How long should the monthly routine take?

For a small charity (under 20,000 records), 2 to 4 hours per month is realistic once the routine is established. The first month is heavier as you build the queries; from month two it becomes much faster.

Should we use the CRM out-of-the-box duplicate tool?

Use it for a first pass, but expect it to miss the trickier cases (subtle typos, hyphenated surnames, married name changes). A second pass using a fuzzy-match approach catches most of the rest.

How often should we re-check email deliverability?

Monthly at least. Bounces, complaints and unsubscribes are signals that need addressing within days, not quarters. Letting them accumulate damages sender reputation across all future campaigns.

CRM Data Quality: The Monthly Routine for Charities

Charity CRM data quality is not a one-off project; it is a monthly routine. A practical, repeatable checklist for keeping supporter records clean, consents current and reporting trustworthy, designed for a single data lead in a small or medium charity team.

CRM data quality is the work that nobody asks for, until something breaks. A duplicate causes a major donor to receive a cold welcome pack. A bad email file blows up sender reputation before a major appeal. A gap in consent data triggers an ICO complaint. None of it is glamorous; all of it is preventable.

The most reliable defence is a monthly routine: a short, repeatable sequence run by the same person on the same day each month. What follows is a practical version of that routine, designed for a single data lead in a small or medium UK charity.

Why monthly, not quarterly

Quarterly clean-ups produce three problems at once: the dataset has drifted further, the work is bigger, and small issues have already affected supporters. Monthly clean-ups keep the work small enough to fit in a morning and catch most issues before they hit production.

The seven-part routine

1. Duplicate scan

Run the CRM duplicate report (or your fuzzy-match query) for the records added or amended in the last 30 days. Merge or flag suspected duplicates. Aim for under an hour. Build the query once; reuse forever.

Check that every record added or amended this month has a current consent status, channel preferences and a source captured. The ICO expects this; future you, drafting an appeal, will thank you.

3. Email deliverability check

Review bounces and complaints from the last 30 days. Hard bounces: remove or quarantine. Complaints: investigate the campaign and the supporter. Soft bounces: retain but flag if persistent. Take ten minutes; it pays off across the next campaign.

4. Address quality

For records used in postal campaigns or for gift aid claims, check address completeness on records added this month. Use a postcode lookup tool. Charities still posting to incomplete addresses lose more income than is comfortable to admit.

5. Gift aid declaration coverage

For donor records added or with new donations this month, check gift aid declaration coverage. Identify donors with eligible giving but missing declarations and queue a gift aid declaration ask. This single check often pays for the entire routine.

6. Critical field completeness

A small set of fields you have decided matter most (e.g. source, primary contact, segment). Run a completeness report against records added this month and chase or auto-default the missing values.

7. Reporting integrity

Open the headline finance and fundraising dashboards. Sense-check the numbers against the source systems (bank feed, payment processor, fundraising platforms). Differences of more than a small tolerance should be investigated and resolved before the month closes.

Tooling that helps

Saved queries

Save each step as a CRM report or query. Reuse it month after month. The point is not to be clever each time; it is to remove the friction that causes the routine to be skipped.

A monthly checklist

A literal checklist (paper or Notion or wherever). Tick each step. Date and sign. Builds a record of when the routine ran and what was found. Useful when you go on leave or someone covers for you.

A small log of issues found

Each month, log the top issues caught. Over six months, the log shows the systemic problems (e.g. one channel consistently produces incomplete records, one team consistently misses gift aid asks). Fix the source, not just the symptom.

What the routine does not cover

The monthly routine is a maintenance routine, not a transformation routine. Major data work (a system migration, a one-off historic clean, a structural change to segmentation) sits outside it as a separate project. The routine keeps things from drifting; bigger projects address structural problems.

Common failure modes

Letting it slip during busy months

December and March tend to be the months the routine is skipped. They are also the months when bad data hurts most (December campaigns, March year-end). Protect the routine particularly in busy months.

No documented owner

Routines without a named owner do not survive staff change. Name a primary and a backup. Make it part of the data lead's job description.

Doing it perfectly once and never again

A perfect annual cleanse is worth less than an imperfect monthly routine. The compounding effect of small monthly attention is what keeps the data useful.

Data quality is a habit, not a project. The charities with trustworthy reporting are not the ones with the best tools; they are the ones with the best routines.

First-month set-up checklist

Pick the day of the month the routine runs (e.g. first Tuesday).
Build the seven saved queries / reports in the CRM.
Write the checklist (literal, one page).
Run the first cycle and time each step.
Log the issues found, fix any sources you can identify.
Lock the day into the team calendar permanently.

Six steps. After that, a few hours a month buys back days of crisis later in the year.

CRM Data Quality: The Monthly Routine

Why monthly, not quarterly