Building dashboards and charts in insights

Insights is where you build your own charts out of the data already in your account, and arrange them on dashboards. It replaces the old reports section.

Go to Data → Insights. The page lists every dashboard in the account. View saved insights, top right, takes you to the library of individual charts, which you can put on as many dashboards as you like.

Dashboards and insights belong to the account rather than to you, so everyone with access to this part of the dashboard sees what you build, and your edits.

Starting a dashboard from a template

Click Add dashboard. The templates are grouped under Contacts, Orders and Transactions, and each one creates a dashboard with its cards already built, so you start by adjusting something rather than by staring at an empty grid.

• Contact Activity Data. Contact engagement by tier, transactions and recency.
• Contacts Data. Total contacts, signups over time and a breakdown by channel.
• Orders. Sales performance, order volume and revenue across channels and locations.
• Product performance. Best-selling products and add-ons by revenue, units and discount.
• Giftcard Transactions. Gift card transactions grouped by channel and by business profile.
• Giftcard Statistics. Amount added and spent, outstanding balance, transaction count and funds on expired gift cards.
• Points Transactions. Points activity by location and channel, redemptions and issued points.
• Voucher redemptions. How vouchers are redeemed across promotions and locations, and over time.
• Statistics. The loyalty figures: outstanding balance, average monthly spending, number of transactions, and points issued and redeemed.
• Financial report. Outstanding gift card, prepaid and loyalty point balances, and how they change over time.

Giftcard Statistics also asks which gift card program it should cover, if you run more than one. Pick a program and every card on the dashboard is filtered to it, with the program name added to the dashboard title. All giftcard programs leaves them combined.

Blank dashboard, under the template list, gives you an empty grid instead. Either way, click Create dashboard.

Building a chart

A chart is called an insight. Open the editor from Add card → Create insight on a dashboard, or from New insight on the saved insights page. The left panel is the question you're asking and the right panel is the answer, which re-runs on its own each time you change something.

Data. Pick the source the numbers come from: Contacts, Points Transaction, Giftcard transactions, Orders, Order line items, Order sub line items, Voucher redemptions, Bookings, Visits, Subscription changes, Form submissions or Issued wallet passes. The source decides which fields everything below it offers, so changing it clears the rest of the panel. Contacts offers your own contact attributes alongside the standard fields.

Form submissions has a row for every time one of your forms was filled in, also when an existing contact filled it in. That makes it the source for how many people signed up through a form. A count of contacts by channel only counts the people a form created, so it comes out lower. Group by Form to compare your forms, or by Source to compare the links you shared with a source added. Status tells Completed submissions apart from Pending approval and Rejected ones, which only happen on a form that needs approval.

Issued wallet passes has a row for every wallet pass a contact asked for, one per wallet: a contact with your loyalty pass in both Apple Wallet and Google Wallet has 2 rows. To count contacts rather than passes, set Measure to Number of distinct values of… and pick Contact, and filter Type on Loyalty card for your loyalty pass. Group by Platform to compare Apple Wallet with Google Wallet, by Wallet Pass to compare your pass designs, or by Created at to see how many passes were handed out each month. A pass counts from the moment the contact asked for it, also before they added it to their phone, and removing it from a phone doesn't always take it off the count. So read these figures as passes handed out, not passes on phones right now.

Filter. Click Add a filter and pick a field. The operators follow the kind of field: is, is not, contains, is empty and is not empty for text, Greater than, Less than and between for numbers, and on, before, after, between, in the last and in the next for dates. A window set with in the last keeps rolling, so an insight saved with "in the last 30 days" shows the 30 days before the day you open it, not the 30 days before the day you saved it.

On Order line items and Order sub line items you can also filter on Contact, the contact the order belongs to, to see what one contact bought and how often. It takes the contact's number, the one at the end of the address when you open the contact's page. Contact is a filter only on these sources: it isn't offered under Measure or Group by.

Summarize. Measure is the number the chart plots: Count of rows, Number of distinct values of…, Sum of…, Average of…, Median of…, Minimum of…, Maximum of… or Ratio. Everything except Count of rows then asks which field to measure. Ratio divides one measure by another and each side takes its own filters, which is how you build a share, such as revenue from one channel over revenue from all of them.

Group by is optional, and it's what turns a single figure into a chart. Group on a date field and you also choose Day, Week, Month, Quarter or Year. Group on a numeric field and you choose how it's bucketed: Auto bin, 10 intervals, 50 intervals, 100 intervals or Don't bin. You can group by more than one field. With a single grouping, a Cumulative switch appears, which adds each bucket to the ones before it for a running total.

On the right, choose how it's drawn: Bar, Line, Area, Pie, Number or Table. Number is the one for a single figure such as a total, and it's what most tiles on the template dashboards use. Next to the chart types is the colour palette. The header counts the rows and totals the values, Show data table opens the rows underneath the chart, and Re-run query fetches the data again.

Give the insight a name in the title at the top, then click Save.

Arranging a dashboard

Click Edit dashboard. An Editing layout badge appears and the cards become movable: drag a card to move it, drag its bottom corner to resize it.

Add card holds three things:

• Create insight opens the editor on a new chart, which is added to this dashboard when you save it.
• Add section heading puts a full-width title across the grid, which you type over, to group the cards underneath it.
• Add saved insights opens a panel of the charts you've already saved, so the same chart can go on a second dashboard.

Click Save to keep the layout and the dashboard title. Cancel puts the layout back the way it was. Removing a card is the exception: that happens straight away, and Cancel won't bring it back.

Outside edit mode, hovering over a card shows a download icon and Open in editor.

Saved insights, and reusing one

Every chart you save lands in Saved insights, whether you built it on a dashboard or in the library. The list shows each insight's source and the number of dashboards it's on.

An insight is one object however many dashboards show it, so editing it changes all of them. That's what makes the library useful, and it's also the thing to watch: when you want a variant, build a second insight instead of editing the shared one.

Removing a card from a dashboard only takes it off that dashboard. Deleting the insight, with the bin icon in the library or in the editor, removes it everywhere, and the confirmation tells you how many dashboards that affects. Section headings belong to the dashboard they sit on, so removing one deletes it. Deleting a dashboard leaves its insights in the library.

Exporting to CSV

The download icon exports what you're looking at: in the editor above the chart, or on a card when you hover over it.

The file has a column per grouping and a column for the value, with the currency in the header when the measure is money. It re-runs the query rather than copying the chart, so it isn't limited to the rows that were drawn, but a single export stops at 50,000 rows. If your file stops at 50,000 rows, it was cut off: narrow the filters and export again. When nothing matches the filters, no file is made and you're told instead.

Things worth knowing

Filters belong to the insight, not to the dashboard. There's no dashboard-wide date range, so the period is set on each chart.

A template's gift card program choice is applied once, when the dashboard is created. Changing it later means editing the filter on each card.

The editor runs your query as you build it, and nothing is stored until you click Save, so it's safe to experiment on an existing insight as long as you leave without saving.