Three PhD students used AI to guess where porcini grow: media gave it a boost, 7,500 installs

Three PhD students at the University of Genoa built a map predicting "which woods to search for mushrooms today," completely free. At the end of October, Italian food outlet Gambero Rosso wrote about it, food accounts kept reposting, and Google Play hit two thousand installs in just over three days. But laying the numbers out, this is a seasonal tool that got a push from one wave of coverage and never really scaled. What's worth watching isn't the size, it's the approach: public data plus a model, then daily auto-generated regional briefs to catch search traffic.

A person who grew up mushroom hunting with their grandparents

In 2026 Oct 27, Gambero Rosso ran a piece that opened like this: Nicolò Badino, as a boy, went up the hills with his grandparents to find mushrooms, learning to read the woods and the ground, and knowing which spots were worth stopping at. After the elders passed, this handed-down craft died out in his family. Now he is a PhD student in environmental and geographic economics at the University of Genoa, and decided to bring it back another way: with AI.

He and classmates Manuel Delfino and Diletta Migliaccio built MyBoletus. The principle is not complicated: stack rainfall, temperature, soil moisture, tree species, altitude and slope aspect, compute a "suitability" score for each patch of woods, green to try, red to skip. It shows by species, with porcini in four types, plus chanterelles and orange-cap amanitas. Beyond the map there's GPS trip logging, offline maps, route scoring per kilometer, and a community: people can log their hauls and earn points on a leaderboard, but the exact spots are not public. In the Italian mushroom-hunting world, spots are family secrets, and "we give scores, not locations" fits the temperament of this crowd.

The web version came in June, Google Play on Aug 17, iOS on Oct 15. Fully free, no in-app purchases.

The numbers aren't that impressive

Source: Google Play install counts, viewed daily from 10-29; 9–October figures are third-party estimates.

It looks like "3 Nov it jumped 2,076 in a single day," but it isn't. Google Play install counts update in batches every few days, and similar apps also jumped on 10-31 and 11-03, so these 2,076 are three-day totals. More important is the ranking: around Oct 30 it reached #12 on the category chart, then kept sliding, and by Nov 6 it was #97. The peak has passed.

Too few ratings on Play for the store to display one; the iOS Italy store has only 2 ratings. It has only appeared on Italy's chart, one country. The developers are active though: four releases in one week, 10-27 to 11-03.

Why now: foraging season, plus a good story

Timing works on two levels.

First, the season. Italy's porcini season runs roughly August to November, so an Aug 17 launch landed right at the start. Late September it added around a hundred installs a day (estimated), and by October it fell to dozens a day, which is what a seasonal tool looks like.

Second, the media. This story was practically made for local papers and food outlets: a university PhD student, AI, a lost grandparents' craft, secret spots. Genoa local paper Il Secolo XIX wrote about it in Jun 14 (back when only the web version existed; "a hundred-odd signups in two months"), then again in Sep 1 ("community signups over two thousand"). Coverage clustered from Oct 21: il T Quotidiano posted on Instagram, Gambero Rosso ran it on Oct 27, food accounts ricettemania and LOVOJO reposted, and AllFoodSicily wrote about it on Nov 2. The climb to #12 on the category chart and the install jump all fall in this stretch.

One detail: the Gambero Rosso October piece still cited "a hundred-odd signups in two months," using June's old figure. You can tell these articles are copying one another, not reporters who actually tested the product.

Nobody discussed it on X, Reddit or Hacker News. Instagram has 815 followers, the Facebook page 54, and no official TikTok account turned up. No evidence of ad spend turned up; judging by its scale and being fully free, it most likely didn't spend on ads (inference).

The one move worth copying: two auto-written "where to find mushrooms today" posts a day

A media wave fades, but a website system doesn't. The myboletus.com sitemap spreads across 40 "country × language" combinations. The Italian-language site alone has about 170 pages, of which 92 are blog posts auto-generated by region, published every day at 5 a.m. and 5 p.m., with titles like "Porcini in Liguria 6 novembre: dove cercare nei boschi migliori" (Porcini in Liguria, 6 November: where to search in the best woods).

Take the 6 November post: the first line of the body reads "model data as of 2026 Nov 6," followed by: average rainfall over the past two weeks 22 mm, mean temperature 12°C, common porcini suitability in the best areas at 49/100, up from 1 on Nov 38; summer porcini reached 100/100 in "high-suitability zone 1," with a recommendation to try chestnut woods on north-facing slopes at 200 to 540 meters.

