Guide
A sample workflow assessment report, on made-up data
By Mavrin Labs. Published . Last updated . Reviewed by Mavrin Labs on . How this guide was researched and checked.
The workflow assessment report records how a task runs today, how ready it is for software, and the fixed plan for a first build. The sample below shows that report for a made-up literacy nonprofit, built on a made-up donor list; every count in it comes from running the fixed rules of Mavrin's donor follow-up demo on that list.
The page is for an owner or a nonprofit lead deciding whether to commission a workflow assessment, who wants to see the report before agreeing to one. Every section below is sample data about a made-up organization, so read it as a workflow assessment example, shortened to the parts an owner reads first.
What is a workflow assessment report?
A workflow assessment report is the written result of the first stage of a Mavrin Labs engagement: the chosen task as it runs today, timed as a starting point called the baseline, where time could come back, and a fixed plan for the first build that the client approves or declines.
The report is the decision document. It says what a build would change, what it would leave with people, how the result would be measured and on what date, all before anything is built. How an engagement works explains the stages around it.
Why it matters: the thank-you is part of the gift
Individual donors gave $394.2 billion of the $617.20 billion given in the U.S. in 2025, according to Giving USA 2026 (Giving USA announcement, checked Monday, November 16, 2026). For a small nonprofit, the follow-up after each gift is a task worth measuring before anything is built.
The thank-you also carries a duty. A donor needs a written acknowledgment for any single gift of $250 or more to claim a deduction (IRS Publication 1771, checked Tuesday, November 17, 2026), so a late or missing letter is more than a courtesy lost.
The rule is older than most donor databases. In a notice in the Federal Register on Friday, August 4, 1995, the IRS explained that Congress added the $250 acknowledgment rule and the $75 disclosure rule to the tax code in 1993, for gifts made from Saturday, January 1, 1994 (the IRS notice, checked Wednesday, November 18, 2026).
Guidance on donor data came much later. The council advice this report follows on telling donors how their data is used, AI tools included, dates from January 2024, as the section on risks below shows. A report written today has to meet both: the old duty to thank and the newer duty to explain.
Sample report: the organization and the scope
In this sample, the organization is a made-up literacy nonprofit with three programs: tutoring, books for families and summer reading. The assessment covers one task, donor follow-up after a gift, as it ran in the week reviewed, which ended on Sunday, November 15, 2026.
Gift entry into the books, the monthly board report and volunteer scheduling are out of scope for this sample. The report lists each one for a later assessment, with the reason it waits: the board report depends on gift entry, and scheduling changes every term.
Sample report: how the task runs today
The sample report writes the task down step by step, as the development lead described it and as the assessment watched it run:
- A gift arrives online or by check and is entered in the donor database.
- Once a week, the development lead exports the gift list to a spreadsheet.
- The lead reads each row and decides who needs a thank-you, a call or an update.
- Letters are written one at a time from an old template.
- The executive director signs each letter, and a volunteer mails it.
In this sample, the weekly export is the only moment anyone looks at the whole list. A gift that arrives just after the export waits most of a week before anyone sees it, which is where the delay in the baseline comes from.
Sample report: the baseline
The sample baseline comes from a made-up donor list of fourteen donor records, reviewed as of Sunday, November 15, 2026. Each count below was taken by running the fixed rules of Mavrin's donor follow-up demo on that list.
| Measure in the sample | Count from the sample list |
|---|---|
| Donor records reviewed | fourteen |
| Gifts with no thank-you on record | five |
| First-time donors waiting for a welcome | two |
| Gifts of one thousand dollars or more waiting for a call | two |
| Oldest gift with no thank-you | thirteen days, from Monday, November 2, 2026 |
| Donors who give every month | two |
Time per letter is the one baseline a sample list cannot show. In a real assessment, staff time each step on real items, and those minutes join the counts above as the baseline the build is judged against.
Sample report: the made-up list behind the counts
The made-up list itself is below, so each count can be checked by hand. The donor follow-up demo that applied the rules is not published on this site for now; its records are numbered one to fourteen here, and amounts and counts are written out in words. Days waiting run to Sunday, November 15, 2026, the date of the review.
| Record | Program | Latest gift | Amount of the latest gift | Gifts so far | Gives monthly | Last thanked |
|---|---|---|---|---|---|---|
| Donor one | Summer reading | Thursday, November 12, 2026 | fifty dollars | one | No | Not yet: three days waiting |
| Donor two | Tutoring | Monday, November 9, 2026 | two thousand five hundred dollars | four | No | Not yet: six days waiting |
| Donor three | Books for families | Friday, October 30, 2026 | two hundred fifty dollars | six | No | Saturday, October 31, 2026 |
| Donor four | Tutoring | Monday, November 2, 2026 | one hundred dollars | two | No | Not yet: thirteen days waiting |
| Donor five | Tutoring | Sunday, November 1, 2026 | twenty-five dollars | twenty-four | Yes | Sunday, November 1, 2026 |
| Donor six | Books for families | Thursday, August 14, 2025 | five hundred dollars | five | No | Friday, August 15, 2025 |
| Donor seven | Summer reading | Monday, March 3, 2025 | seventy-five dollars | one | No | Wednesday, March 5, 2025 |
| Donor eight | Tutoring | Tuesday, June 20, 2023 | forty dollars | two | No | Thursday, June 22, 2023 |
| Family foundation (record nine) | Tutoring | Tuesday, September 15, 2026 | five thousand dollars | three | No | Wednesday, September 16, 2026 |
| Donor ten | Books for families | Saturday, November 14, 2026 | twenty-five dollars | one | No | Not yet: one day waiting |
| Donor eleven | Summer reading | Monday, October 5, 2026 | one hundred fifty dollars | three | No | Tuesday, October 6, 2026 |
| Donor twelve | Books for families | Sunday, November 1, 2026 | ten dollars | nine | Yes | Sunday, November 1, 2026 |
| Donor thirteen | Tutoring | Friday, November 13, 2026 | one thousand dollars | two | No | Not yet: two days waiting |
| Donor fourteen | Books for families | Saturday, January 10, 2026 | sixty dollars | two | No | Monday, January 12, 2026 |
Sample report: readiness of the task
The sample report runs the five readiness tests on donor follow-up and records what it found for each one:
| Test | Finding in the sample | Ready |
|---|---|---|
| The task repeats | Gifts arrive every week | Yes |
| The input is readable | Each gift is a record with a date, an amount and a program | Yes |
| The steps are known | The lead's rules are written down for the first time in this report | Yes, once agreed |
| The result can be checked | Each gift is either thanked by the organization's own goal or it is not | Yes |
| A mistake can be caught | Staff read every draft before it goes out | Yes, with approval kept |
The tests come from the guide to what a small business should automate first, which explains each one.
