A tour of Analytics

Bookings, revenue, customers and where they came from — over any period.

Analytics answers one question: how is the business doing over a stretch of time? Pick a period at the top and everything below refills from it. Here is the screen top to bottom, and the figures people misread.

Set the period first

Open Booked › Analytics. It always opens on Last 30 days — the range is not remembered, so check the button before you trust a number.

• 1The date range button — the period in force.
• 2Compare — turns solid when it is on.
• 3Compared with … — the window you are measured against.
• 4Ask AI: how do I grow? — appears only when your licence is active.
• 5Bookings — orders taken in the range. The hint beneath, "6 today", ignores the range.
• 6Repeat rate — all-time, so its chip reads "± 0%" for ever.

The button opens this popover.

• 1Presets — clicking one applies it. Weeks start on Monday. They count days, not calendar months: Last 30 days includes today, Last 12 months is 365 days. Use This month or Last month for calendar periods.
• 2Custom — clears the selection and keeps the popover open. Click a start date, then an end; an earlier second click flips them.
• 3Future dates are greyed out. Analytics only looks backwards — for what is to come, use the Calendar or Upcoming bookings.
• 4Apply — disabled until a start date is chosen. Escape or Cancel closes it unchanged.

Reading a comparison

Compare measures your range against the same number of days immediately before it — days, not calendar months. Pick This month on the 26th and you are compared with the previous 26 days, which straddle two months.

• Chips appear on six tiles only — Bookings, Expected revenue, Taken revenue, Customers, Repeat rate and Cancellation rate.
• Green means the movement is good. Cancellation rate is inverted, so a fall shows green.
• 100% means the previous period was zero; no chip means both were zero.

What each tile actually counts

Six tiles follow your range. Three ignore it, with nothing on screen to say so.

Two need care. Cancellation rate divides by every booking made in the range, cancelled ones included, while Bookings leaves them out — so it will not divide into the headline count. Lifetime value counts everyone who ever booked, so it reads low if you do a lot of one-off trade.

The two trend charts

Both charts use the same daily buckets — one bar per day, counted by the day the booking was made. Revenue here is expected value; there is no chart of money taken.

• 1The range total — the only whole number the chart shows.
• 2Hover a bar for "2026-08-14: 7". There is no scale up the side; heights are relative to the busiest day.
• 3Three labels only — first, middle and last, in MM-DD form; the middle goes on ranges of four days or fewer.

Services, staff and where bookings came from

Top services and Staff rank by booking count, top eight only. Booking status is the one panel that includes cancelled bookings. Most valuable customers ignores the date range.

• 1This number is a count, not money — on Top services and Staff alike. Bookings with no staff attached show as "Unassigned".
• 2The info circle — hover it for how to tag links, for example ?utm_source=facebook&utm_campaign=spring.
• 3Direct / none — everything Booked could not attribute. It exists only in Source, so the three columns never total the same.
• 4Empty states — Medium and Campaign list tagged traffic only.
• 5Here the number is money — lifetime booked value, all-time, in the style of the panels above.

Attribution is first-touch, kept in the customer's browser for 30 days, and captured on the page carrying the booking widget. So:

• If your ad lands people on a page without the booking widget, the tags are gone before Booked sees them. Private browsing, or clicking on a phone and booking on a laptop, loses it too.
• Bookings staff create in the admin, those from the compact form, and course or bundle follow-ups carry no tags — all land in "Direct / none".
• Casing matters: utm_source=facebook and utm_source=Facebook count as two sources. Underscores show as spaces, so spring_sale reads "Spring sale".
• Every list is capped silently: 10 each for sources, mediums and campaigns, 8 services, 8 staff, 6 customers.

This is not Settings › Tracking & pixels, which reports bookings out to pixels you run. The two rarely match.

Ask AI: how do I grow?

Analyse my numbers returns a breakdown headed Your AI breakdown: a summary, a "Do this first" highlight, a "What's working" list and an "Opportunities" list, each with an action. It reads the selected range only.

It sees aggregates only — the period, the KPIs, top services, staff names, the status mix and traffic sources. Customer names are never sent. You get one run per site per week, so set the period first: changing the range wipes the breakdown, and re-running spends that week's run.

Choosing a period that answers your question

1. Judging a campaignSet its dates with Custom, then read Marketing attribution.
2. Month against monthNote the totals on Last month, then switch to This month.
3. Checking a quiet weekLast 7 days with Compare on — seven days against seven.
4. Looking at the yearLast 12 months. Keep custom ranges under about 400 days, or the charts and the tiles disagree.