These pages target the long-tail search of "funghi + region + date." Foragers search exactly this before heading out, and these pages already show up in results. One model and one template: swap the region, the date or the language, and you have a new page. There are also 28 tree-species pages, 21 regional pages and 13 mushroom-species pages.

Who it's competing with

This niche is more crowded than it looks. Almost every big mushroom-hunting country already has someone charging for it:

The web also has FcP Map, backed by the large Facebook group "Porcini che passione," at €2.99 a month or €24.99 a year; FungiFind is a free map covering all of Europe, with a Plus subscription adding alerts and forecasts, nearly identical in approach to MyBoletus. In the year 2026 alone, Italy saw a batch of "AI porcini prediction" apps emerge: FunghiTrack, MushScout, ShroomCast, FungApp, Control Porcino, mostly small zero-rated apps.

So two points need to be clear. First, among Italian peers MyBoletus is fairly small in scale, and its product moat is shallow: features are much the same, and there's no reputation to prove the model is accurate. Second, this category does make money: Sweden, the UK and France all have people charging, but the ones charging are long-time local players in each country.

If it were you

C. Worth learning from, but copying it for Italian mushroom hunting isn't worth it. Italy is already saturated; MyBoletus itself only reached 7,500 installs on media coverage and is sliding. What's worth taking is the approach: a seasonal hobby, plus public weather and forestry data, computed into a "where to go today" map, with daily auto-written regional briefs to catch search traffic.

Who it suits: a team of 1–2 people, with someone who can handle geographic and weather data (rasters, weather APIs, forestry layers) and someone who can do programmatic SEO; ideally someone who plays this hobby themselves, or a country where the hobby is popular in their native language but no long-time local player is charging yet.

Who it doesn't suit: anyone wanting to fight head-on in Italy, the UK, France or Sweden, where long-time local players and a crowd of new products already exist; anyone wanting year-round revenue, since this kind of product only has two or three months of users a year; anyone unable to produce forecasts that experienced foragers respect, since this community runs on reputation, and an inaccurate model leaves you with a single download.

How to do it: pick the country first, then the hobby. Mushroom hunting is a nationwide pastime in Central and Eastern Europe; Poland, the Czech Republic and German-speaking regions are worth checking first for local paid tools (whether these exist is unverified). Or switch to a different hobby that similarly depends on weather, place and season: sea fishing, stream fishing, cherry blossom and autumn foliage spotting, birdwatching. Start with a web product; a map with a suitability index is enough. Put GPS tracks, offline maps and a community leaderboard (points only, no public locations) later. For pricing, use FcP Map and Sporecast as references: monthly subscriptions during the season, or a one-time annual purchase.

How to acquire users: the main channel is programmatic SEO, one daily "where to find it today" post per region, copying its title format "species + region + date + where to search." Supplement with local hobbyist Facebook groups and forums, then pitch a persona-driven story to local papers and niche media. "University student / AI / lost craft" is a story MyBoletus has already proven media will run.

Deadline: using mushroom hunting as the example, you need to launch before the start of the 2027 August season. Regional briefs on the web should start two to three months earlier so search engines can index them first. Miss the season's start and you wait another year.

How we could be wrong: ① MyBoletus adds a subscription and next mid-season makes the top 30 of Italy's maps bestseller chart, which would show the Italian market can still earn money and we underrated it; ② conversely, paid long-time local players like Hitta Skog, Sporecast and Champi drop off their countries' bestseller charts mid-season next year, showing the business can't support a subscription; ③ a major outdoor app (such as AllTrails or Komoot) adds a "mushroom prediction" layer, leaving no room for a small product built around this one thing.

---

Data and sources: Google Play installs and the Italy category chart (10-29 to 11-06 daily); 9–October installs are third-party estimates; App Store Italy ratings (11-06); coverage by Gambero Rosso 2026-10-27, Il Secolo XIX 2026-06-14 and 09-01, AllFoodSicily 2026-11-02; the myboletus.com sitemap and the Liguria briefs at 11-06; the FcP Map, FungiFind and FunghiTrack official sites. Places marked "estimated" or "inference" are estimates and inferences.