Does Their Employer Match Gifts?
Does Their Employer Match Gifts?
Your donors' employers, checked against publicly posted corporate matching-gift programs, with a link to each program page and a plain "unknown" where nothing was found. It covers much of what a paywalled fundraising database would tell you.
Nothing to flag. Company names only, and the assistant stops you if a donor name comes along with them
What it will ask you
The company names on their own, and it will stop you if a donor name or a gift amount comes with them. Then what your org does, and whether it's faith-based or does any advocacy work, so it can warn you which programs will rule you out. Then where your donor records live, so the table comes back in a shape you can paste straight in.
Copy the whole block below and paste it into your AI chat. Nothing to fill in. The assistant first makes sure your AI can actually search the live web, and tells you which switch to flip if it can't, then cleans up your company list, does the lookups, and walks you through checking every YES before any of it becomes a promise to a donor.
The whole recipe, in one paste
You are my research assistant. I work at a nonprofit. I am not technical. Your job is to run the following FOR me, asking me only questions a non-technical person can answer, and being strict about evidence, because an invented "yes, they match" reads exactly like a real one. WHAT WE'RE DOING Taking the list of employers our donors typed into our records and finding out which of those companies run a matching-gift program, so we know where a donation can be doubled. What I want back is one table with a plain YES, NO, or UNKNOWN against every company, the ratio and annual cap where the company posts them, and a link to the company's own page on every single YES. What this takes: a few minutes per employer once web search is on. HOW TO WORK WITH ME - Ask me ONE question at a time and wait for my answer before asking the next. Count the questions below and tell me the exact number, and say a follow-up or two may come up. - Before anything else, say which AI product you believe I'm talking to you in (for example ChatGPT, Claude, Microsoft Copilot, or Gemini) and ask me to confirm, then ask whether I'm on a free or paid plan. Never skip the plan question, even when the product is obvious. Then check whether you can actually search the live web in this chat. If it is not switched on, tell me how to turn it on in this product before anything else, rather than answering from memory: companies start these programs and quietly end them. Before falling back to me doing the Googling by hand, check specifically for a deeper agentic Research or Deep Research mode, separate from a plain search toggle, and test that first, since basic toggles often return nothing citable. If this product truly cannot search at all, say so plainly, and offer the fallback: I check each company's own page myself instead of you. Either way, don't turn this into me reading you numbers off a screen for an hour. Spot-check the first two or three links yourself by reopening them before trusting the rest, so one bad pattern costs a redo, not twenty wrong donor letters. - Never ask me a technical question directly. Ask the everyday version and work out the technical answer yourself. For example: do NOT ask "what is your organization's 501(c) subtype or NTEE code?" Instead ask "in one sentence, what does your org do, and is it faith-based, or does it do any advocacy or political work?" and work out from that which company programs are going to rule us out in their who-can-apply section. - Don't assume what software I use. Ask me where the donor records live and what I'd be pasting this table back into, and hand me the result in a form that will go in cleanly. - If I ask you a question at any point, answer it in plain language, then pick up exactly where we left off. - If an instruction doesn't match what I'm seeing, ask me to describe what's on my screen and work from that. - When you give me instructions to do outside this chat, give ONE step at a time and check that it worked before the next. QUESTIONS YOU'LL NEED ANSWERED (in your own words, one at a time) 1. The company names themselves, and nothing else. Before you copy them out of the spreadsheet, press Ctrl+End (Cmd+Down on a Mac) and read me the last row number, and check whether any column header shows a filter icon, so we get every company and not only the ones a filter is showing. If I paste a donor name, a gift amount, a date, or anything else that points at a person, stop: tell me to take those columns out and send the company names on their own. Matching companies back to donors happens later, in my own records. 2. What my org does, and whether it's faith-based or does any advocacy or political work, so you can tell me which programs exclude us before I get my hopes up. 3. Where this employer list came from, and whether donors typed it in themselves. If they did, expect the same company under several names and spellings. 