FastStats Beta Audit Guide

Within the new 2018 FastStats system we have created many visualisations. These are to enable you, to audit your data in more detail. These objects produce a simple population audit. Along with the statistics, measures and KPI’s used within the Country Pack analysis. This includes:
  • High-level income trends
  • Acquisition trends
  • Conversion rates
  • Regular Giving Upgrade & Downgrade rates
  • Retention rates
  • Lifetime Value calculations.
Within the “Standard Templates” folder within FastStats, there are two key folders.
My Discoverer Folder / Public / Standard Templates
These folders are:
  • Data Quality Audit
  • Power BI Replication

Data Quality Audit

In this area we have created visualisations to explore the population of your data. These are:
  • Person Level Validation
  • Person Level Derived Fields Validation
  • Gift Level Validation
  • Regular Giving Pledge Level Validation
These visualisation generally provide a cross-tab breakdown of the supplied data. The exception is the “Person Level Derived Fields Validation”. This provides a cross-tab breakdown of fields derived by Adroit at the person level. This includes fields which define a donors Active Status at certain time points. It also shows Donors Status, i.e. is the Donor a Major Giver, Mid-Value Donor, Individual Giver or Micro Donor.
These objects are structured as a workbook. With a single base selection, with several “Tabs” symbolising the different cross-tabs. Run this workbook by pressing the green play button. This will calculate the figures within every tab. Scroll through the tabs, checking the figures match what your expectations.

Selection Workbook

The cubes are structured with a single field denoting the Country/Organisation on the vertical axis. The horizontal axis donates the field to with is to be validated. Note that these audit tables will allow you to see data for all organisations you are entitled to see. This enables peer reviewing of the datasets.

Example Cube – Data has been randomised

Power BI Replication

This folder focuses on providing KPI’s as a data audit tool. From past experience in delivering analysis projects, we have found initial data audits don’t spot all issues. They spot the majority of issues, but not everyone of them. Indeed, it is often the case that only once analysis is produced, do data quality issues arise. As a result, we are providing the objects which produce the KPI’s early in the process. The aim is to ensure that you can spot data issues or raise queries before the final analysis is delivered.
Within the Power BI Replication folder, under the “Country Report” folder. There are several folder areas denoting common fundraising areas. These are:
  • Acquisition
  • Cross-Sell
  • Life-Time Value
  • Overall
  • Reactivation
  • Retention
  • Upgrade.
In each of these folders there may be extra subfolders. These often split out between Regular Giving and Single Giving areas. But ultimately within each area there will be several FastStats objects. These objects replicate the figures from the Power BI Country Report.
For each object you will need to update the selection by pressing the green “Play” button. You will be prompted to select a particular country from a check-box list. Please selected the country for the market you are validating. This will update the selection, using only data from this country. The cube itself will show information for your organisation. It will also show the KPI’s for peer organisations within your market.

Select Country Prompt

Example Overall KPI’s for Auditing

These FastStats cubes are different to those described in the Data Quality Audit section. In that there are usually several statistics or measure show. The Data Quality Audit measure are all a straight count of records (People, Gifts or Pledges). In this sections measure include counts, totals, averages and ratios.

Completion of Audit

Upon completion of the FastStats Beta Audit, please email to confirm that the data is correct, or to raise issues. If there are issues, please clearly explain within the email the issue, referencing the object in question. Screenshots will be appreciated if you are able to provide these.

Leave a comment