Product Guides
Create custom KPIs and collect values from your organization
Custom KPIs let you define your own indicators — water use, waste, certifications, anything — and collect their values from data owners across your organization through periodic campaigns. This guide shows how to create a KPI group and define the indicators it collects.
Before you start
You need an Editor role in your Dcycle organization.
Steps
Step 1
Click “Data” in the left navigation.

Step 2
Open “Dcycle categories”.

Step 3
Find the “Custom KPIs” card among the data categories and click it.

Step 4
The Custom KPIs page lists your KPI groups. A KPI group bundles related indicators (for example “Water consumption per facility”) together with the campaigns that collect their values. Click “Create” to add a new one. If you don’t have any group yet, click “Create your first KPI group” instead.

Step 5
Give the group a “Name” (e.g. “Water consumption per facility”) and an optional “Description”. Then define each indicator under “KPI 1”, “KPI 2”…: its name, the “Response type”, an optional unit and description, and whether a blank response is allowed. Use “Add KPI” for each extra indicator.

Step 6
Choose the response type: Number, Text, Percentage, Date, Yes / No, Dropdown or Number by category.

Step 7
If an indicator breaks down into several types (for example drinking, sanitary or reused water), use “Number by category” instead of creating one KPI per type. Click “Add category” for each breakdown. The unit is the same for all categories, and the KPI value is the sum of all of them. If the KPI doesn’t allow a blank response, the data owner must fill in at least one category.

Step 8
When you’re done, click “Create KPI group”. You’ll land on the group’s page, where you can edit the KPIs later from the “…” menu at the top right.

Step 9
Next, collect the values through campaigns: see “Launch a custom KPI campaign and collect values from data owners”.
Was this helpful?
What was missing or unclear? (optional)
Sent. Thank you for helping us improve.