Batch vs continuous donor data screening: how current does your data need to be

Batch screening refreshes data in cycles; continuous screening keeps it current every day

Most fundraising teams screen donor data in batches. You send the file, a vendor appends wealth and contact data, and you work from that snapshot until the next refresh. Continuous screening flips the model. Your data updates on a rolling basis, so records stay close to live.

The real question behind this choice is simple: how current does your data need to be to make a good decision? The honest answer depends on the field, because some data barely moves in a year while other data goes stale within weeks.

This guide gives you decay rates by field type, the cost trade-offs of each model, and the questions to ask a vendor before you commit to a refresh cadence.

What is batch wealth screening?

Batch wealth screening is a periodic append. You export a donor file, a vendor enriches it with wealth markers, giving capacity and updated contact details, and you import the results. You repeat the cycle on a set schedule, often quarterly or once a year.

The strength is control and predictable cost: you pay per screen, you know what you spent, and you work from a clean snapshot. The weakness is drift, because the moment the screen finishes your data starts aging, and you will not see new signals until the next run.

What is continuous donor data screening?

Continuous screening keeps your donor records updated on a rolling basis. Instead of one big refresh, changes flow in as they happen: a new address, a liquidity event, a change in giving behaviour. Your file stays close to current without a manual cycle.

The strength is freshness and speed to act. The trade-off is cost and noise, because you pay an ongoing subscription and you need a reliable way to decide which updates actually matter.

How current does your donor data need to be?

Currency is not one number. Different fields decay at very different rates, so a single refresh cadence will always be too slow for some data and wasteful for others.

Use the decay pattern of each field to set your cadence, not a blanket rule.

Two rules fall out of this table. First, behavioural data decays fastest and matters most for timing, so it rewards a continuous feed. Second, wealth markers decay slowly, so paying to re-screen them every month buys little.

Batch vs continuous: the trade-offs side by side

Neither model wins outright. Batch suits slow-moving wealth data and tight budgets, while continuous suits behavioural signals where timing drives revenue.

What does each model really cost to own?

Sticker price is only part of the bill. Total cost of ownership includes the hidden work and the missed revenue around the screen.

For batch, count four costs: the per-screen fee, the staff hours to export, clean and import each cycle, the decisions made on aging data between runs, and the revenue lost when a lapse or an upgrade signal sits unseen until the next refresh.

For continuous, count three: the subscription, the integration and maintenance of a live connection, and the triage time to separate signal from noise. The ongoing fee often looks higher, but it can cost less once you add the staff hours and missed timing that batch hides.

A useful test: if your team acts on data weekly, a quarterly batch means you spend most of the year working from data that is one to 12 weeks old. Price that gap, not just the screen.

Wealth tells you who is rich; behaviour tells you who is ready

Screening answers one question well: who has capacity to give. It does not tell you who is ready to act now. A donor can have high capacity and zero intent, and a modest donor can be about to lapse or upgrade this week.

That is why cadence should follow behaviour first, then capacity. Start with who is showing the signals, then apply the wealth lens to size the ask. Fresh behavioural data is what turns a static wealth score into a timely action.

This is also where many teams over-invest. They refresh wealth markers often and watch behaviour rarely, which is backward. The slow field gets the fast cadence, and the fast field gets ignored.

What to ask a vendor before you commit to a refresh cadence

A refresh cadence is a contract commitment, so pressure-test it before you sign. Ask these questions and keep the answers.

• How often is each field updated at the source, not just delivered to me? A monthly delivery of annual data is still annual data.

• Which fields are refreshed continuously and which are point-in-time? Get this field by field.

• What is your coverage and match rate for my region and donor base?

• How do you measure accuracy, and can you show error rates by field?

• Can I trigger a re-screen on an event, like a major gift or an address bounce?

• How do updates land in my CRM, and does that need manual work each cycle?

• What filtering do you provide so my team sees material changes, not every minor edit?

• What is the all-in annual cost, including overage fees and integration support?

• Can I start with a batch pilot and move to continuous for only the fields that need it?

If a vendor cannot answer the field-by-field freshness question, treat the cadence claim as marketing, not a spec.

Practical recommendations

Match cadence to decay. Put behavioural and contact data on a fast or continuous feed, and keep wealth markers on an annual screen with event triggers.

Do not buy one cadence for everything, because a hybrid almost always wins: batch the slow fields and stream the fast ones.

Price the gap, not just the screen, because the cost of acting on stale data is usually larger than the line item for the refresh itself.

Start small. Run a batch pilot, measure how often stale data cost you a timely action, then move only those fields to a continuous feed.

Conclusion

The question is not batch or continuous. It is which fields need to be fresh, and how fresh. Behavioural and contact data decay fast and drive timing, so they reward a continuous feed. Wealth markers decay slowly, so an annual screen is usually enough.

Set your cadence by decay rate, price the cost of stale data honestly, and make vendors prove field-level freshness before you commit. Do that, and your data will be current where it counts, without paying to refresh what barely moves.