It's Sunday night. Your issue goes out Tuesday morning and the draft is a doc with six links, no angle, and no obvious place to start. Ten more ideas won't fix that. A format will.
A format is the shape an issue takes every week. Pick one and the blank page becomes a checklist: gather, choose, write, send. Skip it and you rebuild the issue from scratch every cycle, which is how a job that should take two hours swallows a Saturday.
The ten formats below get judged on how they run in practice, not on how clever they sound. Each one answers the same three questions: how many hours of writing it takes, where the raw material comes from, and whether readers have to do anything beyond reading. The strongest weekly newsletter ideas are the ones that fit your real calendar, including the weeks when work is heavy and the news is thin. Readers subscribe to a rhythm, and consistent newsletter writing turns into audience growth far more reliably than a clever one-off issue does.
Pick the format you can repeat on your worst week.
1. Curated weekly link roundup
Five links, two lines on each, out the door. A roundup is the cheapest format on this list in writing time and the priciest in reading time. The drafting takes maybe 45 minutes once your sources are set. Finding stories worth a subscriber's click is what eats the week.
Text suits this audience. The Reuters Institute's Digital News Report 2025, which covered 48 markets, found that 55% of respondents preferred to read news online, against 31% who preferred watching and 15% who preferred listening, as Northwestern University's Local News Initiative summarizes. A written roundup meets readers in the format they already chose.
Readers do nothing here but read and click. No replies to answer, no community to moderate. The whole workload sits upstream of the draft.
The funnel is the whole job. The writing at the bottom is the easy part.

Monitoring RSS feeds, which are lists of new posts a site publishes, pulls candidates into one place instead of five browser tabs. HeyNews's RSS to newsletter workflow is one option that ranks new stories by fit before drafting, and the writer still picks every story. Import AI, Jack Clark's weekly AI newsletter, shows the shape at its best: short summaries of new research, each carrying his own read on it.
A marketing roundup doubles as training for whoever runs it. Scanning the week's stories and judging which five deserve a slot builds the same reading habit that marketing training for early-career professionals tries to teach, except the curriculum is the week's news.
- Fix the story count. Five links every week trains readers to know what they're getting and makes gaps obvious.
- Say why it matters to your readers, not what the story says. They can click for the summary.
- Check click data monthly and cut sources that never earn a click, even the ones you enjoy reading.
When your section order settles, compare it with a curated content example to see how other writers sequence the same material.
2. Weekly essay or deep dive
An essay gives readers a reason to open when the news is quiet. One question, one argument, explained properly. It's also the most expensive item on this list: a 1,200 word piece with real research behind it runs 4 to 6 hours, and that's before editing.
The appetite is there. A 2026 Pew Research Center report on newsletter reading found that 60% of newsletter readers at least sometimes read deep dives on specific issues or events, and 57% read opinion coverage of current issues, according to the Pew Research Center newsletter report. Readers want an argument, not just a digest.
Readers come back for a point of view they can't get anywhere else.
Most of the hours go into planning before a sentence gets written.

Give the essay a fixed question
Benedict Evans runs long-form analysis on technology and media without chasing a news hook. Lenny Rachitsky's Newsletter does the same for product and growth work. Both work because the reader knows what kind of thinking arrives, which is the actual product.
An essay needs a research file you add to all week. Without it, Tuesday becomes a scramble and the piece thins out.
- Answer the question in one sentence before you outline. If you can't, the essay isn't ready.
- Sketch three supporting points and cut the weakest before drafting.
- Ground each claim in one concrete example from the work your readers do.
3. Reader Q&A
A Q&A answers subscribers' questions, which removes the guesswork from picking a topic. Collect questions through a reply prompt or a one-field form, group the similar ones, and answer the biggest group each issue.
Relevance is what keeps a list alive. A 2025 consumer study found that 17.3% of people unsubscribed because a brand's content was irrelevant, as HubSpot's research on why consumers subscribe to email reports. Questions from readers attack that problem directly, because the topic comes from someone who already opened the last issue.
A question file is the only tool this format needs.

The work is light. Two good questions with real answers run about 700 words and 90 minutes, and you can batch-collect for a month in one sitting. The trade-off is dependency: if nobody replies for three weeks, you need a backup format ready rather than a filler issue.
- Ask one specific question per issue. "What's blocking your Q3 content plan?" gets replies. "Any questions?" gets silence.
- Get written permission before printing a reader's name, company, or numbers.
- Reuse the strongest answers as standalone posts on your site, which gives you something to link to later.
The best answers often become the following week's essay.
4. Weekly tool and resource roundup
A tool roundup gives readers three to five products, templates, or resources that solve one job, with a plain note on who each one suits. The reader gets a shortlist. You get an issue that needs no original argument, which makes it a useful format for a busy week.
