The Channels dashboard tells you where your orders are coming from — dine-in, takeaway, direct delivery, partner platforms (Grab, LINE MAN, Foodpanda, ShopeeFood, Robinhood, etc.), and co-located venues — and how each of those channels is performing on revenue, orders, and average ticket.

Open it when you want to understand channel mix, see whether partner platforms are growing or eating margin, or compare what items sell on which channels.

Throughout every chart the channels are labelled by a single canonical name derived from channelType and channelName:

Wherever the order has no channel information at all, it shows up as "Unknown". This should be very rare in normal data.

Share of revenue from each channel as a pie. The whole pie is total revenue across the selected date range; each slice is one channel.

How it's calculated: sum of orders.total on closed orders (status = 'complete'), grouped by the canonical channel name.

The same pie, but counting orders instead of revenue. Compare the two — a channel can be a big chunk of orders but a small chunk of revenue (or vice versa) if its tickets are unusually small or big.

Average order value per channel — bars side-by-side so you can compare ticket sizes across channels.

How it's calculated: revenue ÷ orders per channel.

A combo chart: bars are revenue per channel (sorted descending), the line on top is the cumulative share ("Cumulative %") of total sales as you walk across channels from biggest to smallest. The right-edge of the line is always 100%.

Reading it: the point on the cumulative line that crosses 80% tells you how many channels make up 80% of your revenue. If two channels get you there, you are highly concentrated.

The same chart, by order count rather than revenue.

Stacked bars: one bar per week, segmented by channel. Week is Monday-to-Sunday in Bangkok time.

Same shape, counting orders.

One line per channel showing how each channel's average ticket has moved over time.

A pivot table with one row per week and one column per channel, showing the underlying numbers behind the charts above. Use it when you need an exact value for one cell.

Same four charts as Weekly, grouped by calendar month.

Same four charts, grouped by calendar year. Most informative for outlets with two or more years of data.

Same four charts, one bar per business day. Useful for spotting outlier days — a partner platform outage, a big private dine-in event, an unusual delivery surge.

This tab restricts every chart to partner orders only — channelType = 'partner'. Use it when you want to compare partner platforms head-to-head (Grab vs LINE MAN vs Foodpanda etc.) without dine-in or takeaway in the picture.

Pie of revenue across partner platforms. Each slice is one partner.

Average ticket per partner platform.

Weekly stacked bars across partners only. Lets you see one partner growing while another shrinks.

For each partner, two bars side by side: the revenue you keep, and the commission paid out to the platform. Together they sum to the gross sales on that partner.

How it's calculated: we apply the commissionRate stored on the partner's channel record (in channels.partnerInfo.commissionRate) to each order's gross. So Commission paid = gross × commissionRate, Revenue you keep = gross × (1 − commissionRate). If a partner channel has no commission rate configured, the rate is treated as 0 (so all revenue is "kept" on the chart, which is usually not what you want — make sure the partner's commission rate is set in Channels.)

The 20 best-selling items by revenue across the whole range, with each item's bar segmented to show what share came from each channel.

Reading it: a stalwart dine-in item suddenly showing a big partner segment means the dish travels well and could be promoted on delivery. An item with no dine-in segment but heavy partner share is a delivery-only winner.

How it's calculated: sum of order_items.itemTotal on confirmed items (order_items.status = 'confirmed') in closed orders, then take the top 20 items by total revenue, and split each by canonical channel.

The same top-20 items as the chart above, laid out as a table: one row per item, one column per channel, revenue in each cell. Columns appear automatically for every channel you use — including custom and partner channels — so it adapts as you add channels. Blank cells mean the item never sold on that channel. Money is on the item-subtotal basis (before discounts, service charge and tax).

A 168-bucket view (7 days × 24 hours) of order volume, with each hour-of-week stacked by channel. The labels along the bottom run Mon 00:00, Mon 01:00, …, through Sun 23:00.

Reading it: the natural rhythm of your week. Dine-in peaks at lunch and dinner; partner platforms often have a different shape (lunch heavy, late-night spike). If one channel suddenly fills an hour where it didn't before, that is a structural change worth investigating.

How it's calculated: count of closed orders, bucketed by EXTRACT(ISODOW FROM reportingDate) and EXTRACT(HOUR FROM orderDate) in Bangkok time, then split by canonical channel.

Two filters at the top of every tab: