Email Lists: The Owned Audience

Everything in the first six lessons runs on rented ground. The platform ranks, the platform recommends, the platform changes the rules, and the followers you built are an audience you reach only when its systems agree. An email list is different in kind. The addresses are in your export file. The messages arrive whether or not a recommendation system approves. Lesson 06 ended with $100 buying roughly 10,000 impressions or 27 email addresses, and this lesson is the argument for the 27.

Why email: the numbers

The best benchmark for advocacy email is M+R's annual study, built from 180 participating nonprofits' actual data, reported as medians, with sector breakdowns (M+R Benchmarks 2026). M+R itself warns against comparing editions, because the participant pool and some calculations change each year, so treat these as 2025 figures for that pool. Its email chapter (Email Messaging):

M+R 2026 email metric (2025 data, medians)	Value
Advocacy email click-through rate	2.3 percent
Advocacy email response rate (completed the action)	1.4 percent
Fundraising email click-through rate	0.59 percent
Fundraising email response rate (gave)	0.05 percent
Revenue per 1,000 fundraising emails	$54 (public media $309)
Email revenue per subscriber per year	$2.40 (up from $1.87)
Email's share of online revenue	11 percent
Email actions per subscriber per year	0.14
Donation conversion after an advocacy action	0.34 percent
Emails sent per subscriber per year	50 (31 fundraising, 9 newsletters, 2 advocacy; wildlife groups 108)
Net list growth	5 percent (public media 12 percent)
Unsubscribes, as share of starting list	12 percent (disaster and international aid 21 percent)
Bounces, as share of starting list	4 percent

Put the advocacy rows beside the social benchmarks from lesson 05. Socialinsider's 2025 average engagement per follower was 0.15 percent on Facebook, 0.48 percent on Instagram, 0.12 percent on X and 3.73 percent on TikTok, counting likes, comments and shares (Socialinsider). Those are reactions. M+R's 1.4 percent is people who finished a petition or sent a letter. The denominators differ, subscribers against followers, and the samples differ, nonprofits against brands, so this is a comparison of magnitudes rather than a controlled test. The magnitudes are not close.

Open rates after Mail Privacy Protection

Every vendor benchmark leads with open rate, and open rate is the one metric that no longer means what it says. Apple's Mail Privacy Protection "hides your IP address so senders can't link it to your other online activity or determine your exact location" and "prevents senders from seeing if you've opened the email message they sent you," though it "does not extend to links or email attachments" (Apple Support). It does this by loading remote content, including the tracking pixel that registers an open, whether or not the person reads. Litmus's September 2026 client data, from over 1.5 billion tracked opens, attributed 76.56 percent of opens to Apple (including MPP opens), 16.10 percent to Gmail and 3.48 percent to desktop Outlook, and Litmus estimates MPP affects roughly 55 to 60 percent of all opens, so it recommends relying on click-through and conversion instead (Litmus). The feature is a setting the user turns on; Apple's guide describes the switch, and the Litmus share shows how many have.

That is why the two most-quoted open benchmarks disagree so sharply and why neither is useful as a target:

Benchmark	Method and period	Nonprofit open	Nonprofit click	Government or politics open / click	All-industry open / click
Mailchimp	Billions of delivered emails from users who reported an industry; campaigns to at least 1,000 subscribers; data as of December 2023	40.04 percent	3.27 percent	Government 40.55 / 4.58; politics not reported	35.63 / 2.62; unsubscribe 0.22
Campaign Monitor	Over 100 billion emails sent on its platforms in 2021; 2022 edition	26.6 percent	2.7 percent	Government and politics 19.4 / 2.8	21.5 / 2.3; unsubscribe 0.1
M+R	180 nonprofits, 2025 data, medians	Not reported	2.3 percent advocacy; 0.59 fundraising	Not applicable	Not applicable

Sources: Mailchimp, which notes that "the accuracy of email open rates may be impacted by Apple's privacy changes"; Campaign Monitor, which says MPP "artificially inflates open rates" and skews click-to-open rate. A 40 percent open rate two years after MPP and a 27 percent rate the year it launched are not two measurements of the same thing. Click rates (2.3 to 3.3 percent for nonprofits across all three sources) agree far better, which is a sign they are measuring something real. M+R dropped open rate from its email chapter entirely.

