The Trends dashboard is your historical view. It answers one question across every time horizon: "How has this outlet's performance moved over weeks, months, years, days, or by weekday?"

Open it when you want to see direction and rhythm, not the live state of today. For real-time pace use the Today dashboard; Trends is for the long view.

Every chart on Trends is split by timeframe — the parts of the day you have configured for the outlet (for example Breakfast 6–11am, Lunch 11am–4pm, Dinner 4pm onwards). Each order is attributed to one timeframe based on the time the order was placed in Bangkok time and the timeframes you defined on the outlet.

If you have not configured timeframes on the outlet, every order is grouped under Total. If an outlet has timeframes set, the top 9 by revenue are kept and anything else falls into Other. There is also a flat Overall avg line drawn on most charts as a reference — that is the average of the whole-outlet weekly total across the entire range, regardless of timeframe.

At the top of every tab there is a timeframe picker table (click a row to filter the rest of the tab to that timeframe; click again to clear). The picker uses the same attribution logic the charts use.

The four "time-bucket" tabs (Weekly, Monthly, Yearly, Daily) have the same seven charts in the same order. The Weekday tab has the same seven metrics, just collapsed to day-of-week.

Stacked bars: one bar per week, segmented by timeframe. The flat line on top is the Overall avg — the average weekly total across the whole range.

Reading it: tall bars above the line are strong weeks; short bars below are quiet ones. If a single timeframe segment is growing or shrinking faster than the others, that is where the movement is coming from.

How it's calculated: closed orders only (status = 'complete'). Each order's revenue is its total field, summed by the week of reportingDate. Bangkok time.

Same shape as Weekly Revenue, but counting orders instead of revenue. One bar per week, stacked by timeframe.

Reading it: compare to the Revenue chart above. If revenue is climbing but orders are flat, your average ticket is going up. If orders are climbing but revenue is flat, you are selling more, smaller orders.

How it's calculated: count of closed orders (one per cheque) by week of reportingDate.

Number of guests served per week, restricted to dine-in orders. The guest count is whatever your staff entered on the order at the POS.

How it's calculated: sum of guestCount on closed dine-in orders (channelType = 'dine-in'). Takeaway, delivery, and partner orders are excluded — they don't have a meaningful guest count.

Average order value per week, drawn as a line. One line per timeframe, plus an overall reference.

How it's calculated: revenue ÷ orders, per timeframe per week. Closed orders only.

Reading it: a rising AOV means each cheque is getting bigger — either guests are buying more items or the menu mix has shifted to higher-priced items. A falling AOV usually means more cheaper orders (often takeaway or partner).

Average number of items on a cheque per week.

How it's calculated: total confirmed items (order_items.status = 'confirmed' — cancelled items excluded) ÷ number of closed orders, per timeframe per week.

How much each individual guest spent on average, dine-in only.

How it's calculated: dine-in revenue ÷ sum of guestCount, per timeframe per week. Cheques where staff did not enter a guest count don't contribute to the denominator.

Average number of items consumed per guest. Dine-in only.

How it's calculated: confirmed items on dine-in cheques ÷ total guest count, per timeframe per week.

Same seven charts as Weekly, grouped by calendar month instead of week. Read them the same way — they smooth out week-to-week noise and show seasonal direction.

Same seven charts, one bar per calendar year. Most useful for outlets with two or more years of trading.

One bar per business day. Use this tab when you want to spot specific outlier days — a public holiday, a private event, a closure, a service failure.

The same seven metrics, collapsed to day-of-week. Every Monday in the range is averaged together, every Tuesday together, and so on. This is the only tab where the bars are not a date — they are Monday through Sunday.

Use this when you are planning staffing, deciding which day to close on, or asking "is Wednesday always this quiet, or just this week?"

Average revenue per Monday, per Tuesday, … per Sunday — computed across all weeks in the selected date range.

How it's calculated: total revenue per weekday ÷ number of that weekday in the range. So if your range covers eight Tuesdays and total Tuesday revenue is 800,000, this chart shows 100,000.

The dashboard has three filters at the top of every tab: