1. Analysis built into the system, not bolted on at year end
Organisations collect data diligently and then analyse it rarely. When analysis happens, it is a manual exercise that one person knows how to repeat. The better approach is to decide the questions up front and let the system calculate the answers every time new data arrives.
On screen: our design, with invented sample data.
2. From a yearly exercise to a live picture
What changes
Today
- Analysis done once a year by one person
- Figures that differ between reports
- Definitions kept in someone's head
With the system
- Indicators recalculated as data arrives
- One definition per indicator, written down
- Every figure traceable to its source records
3. The questions, answered continuously
What the system can calculate
Descriptive summaries
Counts, proportions, rates and averages, with denominators shown.
Targets and trends
Progress against targets and change over time.
Disaggregation
By sex, age, location, site or any captured grouping.
Maps
Results by area where locations are recorded.
Data-quality flags
Missing values, outliers and inconsistencies highlighted for review.
Automated reports
Scheduled summaries sent to the people who need them.
4. How we treat numbers
Method, stated plainly
Do you offer forecasting?
Only where the data supports it and the method can be defended. Most organisations are better served first by accurate descriptive analysis.
Can we see how a figure was calculated?
Yes. Indicator definitions, numerators and denominators are documented in the system.
Can analytics work with our existing data?
Usually. Spreadsheets, databases and exports from survey tools can be brought in and cleaned.
5. Users, connections and data
Who uses it
- Data officersClean, review and publish data
- ManagersDashboards and scheduled reports
- Board and fundersSummary views, read only
Connects to
- Existing databases
- Excel and CSV
- KoboToolbox and ODK
- DHIS2 where authorised
- Scheduled PDF and spreadsheet exports
Data sources are connected with the owner's permission and credentials.
Data and security
- Small groups suppressed in public views to protect individuals
- Access to record-level data limited by role
- Every published figure traceable to its source data
Talk to us about data, analytics and dashboards
Tell us how the work runs today. The first conversation is about understanding it, not selling a package.
Discuss an Analytics Solution