Building the list

Lead magnets that fit advocacy

People give an address for something they want. For an advocacy page the honest offers are the things you already make: the full report behind the carousel, the public-records toolkit, an alert when the agenda for a meeting posts, the comment-letter template, the petition. The controlled evidence on which magnet works best does not exist in the public literature we could find; what exists is M+R's cost data, which tells you what the market pays. Cost per lead across channels averaged $2.81; on Meta it was $3.64; through lead vendors such as Care2 and Civic Shout, whose model is petitions, $1.59; display ads $86 and Snapchat $78 (M+R Advertising). Petitions are the cheapest known way to collect an address, which is why the lead vendors sell them.

Two cautions on petitions. First, disclosure: a petition page that quietly adds signers to a mailing list without saying so is the kind of practice the double opt-in section below is designed to prevent, and a signer who did not expect your newsletter is a future spam complaint, which the deliverability section says you cannot afford. Say on the form what the address will be used for. Second, the slacktivism finding from lesson 05: Kristofferson, White and Peloza found that a public token act such as signing a petition led to less subsequent meaningful help than a private one (Journal of Consumer Research). A petition that posts the signer's name publicly may satisfy the urge to help; follow it, within days, with the specific next action framed as a continuation of what the signer already did. M+R's 0.34 percent donation conversion after an advocacy action is what that follow-up looks like at the median.

Placement, landing pages and QR codes

We found no randomized study of signup placement for small organizations this session, and vendor claims about pop-ups and placement conversion rates are omitted for that reason. Treat placement as a test variable with the sample-size discipline from lesson 05: a signup form's conversion rate (signups per visitor) is a high-rate outcome, often in the single digits per hundred, so a test of two placements can resolve in a few hundred visitors each. The practical defaults, which cost nothing and need no evidence: a signup link in every social bio, a one-line ask at the end of every post that offers something ("the full report is in the newsletter"), a landing page that offers one thing and asks for one field, and a QR code to that page on every handout, sign and slide at every meeting you attend. People at a public hearing have their phones out and a reason to care; a QR code on a half-sheet is the cheapest lead you will ever collect, and the code can be generated free in any newsletter tool or by a search for a QR generator.

The social-to-email funnel

The funnel runs: impression, click, landing page, signup, confirmation, first action. Published benchmarks cover the first two steps and the total. WordStream's 2,946-campaign Facebook benchmark gives a median click-through rate of 1.57 percent and cost per click of $0.77 for traffic campaigns (WordStream). M+R gives the cost per completed lead, $3.64 on Meta. Between them sits the landing page, whose conversion rate you must measure yourself.

Double opt-in and hygiene

Double opt-in adds a confirmation click before an address joins the list. Mailchimp's help describes it as "an extra confirmation step that checks each email address or phone number," says it "can help improve open rates," that it is optional for email and on by default for some EU audiences, and that its reCAPTCHA step "keeps spambots from adding fake data to your audience" (Mailchimp Help). You lose some signups at the confirmation step. You gain a list in which every address was typed by a person who controls it and expected you, which is the only kind of list that stays under the spam complaint ceilings below. For an advocacy list that may be read by opponents looking for a complaint, the confirmation record is also your evidence of consent.

Hygiene is the routine that keeps M+R's 12 percent unsubscribe and 4 percent bounce from compounding into a deliverability problem. Yahoo's requirements put it plainly: send only to people who asked, do not buy lists or pre-check the opt-in box, and "honor the frequency of the list's intent" rather than turning a monthly list into a daily one (Yahoo Sender Hub).

