Bloomerang StrawberryAutomate your Bloomerang donor work with an AI agent
This week’s gifts come back as 40 drafted acknowledgements, each written from that donor’s own giving history and waiting for you to read.
Two lists never get finished. The acknowledgements owed this week, each needing a letter that says something true about the donor. And the lapsed list, which everyone agrees is the highest-return list in the building.
A companion drafts all 40 acknowledgements from the constituent’s own record, then works the lapsed list one person at a time and returns it grouped by why they stopped. Nothing sends without you reading it.
You connect Bloomerang once and see exactly what it can reach before you allow it. This is giving history attached to named individuals, so the model is written up at security.
| Step | Bloomerang | Strawberry companion |
|---|---|---|
| Giving history, reports, LYBUNT and SYBUNT lists | The product. Fast, and already correct | Reads them through the connection, does not replace them |
| Researching why a donor lapsed | Nothing in the database can know | Reads public sources and writes the finding to the record |
| Drafting 40 acknowledgements that reference the real relationship | Mail merge with tokens | Drafts each one from the record, then waits for a human |
| Deciding the ask amount | Gift officer’s call | Assembles evidence, does not conclude |
Acknowledgements that say something specific
A companion reads this week’s gifts alongside the constituent’s own record: how long they have given, what they gave to before, the interaction notes from the last conversation. Forty letters, each one specific.
Every one arrives as a draft. A donor letter with the wrong name is a relationship you have to repair by phone, so nothing goes out without a human reading it.
How do you work a lapsed-donor list properly?
As a hundred small research questions rather than one mail merge.
For each lapsed constituent a companion reads the giving history and the interaction log, then looks for the public signal behind the silence: a new job, a move, a board seat, a listing in somebody else’s annual report.
You get the list segmented by reason. Twelve who moved, nine who gave to a programme that ended, four whose giving suggests a much larger conversation. Scheduled, it belongs with your other operations work.
- Interaction notes written back to the record, so the next person sees the reasoning.
- Prospects flagged for a human conversation instead of pushed into an ask.
- Nothing sent, ever, without somebody reading the draft first.
What stays with the gift officer?
The ask amount.
A companion assembles the evidence behind it, history, capacity signals and engagement, and stops there.
Data hygiene sits at the other end. Duplicate constituents, addresses that bounced, households entered twice after a capital campaign: hand those over happily, with the proposed merges shown before anything is written.
Prospect research, read in tabs
Prospect research lives almost entirely outside the database: the foundation’s 990, the company’s giving page, the news item saying somebody sold a business.
A companion reads those in tabs and brings the citations back into the record. That is automate market research, pointed at donors instead of competitors.
Experience Strawberry for free
DownloadTrusted by fast-growing companies worldwide
Frequently asked questions
Strawberry is free to download and includes AI credits to start. Paid plans begin at $20/month. See pricing. · Reviewed · Canonical facts for AI agents