The week on the radar
Emails, news, videos, podcasts and AI research from the week of October 26 – November 1, 2026.
October 26 – November 1, 2026
From Monday to Sunday the radar picked up 566 items: 202 emails, 87 news articles, 234 videos and 43 podcast episodes. AI models did a first reading of 508 of them. This is what stood out.
Emails
202
from 80 organizations
News articles
87
from 24 sources
Videos
234
from 68 channels in 26 countries
Podcast episodes
43
from 18 shows
25 of 35
emails on October 30 asked for money, the most of any day. On Saturday, 8 subject lines raced a deadline.
Fundraising emails per day
16
subject lines dressed up for Halloween, against 1 the week before. Some made it a deadline.
Subject lines about Halloween, per day. Counted by code
52%
of emails made money the main ask, against 46% the week before. 106 appeals, up from 89.
Share of emails. 202 this week, 191 the week before
21%
of the week’s 2.44M views went to one Teletón Short. Teletón Chile posted 7 of the top ten.
Views of the week’s 234 videos, by November 3
12 of 87
headlines were about AI, against 4 the week before. Public policy fell from 28 articles to 14.
Articles on each topic. Headlines about AI: counted by code
99 of 286
statements were recommendations. Opinions stayed worried: 48 of 82, against 44 of 78.
The 82 opinions, by tone
The same AI model read both weeks, so their labels can be compared, with care: feelings and tone are soft labels. Subject-line counts are plain code.
What landed in the inbox
202 emails reached us from 80 nonprofits, all in the United States: more than the week before (191), with more appeals (106, against 89). They peaked on Friday, October 30; on Halloween, deadlines ran to midnight.
Emails received per day. Dark: fundraising appeals. Light: every other purpose. n = 202
Emails per day using a matching gift, and subject lines about Halloween
Device: a matching gift
Subject line: Halloween
What each email is for, one label per email. n = 202
The one thing each email most wants you to do. Dark: asks for money. n = 202
Six of the 12 fundraising emails that arrived on Saturday, October 31, the month’s last day. Central Time
01 · Emails, continued
Devices and feelings are softer labels than purpose and ask. Read the first two charts as a rough guide.
Emails using each persuasion device; up to three per email. n = 202
Feelings each email aims for; up to three per email. n = 202
Emails by day and hour of arrival, Central Time. n = 202
Emails per cause, for causes with 18 or more. n = 202
The date or season an email is tied to. n = 123 of 202
The voice of each email. n = 202
41%
use the reader’s first name in the text; 20 put it in the subject line, and 28 subject lines carry an emoji.
Main words, with the number of emails each one marks
Main words are the terms that make one email different from the others. Statistics, not AI.
What the trade press wrote
87 articles from 24 sector publications, fewer than the week before (97). Running a nonprofit led the topics again, then philanthropy and fundraising; public policy fell by half, and 12 headlines were about AI. Most stories read as neutral; 12 read as negative, against 16 the week before.
Articles published per day. n = 87
Topics of the articles; up to three per article. n = 87
Tone of each article, overall and for the eight most covered topics. n = 87
Feelings each story leans on; up to four per article. n = 87
Topics, tone and feelings are an AI model’s reading. The news labels have not had our hand-labelled test yet: read them as a rough map, not as findings.
02 · News, continued
Six stories chosen by us for range, then what the news could start. Each headline opens the article.
In the order of the list, not of importance
The Chronicle of Philanthropy · Thursday
The NonProfit Times · Tuesday
The Chronicle of Philanthropy · Thursday
The Agitator · Friday
M+R · Thursday
Nonprofit Quarterly · Wednesday
Articles per source: the seven most active of 24. n = 87
For each article the AI drafts three ideas a nonprofit could use, each with a channel. n = 261
Drafted by the AI model from one article each. We picked the six; the wording is the model’s
Test cutting appeal frequency and track net revenue over two years.
