How a trend gets scored, and how it gets judged.
Two different things. The score is arithmetic you can check. The verdict is an operator's call, dated and kept on the record.
0 to 100, from six inputs
Recomputed nightly and whenever the editor adds a call or an article. Every trend page shows the inputs, not just the total.
10 points per separate forecaster naming the trend. A retailer's annual list, a trade-association panel and a research report each count once.
15 if the newest forecaster call is under 90 days old, 9 under 180, 5 under a year, 2 after that.
3 points per linked article in the last 90 days, capped at 15.
Google Trends (US) when available, otherwise Wikipedia daily pageviews for the trend's articles, last 90 days. Up to 9 points for the last four weeks against the eight before, up to 6 for the level of interest. The trend page says which source was used.
Reddit threads across the grocery and food subreddits in the last 30 days against the 30 before, plus new YouTube videos. Volume and growth each count.
Published products tagged to the trend: 1 point each at Scaling or Established, half a point at Emerging, capped at 5. Stage is distribution reach, with the evidence and date recorded. The weight rises as the product library grows.
Real, Niche or Fad
Never automatic. The verdict is made by an operator who has merchandised, priced and sold the category, carries two lines of reasoning and a date, and every past verdict is kept so the quarterly scorecard can show the record.
Will earn and hold shelf space across formats and price points. Worth building assortment, menus and launches around.
Durable, but for a defined shopper or channel. Stock it where that shopper is; don't bet the planogram on it.
Press and launches outrun repeat purchase. Expect the velocity to fall before the next reset.
On the radar. Enough forecasters or press to track, not yet enough product or sell-through to call.
Where the press is going
The sparkline is six months of linked press mentions. Rising means the last two months beat the two before; Steady means they matched; Cooling means they fell; Quiet means no recent coverage.
Press momentum is descriptive. Search and social are scored, but together they are a quarter of the total: a quiet trend with four forecasters behind it still outranks a loud one with one. When talk runs well ahead of forecaster and product support, the trend is tagged Talk ahead of shelf and surfaces on the Buzz Board.
Who we track, and who tags
Retailer forecasts (Whole Foods Market, Kroger), the Specialty Food Association's Trendspotter panel, Expo West and Fancy Food Show coverage, research houses such as Tastewise, and the trade press (Grocery Dive, BevNET, Nosh, FoodNavigator, Supermarket News, The Shelby Report).
Nothing publishes without a human tag. Trend calls and articles come in through an editorial desk and a weekly assisted scan; each call is credited to its forecaster with a source link.