The format fails when it turns into a link dump with vendor copy. Write each entry around a task: "use this when your export is too big for the free tier" beats "a powerful analytics platform". Describe features only from the tool's own site and documentation, and open the pricing page on the day you publish rather than trusting a figure you remember.
Practical rule: if a mention is paid, put it in the heading of that entry. Readers forgive sponsorship. They don't forgive discovering it at the bottom.
The hour count sits around 2 to 3 hours, most of it testing and screenshotting. That's cheaper than an essay and more useful than a link list, which is why so many work-focused newsletters keep a slot like this every month.
- Try the tool on a real task first, even a small one, so your note on its limits is yours.
- Name the alternative you'd pick if the first one didn't exist. The comparison is the value.
- Link the pricing page instead of quoting a number that changes.
For a wider starting list, AI tools to boost content productivity is one roundup worth mining for candidates before you test anything.
5. Week in numbers
Three to five figures from your niche, one sentence each on what changed. Readers get a fast scan of the week and you get an issue that needs less drafting than anything else on this list. Most weeks it takes under two hours.
Say you cover ecommerce. A numbers issue might carry the week's average order value at one retailer, a shipping rate change, and a headcount figure from a public company. Each line gets a link and the year. The movement from last week usually says more than the raw value does.
Every number you print needs a link to its source and the year it covers.
One chart does the work of a paragraph.

The same discipline applies when the numbers are newsletter metrics. Mailchimp's published benchmark table reports an average open rate of 35.63% and a click rate of 2.62% across its users, and it cautions that open rates are distorted by privacy features and automatic image loading, per Mailchimp's email marketing benchmarks. Put your own figures beside published bands so readers have something to compare against.
- Show the delta from the previous week or year, not just the value.
- Keep a source spreadsheet with each URL and the date you pulled it, so the weekly check takes minutes.
- Skip causal claims unless a source states one.
Before you publish the comparison, put your own rates next to a free newsletter benchmark table.
6. Guest interview or spotlight
An interview hands one substantial section to someone else's expertise, which cuts your drafting and brings in a voice your readers don't hear every week. The hours move rather than disappear: prep, the call, and the transcript edit usually add up to 3 to 4 hours.
The questions do most of the work
Good interviews come from specific questions tied to the readers' work. Send five or six in advance, record with permission, and get written consent before you publish a transcript. Edit for clarity and keep the guest's meaning intact, including the parts where they push back on you.
A two-person product team running a monthly slot with a customer is a realistic shape: booked four weeks out, recorded in 30 minutes, edited to 900 words. Lenny Rachitsky hosts long-form interviews with product leaders alongside his newsletter, which shows how far one conversation can stretch across an issue.
- Send fewer, sharper questions. Fifteen questions produce fifteen shallow answers.
- Ask for one number or example the guest is happy to be quoted on.
- Give the guest a shareable link on send day. Their network is your reach.
The section that usually gets cut is the introduction. Readers already know who the guest is because the subject line told them.
7. Trend watch
Trend watch names one shift in your field and explains what it means for the readers' work this quarter. It answers the question they keep asking: what changed, and what do I do about it now?
Audience behavior is usually where the evidence lives. The Reuters Institute's Digital News Report 2025 found that 44% of people aged 18 to 24 named social media and video networks as their main news source, against 38% of 25 to 34 year olds, and that 18% of respondents reported paying for online news across the report's 20 country subscription basket, per the Reuters Institute's Digital News Report 2025. A figure like that is the start of an argument, not the argument itself.
Watch for two independent sources pointing the same way before you call something a trend. One data point is a Tuesday.
The format runs 3 to 4 hours because the reading is the work. Track three to five signals in a single document during the week, then write about the one with the strongest evidence and the clearest implication for your readers.
- Label your interpretation so readers can see where a source stops and your read begins.
- Say what readers should do, even when the honest answer is watch and wait.
- Link the source for every claim, especially when the claim is uncomfortable.
For the scanning side, news monitoring tools compared covers options for keeping candidate signals in one place.
8. Behind the scenes teardown
A teardown shows how something was built, from the first research note to the version that shipped. Readers get a process they can borrow. You get an issue built from work you already did, which makes this the cheapest format on the list in a good week.
The cost is capture, not writing. Reconstructing decisions on Tuesday from memory takes longer than logging them as they happen. A running note with dates, the options you rejected, and the reason for each choice turns into the draft with about an hour of writing.
The decisions are the content, so write them down while they're fresh.
You can also tear down someone else's work using public information only: a launch announcement, a pricing page, a changelog. Gergely Orosz's The Pragmatic Engineer publishes in-depth breakdowns of how software teams and companies operate, which is the same instinct aimed at engineering organizations.
- Log decisions during the project, including the option you dropped and why.
