, was that the more complicated the sheet, the sooner I stopped filling it in. A spreadsheet you abandon in April produces no insight at all. ![A wooden crate of freshly picked tomatoes, courgettes and beans](https://picsum.photos/seed/fern-vegetable-harvest/1600/900) A good August morning, and a few hundred grams to log. ## Finding one: sowing earlier isn’t always better I’d always assumed that earlier sowing gave better results. The data suggested otherwise. My tomatoes sown in early February germinated well, but they became leggy and stressed on the windowsill long before it was warm enough to plant them out. The ones I sowed in mid-March caught up within weeks and, in the end, produced more fruit per plant. That single comparison saved me a great deal of fuss this year. I’ve moved my tomato sowing to March, freed up space on the windowsill, and the plants look healthier for it. ## Finding two: I was growing the wrong things The scale was humbling. Courgettes accounted for nearly a third of everything I harvested by weight, a fact that surprised nobody who has ever grown courgettes. Meanwhile, my runner beans, which I’d been enthusiastic about, barely produced anything worth the space they occupied. Once I looked at yield per square metre, a clearer picture emerged. Some crops earned their place easily, like salad leaves, herbs and tomatoes. Others, like the main-crop potatoes, took up a huge amount of room for a modest return. This doesn’t mean I’ve stopped growing potatoes; I just now do it on purpose, because I love the taste of fresh ones, and not because I assumed they were efficient. ## Finding three: bolting has a calendar Lettuce was the most reliable pattern in the whole dataset. Anything I sowed after the middle of May bolted early, usually within five or six weeks. Anything sown before that, or in the cooler weeks of late summer, lasted noticeably longer. It’s the kind of thing experienced gardeners just know. But as someone with no gardening background, I found it quietly thrilling to see the pattern appear in a chart. ## The failed experiments Not everything worked. I tried logging the daily weather and quickly found that I had no way of connecting it to anything useful with only one season of data. I also tried to calculate whether my vegetables were cheaper than shop-bought, and abandoned that when I realised I’d have to estimate the cost of a bag of compost, half a shed and about four hundred hours of Saturday mornings. Some things aren’t worth measuring. ## Tools, such as they are I use a plain spreadsheet with one tab, one row per sowing, and one rule: if I can’t fill the sheet in within thirty seconds, I’m recording too much. A pivot table at the end of the season is enough to answer most of my questions. I’ve tried garden-planning apps, and some are lovely, but none of them gave me the pleasure of building my own questions. ## What I’d tell someone starting out 1. Track fewer things than you think you should. Five columns is plenty. 2. Weigh your harvest. Counting is easier, but weight lets you compare crops fairly. 3. Write a note every time something goes wrong. Those notes become the most useful column of all. 4. Don’t expect conclusions from a single season. Treat year one as a baseline. 5. Look at the data once, in winter, with a hot drink, and decide what to change. ## A garden isn’t a dataset I’ll finish with a caution, because I can see how this could turn into an obsession. A garden isn’t a dataset, and the numbers aren’t the point. What the spreadsheet really gave me was a reason to pay closer attention, week after week, to what was happening on that small terrace. The attention is what made me a better gardener. The data is just the souvenir. If you’ve ever wondered whether your garden is trying to tell you something, I’d encourage you to write a few things down. You may find it’s been talking all along.