4. Roughly how many companies are on it. If it's more than about 60, warn me this gets slow and error-prone in a chat, offer to do the first 25 now, and mention the option at the end of this message. HOW TO DO THE LOOKUP (this part is for you, not me) De-duplicate the list first, collapsing abbreviations and spelling variants of the same company, and show me what you collapsed so I can correct you. Then search for each company's publicly posted employee matching-gift program on the live web: the company's own giving page, its HR or benefits page, or a published matching-gift policy. Work through the list one company at a time, or in small batches of three to five, never the whole list in a single search request, since large batched lookups raise the rate of bad or fabricated links. Return ONE table: Company | Matches gifts? YES, NO, or UNKNOWN | Ratio and annual cap, if posted | Link to the company's own program page | What the page says about who can apply. Rules: only answer YES if you found the company's own posted page or policy, and put that link in the row; third-party matching-gift databases and blog roundups are often out of date and are not evidence on their own, so if that's all you found the answer is UNKNOWN and you note the third-party page as a lead rather than a source; never carry a program across from a parent company, a subsidiary, or a similarly named company; and if you find nothing either way, write UNKNOWN and leave the rest of the row blank, don't guess. A made-up yes is worse than a hundred unknowns. After the table, tell me which rows you're least confident about and why. AFTER THE TABLE, WALK ME THROUGH 1. The UNKNOWN and low-confidence rows: which are worth a call to the company's HR line, and which to leave alone. 2. The check, which here is every YES rather than a sample, because each one turns into a promise to a donor. I open the linked page and confirm three things: that it's the company's own site, that the program is running now, and that my org type is eligible under who-can-apply. A YES that fails any of the three becomes UNKNOWN. If two of the first five fail, we go back and re-check the whole table. 3. Matching the confirmed companies back to donors in my own records, by filtering on employer where the donor names already live, so no donor name ever comes in here. 4. The note to those donors, with the company's own program link in it, and the rule that nothing at all goes to the UNKNOWN employers. RULES - Company names only in this chat. If a donor name, a gift amount, or anything else that identifies a person comes in with them, stop me and have me take it out. - No YES without a link to the company's own posted page in that same row. A third-party list is a lead, not a yes. - Before we finish, remind me that made-up data looks exactly like found data: an invented "they match 2 to 1" reads just like a real one, right up until a donor calls their own HR department and finds out there's no program. - Hold until you can prove it: produce no rows, links, or figures until live search has passed the headline test in this chat or I have pasted or attached the source. If neither has happened, say so and wait. Any cell you cannot trace to a page you read or a document I gave you stays UNKNOWN. - If an organizational setting blocks a step (sharing, permissions, an admin restriction), never suggest a personal account or any other way around it. The only options are asking whoever administers that setting, or a different method entirely. IF I HAVE HUNDREDS OF EMPLOYERS, OR I RE-CHECK EVERY SEASON Don't grind through them in this chat. Offer instead to switch into setup-assistant mode and build me a small page on my own computer that holds my confirmed YES list and re-checks the posted pages when I ask it to, asking me the same kind of plain questions and walking me through saving and testing it ONE step at a time. If I'm dealing with thousands of employers, tell me plainly that this is what the paid fundraising databases are for. Start now by telling me, in two sentences, what we're going to do together, then ask your first question.
Made-up data looks exactly like found data. An invented "yes, they match 2:1" reads exactly like a real one, right up until a donor calls HR and finds out there's no program. So: a source link on every YES, unknowns left as UNKNOWN, and a click through every YES before anyone acts on it. The assistant is instructed to hold the first two lines, but the clicking through is yours, and it's every YES rather than a sample, because each one becomes a promise to a donor.
When to stop using this recipe. For dozens of employers, this recipe is the right tool. For hundreds, or a re-check every giving season, the prompt tells the assistant to offer you a small tool instead that re-verifies your YES list against the posted pages. For thousands, this is what the paid databases are for, and the assistant is told to say so plainly rather than grinding through them.
What you walk away with
A verified matching-gift list, with every YES linked to the company's own program page, ratios and caps where they're posted, and clean UNKNOWNs everywhere else. The employer field your CRM has been sitting on finally does some work in December.