Segmentation, welcome sequences and frequency

Segmentation means sending different things to different parts of the list: county, issue, action history. The controlled evidence on segmentation for advocacy is thin; what is verified is the personalization experiment below, and the logic of M+R's response rates, which are medians across whole lists. People who signed up for a water alert will act on water emails and unsubscribe from the rest. Tag by the offer they signed up for and by the actions they take, and send the ask to the segment it fits.

The welcome sequence is the two or three emails a new subscriber gets in the first week: who you are, the offer they signed up for, one easy action. We could not verify a controlled study of welcome-sequence effects this session, so no effect size is claimed. The practical reasons stand: a subscriber who just confirmed is at peak attention, and the first email is the one most likely to be read and to set the expectation of what follows, which matters for the complaint rate. Make it plain, short and useful, and put the unsubscribe link where it can be found.

Frequency has the best descriptive data. M+R's nonprofits sent 50 emails per subscriber per year, about one a week, with wildlife groups at 108, and still grew their lists 5 percent net after 12 percent unsubscribes (M+R). A weekly cadence is therefore the norm for organizations whose results these benchmarks describe. Yahoo's rule is the useful constraint: send at the frequency people signed up for. A list recruited for "the weekly update" that starts getting four fundraising emails a week will show it in the complaint rate before you see it anywhere else.

Subject lines: the finding that replicates

Most subject-line advice is vendor folklore about word counts and emoji that does not survive a second dataset, and it is omitted here. The one large, randomized, replicated result is Sahni, Wheeler and Chintagunta's: in field experiments with three companies sending tailored emails to millions of people, adding the recipient's name to the subject line raised the open probability by 20 percent (from 9.05 to 10.80 percent), raised leads by 31 percent (0.39 to 0.51 percent) and reduced unsubscribes by 17 percent (1.2 to 1.0 percent), with similar effects across the other firms and both acquisition and retention goals (Marketing Science 37(2), 2018). The authors' explanation is that the personal detail raises the effort readers put into the rest of the message. Two caveats: the setting was commercial, and the open figures predate MPP, so replicate on your own list using clicks. The lift in leads and the drop in unsubscribes are the numbers that matter, and both are measurable after MPP.

Deliverability: authentication and the 2024 rules

On February 1, 2024, Gmail and Yahoo began enforcing matching requirements for anyone sending to their users, with a stricter set for senders of 5,000 or more messages a day (Gmail; Yahoo). A county advocacy list is usually under 5,000 a day, and should meet the bulk rules anyway, because the newsletter service it uses sends well above that volume across all its customers and because the bulk rules are the ones that keep mail out of spam.

Requirement	All senders	Bulk senders (5,000 or more a day to Gmail; Yahoo's bulk tier)	Who does it
SPF and DKIM	SPF or DKIM	Both; DKIM key at least 1024 bits (2048 recommended)	You publish the DNS records your newsletter tool gives you
DMARC	Recommended	Required; policy may be p=none; Yahoo recommends a rua reporting address	You publish one DNS TXT record
Alignment	Not required	From-header domain must match the SPF or DKIM domain (relaxed alignment acceptable)	Send from your own domain, not a free webmail address
Forward and reverse DNS (PTR)	Required	Required	Your tool, for its sending IPs
TLS in transit	Required	Required	Your tool
Message format	RFC 5322	RFC 5322 (Yahoo: 5321 and 5322)	Your tool
One-click unsubscribe (RFC 8058)	Recommended	Required for marketing and subscribed mail: List-Unsubscribe and List-Unsubscribe-Post headers plus a visible link in the body	Your tool; check the headers
Honor unsubscribes	Promptly	Within 2 days	Your tool; you verify
Spam complaint rate	Below 0.3 percent	Below 0.3 percent; Gmail says aim below 0.1 percent	You, through list quality
No impersonating Gmail From: headers	Required	Required	You

RFC 8058 is the one-click standard: a List-Unsubscribe-Post header that lets the mail client send an HTTPS POST to unsubscribe the reader without a confirmation page, provided the message carries a DKIM signature covering the headers (RFC 8058, Jan. 2017). Every tool in the comparison below supports it; what you must do is send from a domain you control, so the alignment rule can be met and the DMARC record is yours.