Shows whether fewer asks can raise total giving. From an article in The Agitator.
Replace ‘transformational gift’ with the specific change it will fund.
Donors respond better to concrete, measurable outcomes. From an article in The Chronicle of Philanthropy.
Rewrite top program page openings to answer common questions in 30 words.
Answer-led openings are easier for AI tools to quote. From an article by Getting Attention.
Post a short video update that stands alone, with the donate link saved for Giving Tuesday.
Preserves limited free organic links for urgent fundraising moments. From an article by M+R.
Audit your fiscal sponsor agreement for termination and transition terms.
Shows donors and boards you are planning for continuity. From an article in Nonprofit Quarterly.
Ask donors for multi-year recurring gifts, not one-time election-season contributions.
Article argues one-time grants cannot sustain democracy work. From an article in Nonprofit Quarterly.
What they posted on YouTube
234 videos from 68 channels in 26 countries, in 13 languages. 84% run under three minutes and 53% are Shorts. Teletón Chile, on tour before its broadcast on November 6 and 7, was watched most.
Videos posted per day. n = 234
Videos by the organization’s country. n = 234
Videos by language. n = 234
Videos by length, as YouTube reports it. n = 225; 9 scheduled videos have no length yet
53%
are Shorts: 125 of 234 videos.
1:12
is the median length, in minutes and seconds.
358
views for the median video. Half had fewer.
With their views on November 3. All 234 videos together: 2,444,988. The one-line descriptions are ours.
70% of the telethon’s money comes from families.
Teletón Chile
Chile · Short, 0:21
Comedian Gila’s 1981 telethon sketch, back for laughs.
Teletón Chile
Chile · Short, 1:02
Actress Kim Hye-ja on why she still acts, 65 years on.
World Vision Korea
South Korea · Short, 0:59
Five hours live from the tour stop in Puerto Montt.
Teletón Chile
Chile · Video, 5:07:20
Thirty seconds asking for monthly donors.
SickKids Foundation
Canada · Video, 0:30
03 · Videos, continued
181 of the 234 videos had a transcript. AI asked each one five questions: who speaks, about what, why, how, and the ask.
Topics of the videos; up to three per video. n = 181
Why the video was made; a video can have several aims. n = 181
How it is told. n = 181
Who speaks. n = 181
What the video asks you to do. n = 181
Feelings each video appeals to, and how strongly. n = 181
Videos by aim (rows) and ask (columns). n = 181
The video labels have not had our hand-labelled test yet: read them as a rough map, not as findings. A video without a transcript is not labelled.
What the sector said out loud
43 new episodes from 18 shows about the sector. We read the transcripts of 38, about 16 hours of talk, and pulled out 286 statements: what hosts and guests think, recommend, learned and expect.
Problems the episodes discuss; an episode can raise several. n = 38 episodes with a transcript
Statements by kind. n = 286
Opinions by tone. n = 82
Topics of the episodes; up to four per episode. n = 38
New episodes and clips per show: the seven most active of 18. n = 43
04 · Podcasts, continued
Eight statements from the week, in our words. Code matched each one to its place in the transcript; the show’s name opens the episode.
Heroes for Children’s monthly giving, rebuilt with a name, its own page and one consistent ask, grew from under $1,000 a month to over $7,000 in a year.
Dana Snyder · on Missions to Movements
She was sure volunteers already gave too much to be asked. A low-budget video asking for $5 a month signed up 300 of them.
Lucy Madden, Letters to a Pre-Scientist · on Sustainable Giving
After a call went wrong, he promised a donor never to ask until invited to. About 20 months later the donor gave almost $400,000.
Dolph Goldenberg · on The Nonprofit Leadership Podcast
An email from a trusted vendor changed the payment details on an expected invoice. A phone call stopped a $250,000 fraud.