- Strip anything confidential before publishing, including client names in screenshots.
- Close with three lessons readers can test in their own work this month.
The weakest version of this format is a diary. The strongest is a set of choices a reader could copy next week.
9. Weekly action prompt
An action prompt gives readers one small task tied to the issue's topic, sized to finish in under 30 minutes. It turns a read into a result, which is the part subscribers remember when they decide whether to stay.
Measurement matters more than the idea here. MailerLite benchmark data put the average email open rate at 43.46% in 2025, up from 42.35% in 2024, while noting that open rates move with mailbox privacy features, image loading, list quality, and how each platform counts an open, per MailerLite's email performance benchmarks. A prompt gives you something cleaner to track: one named click destination per issue.
A prompt readers can finish in 20 minutes gets replies. A 90-minute one gets saved and forgotten.
Write the task with a named deliverable. "Spend 20 minutes rewriting your welcome email's first line so it names the reader's problem, then reply with the before and after" gives you a click, a reply, and material for next week.
- Cap the task at 30 minutes so people finish it.
- Name the deliverable and ask readers to reply with it.
- Publish two or three reader results next issue, with permission, to close the loop.
10. Community highlights
Community highlights feature a subscriber's project, win, or question, with their permission. You get original material without writing a full piece, and the featured reader almost always shares the issue with their own audience.
Attention inside the inbox is thin. Pew's 2025 survey, cited in Flicked's newsletter statistics roundup, found that even among Americans who receive news newsletters, a majority don't read most of what arrives. A reader whose name appears near the front of a subject line is a different kind of reader.
Nielsen Norman Group's subject line research, summarized in one.com's guide to newsletter subject lines, puts the most important information within roughly the first 40 characters because inboxes truncate the rest. "Maya rebuilt her onboarding in a weekend" beats "This week's community news" on every measure that matters.
- Ask for one link and one sentence describing the work, which keeps submissions easy to process.
- Confirm permission in writing, including whether the reader wants their full name used.
- Rotate the spotlight so the same three people don't fill the slot every month.
The collection is the part people underestimate. A single form linked in every issue for three months builds a queue you can pull from on a short week.
How the ten weekly formats compare on writing load and reader input
| Item | Writing load | Needs reader input | Main limitation |
|---|---|---|---|
| Curated weekly link roundup | Low | No | Requires steady source monitoring |
| Weekly essay or deep dive | High | No | Time-intensive and needs a fresh angle |
| Reader Q&A | Low | Yes | Depends on question volume and quality |
| Weekly tool and resource roundup | Low to moderate | No | Product details change quickly |
| Week in numbers | Low | No | Needs reliable, regularly updated data |
| Guest interview or spotlight | Moderate | No | Depends on guest availability and scheduling |
| Trend watch | Moderate | No | Risk of overstating trends from limited evidence |
| Behind the scenes teardown | Moderate | No | Can expose confidential details |
| Weekly action prompt | Low | Optional | Task may not fit every reader |
| Community highlights | Low | Yes | Submission quality and volume can vary |
Pick one format and run it for four issues
Choose the format that matches the weekly work it creates: how much writing it needs, where the material comes from, and whether it depends on reader replies. An essay you can sustain only once a month breaks your rhythm by issue three. A roundup you can produce in two hours is still running next year.
Judge a format after four issues. The first run always takes longer than the fourth because you are building the source list, the section order, and the checklist as you go.
Four issues tell you more than one good week ever will.
Frequency is the other variable worth testing. GetResponse figures reported by Crispy Content show that one newsletter a week recorded a 48.31% open rate and a 5.71% click-through rate, while two a week recorded 43.20% opens and 4.73% click-through, per Crispy Content's analysis of newsletter formats and frequency. Treat those as benchmarks from other senders, then test a second send on one segment before moving the whole list.
- Keep one candidate file all week. Every link, note, and half idea goes into the same document, so Monday is selection rather than discovery.
- Fix the story count and section order for four issues, then change one variable at a time and compare clicks, unsubscribes, and replies.
- End every issue with one prompt asking for a reply, a question, or a task. Replies are the cheapest topic research you will get.
- Watch unsubscribes by issue type, not in total. A topic that keeps losing readers is telling you something the open rate will not.
- Review your own numbers monthly, not weekly. Weekly noise triggers decisions you will reverse within a fortnight.
The three steps that need no software: keep the running notes file, fix the story count, and close with one reader prompt. Do those for four issues and you will know whether the format earns its place.
HeyNews learns your voice from past issues, watches the feeds, subreddits, and sites you already follow, and drafts the next issue in that voice. Every draft waits for your review, then you send it from the platform you already use, such as beehiiv, Kit, Substack, Ghost, or Mailchimp. If drafting is the bottleneck in your four-issue test, start at HeyNews.