Duplicate record rate
The most visible sign of a messy database and the easiest to track over time.
Ask a development director how many active donors they have and the answer usually comes with a caveat. Event registrations from Eventbrite, online gifts from Givebutter, a list imported from the gala five years ago and manual entry by three different staff members have left duplicates, split households and gifts coded to campaigns that no longer exist.
We would clean the database in a controlled way: agree the rules for donors, households, soft credits and gift coding, merge duplicates in reviewed batches, and fix the integrations that keep creating new mess. The goal is a LYBUNT report, a year-end acknowledgment run and a board dashboard that all match without a spreadsheet on the side.
Decide with development and finance how donors, households, organizations, soft credits, tributes and gift codes should be recorded, and write it down as a short data guide.
Run matching on names, emails, addresses and phone numbers, then have a staff member review suggested merges in batches before anything is combined.
Map old campaign, appeal and fund codes to a smaller current list, with finance confirming that restricted funds and grant codes still reconcile to the general ledger.
Adjust how online giving, event and email tools sync to the CRM so new records are matched rather than duplicated, and set required fields on manual entry forms.
A short scheduled report flags new duplicates, missing addresses and uncoded gifts, so the database stays clean after the project ends.
How we deliver this: AI automation for small business or fractional CTO leadership, scoped in writing and run over 90 days. See pricing or the sample scope.
The most visible sign of a messy database and the easiest to track over time.
Shows whether development and finance finally agree on totals.
Bad contact data costs postage and loses donors quietly.
Clean data should shorten the busiest reporting period of the year.
For example, an illustrative food bank with about 9,000 donor records plans its December appeal and Colorado Gives Day push. A first matching pass suggests roughly 900 likely duplicates and 1,400 spouses who should be householded. The database manager reviews merges in batches over two weeks. The appeal goes to fewer, correct addresses, giving histories finally show true lifetime value, and the major gifts officer retires her private spreadsheet.
In the first 30 days we would export a snapshot of the database, measure duplicates and coding problems, interview development and finance about how they use the data, and agree written data rules. No records would change until you approve the plan.
You should, and we would start there. Bloomerang, DonorPerfect, Raiser's Edge and Salesforce all have some duplicate management. Built-in tools find obvious matches but rarely decide your household rules, clean gift coding or fix the integrations creating new duplicates. Those decisions need someone to lead them with development and finance.
For a database of a few thousand to tens of thousands of records, the main cleanup usually fits within the 90-day engagement. Duplicate review is often the slowest part because a person checks each batch. We would schedule it around your campaign calendar so nothing changes during a year-end appeal.
Sometimes a cleanup reveals that the current platform is a poor fit, but migrating dirty data just moves the problem. We would clean first in either case. If a switch makes sense, a clean database makes migration faster, cheaper and less risky.
Illustrative defaults for a nonprofit. Replace them with your numbers.
a month in staff time, before any lost sales
Illustration only, not a forecast or guarantee. Recovered staff time becomes a cash saving only if spending actually declines or the time goes into paid work. Tool costs are an assumption; we replace every number here with yours before recommending a scope.
CRM setup and cleanup: full guide
In 30 minutes we look at one recurring workflow, what it costs you today and whether there is a credible case to improve it. If there is not, we will say so.
30 minutes with Phil or Martin. No technical brief, no pitch deck. Phil or Martin replies personally, usually within one business day.