The law

Law or rule	What it covers	Core requirements	Penalty or consequence	Applies to an advocacy list?
CAN-SPAM Act (15 U.S.C. 7701 and following; FTC rule)	"Commercial electronic mail messages": primary purpose is "the commercial advertisement or promotion of a commercial product or service" (7702(2)(A))	Accurate From, To and routing; honest subject lines; disclose ads as ads; valid postal address; clear opt-out honored within 10 business days without fees; no selling opted-out addresses; the promoted party and the sender are both liable	Up to $53,088 per email in violation	Usually outside the definition for pure advocacy; inside it when you sell merchandise or tickets. Meet the requirements regardless
Gmail and Yahoo sender requirements (private rules, since Feb. 1, 2024)	All mail to their users; stricter for 5,000 or more a day	Authentication, alignment, one-click unsubscribe, two-day honoring, complaint rate under 0.3 percent	Mail rejected or sent to spam	Yes, to every message
TCPA (47 U.S.C. 227) and FCC rules for political calls and texts	Calls and texts	Autodialed or prerecorded calls and autodialed texts to cell phones need prior express consent; manually sent texts do not; prerecorded calls to landlines allowed without consent, at most three in 30 days, with caller identity at the start and a phone number; consent revocable in any reasonable way, including replying "stop"	Statutory damages per call or text; FCC enforcement	Yes, the moment you add texting
State privacy laws	Collection and use of personal data by covered businesses	Varies by state; many exempt nonprofits and small entities	Varies	Covered in the coming module on state privacy law; check whether your state's law reaches nonprofits
Idaho Code 67-6614A	Paid express advocacy through broadcast, newspaper, magazine, outdoor advertising, direct mail "or any other type of general public political advertising"	Clearly identify the person responsible	Enforced by the Secretary of State	Email is not named; an email to your own list is not plainly "general public" advertising. Identify yourself anyway
Idaho Guidance 2026-5 (Secretary of State)	What counts as reportable political activity	"Emails to members" and unpaid social posts generally not reportable; "paid broadcast messages via radio, TV, text, or social media" are	Reporting thresholds in 67-6606, 67-6611, 67-6628	Your newsletter is generally not reportable; a paid text blast to voters is

Sources: FTC compliance guide; 15 U.S.C. 7702; FCC political calls and texts; Idaho Code 67-6614A; Guidance 2026-5. On CAN-SPAM, the FTC guide says the law covers "all commercial messages" and does not address nonprofit or political mail; the statutory definition turns on commercial purpose. An advocacy newsletter asking people to attend a hearing is not advertising a commercial product. One that sells tickets to a fundraising dinner is, for that message. Since the requirements (honest headers, an address, a working unsubscribe) are what Gmail and Yahoo demand anyway, the practical rule is to meet CAN-SPAM in every message and stop worrying about the classification.

Texting is where small groups get hurt. The FCC's rules are specific: autodialed texts to cell phones need the recipient's prior express consent; manually sent texts do not; consent can be revoked any reasonable way and must be honored (FCC). A texting tool that sends to a list is autodialing. Collect consent to text separately from consent to email, on the form, in words, and keep the record. Campaign finance lesson 06 covers the FCC's 2024 ruling that AI-generated voices count as artificial voices under the TCPA.

We searched for an Idaho statute requiring a disclaimer on campaign email and found none. 67-6614A names its media and ends with "any other type of general public political advertising"; it was written for ads, and an email to people who asked to receive it is not advertising to the general public. Guidance 2026-5 treats emails to members as generally not reportable. Put your name and address at the bottom of every email anyway, because CAN-SPAM wants a postal address, Gmail wants a visible unsubscribe next to it, and a reader should never have to guess who is writing.

Tools compared

Six tools, compared on what their own pricing pages said in October 2026, with no endorsement. Prices change; check before you sign. Nonprofit discounts require proof of status.