Sample report: the recommended first build
The sample report recommends one build: a daily review of new gifts that sorts each donor by the organization's own rules, puts the week's thank-yous, calls and updates in order, and drafts each letter for staff to approve. Nothing is sent without a person's approval.
The rules come from the development lead's own practice, written down during the assessment. First gifts are thanked by the second day, other gifts by the seventh day, and gifts of one thousand dollars or more get a personal call from the executive director by the second day.
Sample report: what runs on its own and what staff approve
The sample report splits every step of the new process between the software and the staff who answer for it:
| Step | Runs on its own | Staff approve |
|---|---|---|
| Reading new gifts each morning | Yes | Nothing to approve |
| Sorting each donor by the rules | Yes | Any label can be changed |
| Drafting thank-you letters | Yes | Every letter before it goes out |
| Notes for a large-gift call | Drafts the notes | The executive director makes the call |
| Changing a donor's record | Never | Always a person |
The build reads the donor database and writes only drafts and labels. Under least privilege (giving each program only the access its job needs), it cannot change a gift, a name or an address (NIST glossary entry on least privilege).
Sample report: acceptance tests
The sample report lists the acceptance tests the build must pass before the organization accepts it:
- A first gift entered on a Monday appears in Tuesday's list with a welcome draft.
- A gift of one thousand dollars or more produces call notes, never a letter.
- A donor already thanked for the latest gift gets nothing this week.
- A record with a missing date is set aside for a person, not guessed.
- Every draft waits for approval, and the system sends nothing itself.
Each test runs on the sample list before launch and again after any change to the rules. The guide to testing an agent before clients see it explains how tests like these are written.
Sample report: the result to measure and the review date
The result the sample report proposes is that every gift gets its thank-you, welcome or call by the organization's own goal, checked on a review date the organization picks. The baseline to beat is the sample list's five gifts with no thank-you, the oldest waiting thirteen days.
Staff time per letter is the second measure, taken the same way before and after. The report sets its target only once the timings exist, so the number is measured rather than guessed. The page on agreeing a measurable result before the build explains the method.
Sample report: risks and what stays with people
The sample report names the risks of the build and who handles each one:
- A draft can state a detail wrong, such as a program the donor never funded, so staff read each one.
- Large gifts and complaints stay with the executive director.
- The donor database stays the system of record (the one program whose copy counts), and the build never edits it.
- Donors are told how their data is used, AI tools included.
The last point follows guidance the National Council of Nonprofits gave in January 2024 on telling donors how their data is used (the council's article on data privacy). The guide to a nonprofit AI-use policy turns it into a policy.
What happens after a real report
A real report closes the first stage. The client approves the plan or says no. A Build starts only after a written design plan and acceptance tests are agreed, and nothing is sent without the client's approval by default.
If the client goes ahead, the result to measure and the review date written in the report become the targets the build is checked against. Either way, the baseline in the report is the starting figure for any later comparison, because a change can only be judged against how the task ran before it.
The donor database to accounting guide covers the gift-entry task this sample left for later.
Key takeaways
- The sample report covers one task, donor follow-up, for a made-up literacy nonprofit.
- Every count in it comes from running the donor follow-up demo's rules on a made-up sample list.
- The recommended build sorts and drafts, and staff approve every letter.
- The result to measure and the review date are agreed before anything is built.
Related guides
To choose the first task for your own organization, read the guide to what a small business should automate first, and run the readiness checks for automation on it.
Where to go next
To find the task your own week points to, the six-question project finder asks where your team's hours go and asks for no contact details. To talk through an assessment of your own, start a conversation.
Frequently asked questions
Is this a real client's report?
The report is not a real client's report. The organization, its donors and every count are made up: each count was taken by running the fixed rules of Mavrin's donor follow-up demo on a made-up sample list. A real report uses the client's own records and timings, and the client decides who may see it.
What does the assessment report include?
The assessment report includes the task as it runs today, a baseline measured from real items, a readiness check, the recommended first build, what runs on its own and what staff approve, the acceptance tests and the result to measure with its review date. The client then approves the plan or says no.
Why does the sample cover only one task?
The sample covers one task because a first build works best when it changes one thing that can be measured. A real assessment can map several tasks, but the report still recommends one first build and lists the others for a later assessment, each with the reason it has to wait.
Can the recommended build send letters on its own?
The recommended build in the sample cannot send letters on its own. It sorts donors and drafts each letter, and a staff member approves every one before it goes out. Loosening that rule for one message type would be a later decision, made only after the tests show the drafts are right.
How is the result checked after launch?
The result is checked on the review date the organization agreed, by measuring the same things the same way as the baseline: gifts waiting for a thank-you, how long the oldest has waited, and staff time per letter. The next step, whether a change or a second build, comes from what that review shows.