Justine Townsend, Your Part-Time Controller · on The Nonprofit Show
Nonprofits running a deficit went from about 22% to 39% in a year. The 2026 filings will show 50, 60, even 70%, he expects.
Dolph Goldenberg · on The Nonprofit Leadership Podcast
Seven in ten nonprofit workers in the U.S. are looking for new work, over three years of his survey, while other sectors’ rates fall.
Evan Wildstein, Social Impact Staff Retention Project · on Nonprofit Nation with Julia Campbell
With tens of billions of dollars leaving the sector, why haven’t foundations given more? He cannot understand it.
Phil Buchanan · on Giving Done Right
Follow up with event guests within 24 hours, ideally sooner: her email is already in their inbox as they drive home.
Julie Ordonez, Courage Lab · on Hey Nonprofits!
Statements by topic (rows) and kind (columns). n = 286
What the readings add up to
AI models did the first reading of 508 of the week’s 566 items. Side by side, the readings show what no single inbox can: a different feeling leading on each channel, two separate conversations, and two different asks.
What the readings show when they are put side by side
1
Compassion leads in the inbox (51% of emails, urgency 50%), pride in the news (59%) and hope in videos (69%). Fear and anger never pass 20%.
2
News and podcasts talk about running a nonprofit: management is a topic in 67% of articles and 63% of episodes, and in 1% of videos. The videos talk about children and health.
3
Email leads with the gift: money is the main ask in 52% of emails, and 17 subject lines offered a match. On video, 62% say “learn more” and 24% ask for a donation.
Items with their AI reading done, of the week’s items
53 had no transcript, so no labels
5 had no transcript
Code looks up every quote in the transcript; the rest are trusted less
Share of items on each topic, %. One topic list for all three sources
05 · AI research, continued
The same questions, asked of different channels. Feelings first, then what each channel asks you to do.
Share of items that aim for each feeling, on one 0–100% scale
A dot: that source’s list of feelings has no such entry. Email labels say “compassion”; news and video labels say “empathy”.
The main ask, one per email, % of emails. n = 202
Asks per video, % of videos; a video can make several. n = 181
Dark bars: asks for money. Two different samples: the emails come from 80 nonprofits we subscribed to, nearly all in the United States; the videos from 68 channels in 26 countries.
The same rules for every source
Every item of a kind gets the same questions, and each answer comes from a closed list. Labels cannot be invented, so they can be counted.
A video or episode without a transcript is not labelled. Nothing is guessed from a title.
Email labels are tested against 200 emails labelled by hand. News and video labels have not had that test yet: read them as a rough map.
About this report
So you can decide how much to trust them
Everything Giving Radar collected from Monday, October 26 to Sunday, November 1, 2026, counted by the day it arrived or was published, U.S. Central Time. Views are YouTube’s own counts, read on November 3.
We subscribe to nonprofits’ public mailing lists with a separate address for each one, since September 18, 2026. We read as a recent subscriber who has never given, so much of what arrives is a welcome series. All of this week’s emails came from organizations in the United States.
39 news feeds, 158 YouTube channels of organizations on our list, and 27 podcasts about the sector. All of it is public.
Anything with a right answer (dates, links, words, views) is counted by plain code. AI models answer fixed questions from fixed lists of answers. A video or episode without a transcript is not labelled.
Email labels are tested against 200 emails we labelled by hand; persuasion devices and feelings are softer and serve as a rough guide. The same model read this week and the week before, so the two weeks’ labels can be compared, with that care. News and video labels have not had that test yet: read them as a rough map, not as findings. Podcast statements are matched to the transcript by code; the eight shown here were found there.
No figure about one organization’s emails is published below 15 emails, so this report has none. A blank means not researched, never “no”. The full method is at givingradar.com/method.
Dates fundraisers plan around, from our Giving Calendar
The whole year is at givingradar.com/calendar.
Every morning
Giving Radar Daily is a short, free morning email for fundraisers: what nonprofits sent, said and posted the day before.