Tool (as of October 2026)	Free tier	Paid pricing shown	Nonprofit discount	Notes
Substack	"Publishing is free on Substack if your content is free, no matter how many subscribers you have"	10 percent of paid subscription revenue, plus Stripe's 2.9 percent and $0.30 per transaction and a 0.5 percent recurring fee	None stated	Newsletter plus a public archive and Notes; limited segmentation and automation compared with the marketing tools; export of subscribers available
Mailchimp	250 contacts; 500 sends a month, 250 a day	Essentials $13, Standard $20, Premium $350 a month at 12-month promotional rates; contact tier for those prices not stated on the page	15 percent, by request through billing	Full marketing suite: segmentation, automation, A/B tests, landing pages; price scales with contacts
Buttondown	First 100 subscribers free	Add-on pricing: $9 a month tier (tags and segmentation, paid subscriptions, surveys, analytics, RSS-to-email), $29 (custom-domain archives, multiple newsletters, automations), $79 (white-labeling, teams); subscriber-count pricing not legible on the page	50 percent for registered 501(c)(3)s	Plain-text-friendly; developer-oriented; small team
Kit (formerly ConvertKit)	Free up to 10,000 subscribers	Creator $33 a month billed yearly ($390) at 1,000 subscribers; Pro $50 a month first-year promotion, $66 after	None stated on the pricing page	Creator-focused automation and landing pages; the free tier is the largest of the six by subscriber count
Action Network	Free tools for getting started	Movement plan from $15 a month minimum at $1.25 per 1,000 emails; Network plan from $125 minimum at $2.50 per 1,000 emails for federated groups; Enterprise custom for 1 million or more emails a month; mobile messaging add-on	Not stated; itself a certified nonprofit	Built for organizing: petitions, events, letter campaigns, district matching. States it "exclusively partners with progressive causes," so it is not available to every advocacy effort
MailerLite	250 subscribers; 2,500 emails a month	Comfort from $12 a month; Power from $25; Enterprise custom	30 percent on paid plans	Marketing suite with automation and landing pages; lower entry price than Mailchimp at small list sizes

Sources: Substack FAQ; Mailchimp pricing; Buttondown pricing; Kit pricing; Action Network plans; MailerLite pricing.

Questions and answers

Our open rate is 52 percent and our click rate is 1.1 percent. Is the list healthy?

The open rate tells you about your subscribers' mail clients; Litmus attributes about 77 percent of tracked opens to Apple, whose Mail Privacy Protection registers an open whether or not anyone reads (Litmus). The click rate is the honest number, and 1.1 percent is below the nonprofit benchmarks (2.3 percent advocacy at M+R; 2.7 to 3.3 percent across Campaign Monitor and Mailchimp). Check whether every email has one clear thing to click, whether the list was recruited for what you now send, and whether six-month non-clickers are still on it.

Should we send the same newsletter to the whole list or segment it?

Segment by what people signed up for and what they have done, because an ask that fits earns clicks and one that does not earns unsubscribes, and M+R's nonprofits lose 12 percent of their list a year to unsubscribes already. We found no advocacy-specific experiment on segmentation; the general principle is supported by Sahni's finding that relevance cues (there, the recipient's name) raise attention to the rest of the message. Start with two segments, by issue, and measure clicks and unsubscribes per segment for a quarter.

Is Substack an email tool or a publishing platform?

Both, which is its trade-off. It is free for free publications regardless of list size and takes 10 percent plus card fees of paid subscriptions (Substack FAQ), gives you a public archive and a place in its network, and gives you less segmentation and automation than the marketing tools. For a writer-led advocacy effort whose list is its readership, it fits. For a coalition running petitions and events, the organizing tools fit better. Either way, export your subscribers regularly.

We collected 300 emails on a paper sign-up sheet at a fair. Can we add them?

Send each a confirmation email first, asking them to click to join, and add only those who do. Yahoo's and Gmail's rules both treat unrequested mail as the path to the spam folder, and a paper sheet gives you no proof of consent that the double opt-in record would. You will lose some of the 300. The ones you keep will not mark you as spam.

Do we need DMARC if we only send 800 emails a week?

Gmail requires DMARC only for senders of 5,000 or more a day, but your newsletter tool sends millions a day across its customers, and your own domain's reputation is what receivers judge. A p=none DMARC record costs nothing, gets you reports on who is sending as your domain (including anyone spoofing it), and satisfies the alignment rule when you later grow. Do it now, while there is nothing to break.

Can we text our email subscribers about a hearing tomorrow?

Only those who consented to texts. The FCC's rules require prior express consent for autodialed texts to cell phones; a texting platform sending to a list is autodialing, and email consent is not text consent (FCC). Add a separate, unchecked text-consent box to your form, record it, and honor "stop." A manually typed text from a volunteer's phone to a handful of people does not need prior consent, which is a different and slower thing.

Does an email to our list urging a yes vote on a levy need an Idaho "paid for by" line or a filing?

Guidance 2026-5 says emails to members are generally not reportable, and 67-6614A names media that do not include email and speaks of "general public political advertising" (67-6614A). Paying to send to a purchased list of voters, or paying for a text blast, is different: Guidance 2026-5 lists paid text and social messages as paid political activity, and the thresholds in lesson 04 apply. Sign every email with your organization's name and address regardless; it costs nothing and removes the question.

How often is too often?

The benchmark median is about one email a week (50 a year), and the groups at that pace lost 12 percent of their starting list to unsubscribes and another 4 percent to bounces, so about 84 percent stayed through the year (M+R). Yahoo's requirement is to honor the frequency people signed up for. Watch unsubscribes and complaints per send; when they rise with frequency, you have found your list's limit, and it is probably at or below weekly for a volunteer-run effort whose emails take time to get right.

Hard questions

Why build an email list when everyone is on social media?

Because "on social media" means reachable when a ranking system decides, and the benchmark gap is an order of magnitude: 1.4 percent of email recipients completed an advocacy action in M+R's 2025 data against 0.15 percent of Facebook followers who so much as reacted in Socialinsider's. Pew's 2025 survey does show 71 percent of US adults on Facebook and 84 percent on YouTube, which is why social is where you recruit (Pew). It is not where you can count on being heard. The list is the part of your audience that survives a platform's next policy change, and the only part you can take with you.

Isn't an email list just a way to spam people who signed a petition once?

It is when the signer was not told. The lesson's answer is procedural: say on the form what the address is for, require a confirmation click, send at the frequency promised, and remove people who stop clicking. Those steps are also what Gmail and Yahoo demand under the 0.3 percent complaint ceiling, so a list run as spam stops being delivered. The Kristofferson finding adds a reason to be careful with petitions in particular: a public signature can substitute for later action, so the follow-up has to be worth the signer's time or it will produce an unsubscribe rather than a hearing attendee.

Open rates are inflated, click rates are low, and the benchmarks disagree. Is any of this measurable?

Clicks, completed actions, unsubscribes and complaints are all measured at your end, not the reader's mail client, and they agree across sources far better than opens do (nonprofit clicks of 2.3 to 3.3 percent in three different datasets). Cost per lead is measurable to the dollar. What is not measurable is the open, which is why this lesson tells you to stop using it. The honest position is that email is the best-measured channel you have, once you drop the one metric that broke.

The deliverability rules favor organizations with technical staff. A volunteer group cannot manage DNS records.

The records are three or four lines of text pasted into a domain host's control panel, with the exact values supplied by the newsletter tool, and the how-to above is the whole job. It takes an afternoon once. The alternative, sending from a free webmail address, fails the alignment requirement and lands in spam, which punishes the volunteer group harder than the staffed one. If no one in the group has ever logged into the domain host, that is the first task, before the first send.

Why compare six tools and refuse to recommend one?

Because the right one depends on list size, budget, whether you run petitions and events, and in one case on which side of politics you are on, and because recommendations in a course that will outlive this year's prices age badly. The table gives verified prices, limits and eligibility as of October 2026 and a method for re-checking them. A reader who works through the six steps will pick the right tool for their effort, which is the only recommendation that holds.