Articles / AI Marketing Newsletter: 4 Picks by Job and Reading Time

ai marketing

AI Marketing Newsletter: 4 Picks by Job and Reading Time

Finn ·

An AI marketing newsletter is a recurring email that explains what changed in AI and how to use it in marketing work. For a solo marketer in September 2026, four deserve a trial: This Week in AI for weekly news, Almost Timely for long how-to essays, Growth Memo for AI search, and The AI Marketing Newsletter for daily news. Start with two: one for news, one for method.

Four AI marketing newsletters, sorted by the job they do

I went through each signup page and public archive on 26 September 2026. Reading times convert word counts at 238 words per minute, the average silent reading rate for English non-fiction in Brysbaert's 2019 meta-analysis, unless noted otherwise.

This Week in AI, from Marketing AI Institute: weekly news, free. The institute says more than 40,000 marketers and business leaders read it and promises "in-depth analysis of the top stories in AI each and every week." The sample on the signup page, a screenshot from November 2023, is a short digest: three top stories with one line each, then a list of other news, each item linking to a full article. The page shows no archive of recent issues, so check the length of the first one you receive. Your reading time depends on how many links you open.

Almost Timely News, by Christopher Penn: one long how-to essay a week, free. It arrives on Sundays. The 16 Sunday issues from 7 June to 20 September 2026 ran from 3,164 to 6,794 words, a median of about 4,500, or 19 minutes. That count includes the plugs, event lists and links around the essay: in the 20 September issue, the essay is about 3,100 of 4,588 words, roughly 13 minutes. Every post in the recent archive is free, each issue links a video and an MP3 version, and the four other emails in that window were a reader survey and three promotions.

Growth Memo, by Kevin Indig: AI search and organic growth, free essays and paid depth. It describes itself as written "for senior marketing leaders", and recent topics include how brands show up in AI search and whether Google has demoted listicles. Of its 23 posts since 15 June, four are open to everyone: the Monday essays from 31 August to 21 September, 1,691 to 2,289 words each, 7 to 10 minutes. The other 19, including the Growth Intelligence Brief and new assets such as a Claude skill, are for paying subscribers: $15 a month or $150 a year on the website, $21 a month and $205 a year through the app (subscribe page, checked 26 September 2026).

The AI Marketing Newsletter, bylined Aleric Heck: daily news, free. Several newsletters share this name; this one is published on beehiiv. From 13 to 26 September, its archive lists 24 emails. Fourteen are morning news editions, labeled 8 or 9 minutes each; the 24 September one carries a sponsor line, a top story, a takeaway box, quick hits, tool picks and a block for Heck's ThumbnailCreator. The other ten, labeled 2 or 3 minutes, all link to a product or a workshop from Heck's own companies: AdsCreator, ThumbnailCreator and AdOutreach. By the site's own labels, that is 140 minutes over two weeks, about 70 a week.

After each issue, write one line naming something you will do differently this week because of it.

Choose by the time you have

All four send free issues, so the main price is your reading time. Take 30 minutes a week as an example budget. The daily newsletter alone asks for more than twice that, and the Almost Timely essay takes almost half. This Week in AI plus the Almost Timely essay should fit, if the digest is still as short as its sample and you open only the links that concern your work.

The pair has one of each for a reason. A news digest tells you that something changed; a method essay tells you what to change in your own work. A second news source adds headlines, not method.

The keep test: one line per issue

The rule I propose: after each issue, write one line naming something you will do differently this week because of it. After four weekly issues, or two weeks of a daily, unsubscribe from any newsletter that produced no line. Four issues is a rule of thumb, not a measured threshold; it covers about a month, enough to see more than one topic. The rule changes if your work is briefing clients on AI news: then knowing the news is the job, and a digest earns its place without producing actions.

Here is the test on real issues from this month. The 20 September Almost Timely, "What's Changed in Basic AI Prompting", argues that current models are tuned to act rather than to know, so you supply the knowledge and define what finished looks like. It gives a before and after of one prompt line. The line it earns in your note:

2026-09-20  Almost Timely
Change: in my saved prompts, replace
  "Ask me one question at a time until you have enough information
  to succeed at the task."
with
  "Ask me up to 20 clarifying questions until you have enough
  information to succeed at the task."
Also delete "think step by step" and "take a deep breath".
Check: the next chat asks its questions in a single reply,
and I answer them all at once.

Dropping "think step by step" is Penn's reading of how current models behave, not a result measured for this article; the check line shows whether the new instruction does what it says in your own assistant. His argument rests on models becoming agentic, a word that Define Agentic: What the Word Means and a Three-Question Test pins down. If you have not settled on one assistant yet, Best AI for Marketing: One Assistant, Then One Publisher covers that choice first.

The daily scores one email at a time. Its 24 September morning edition, "AI Got Cheaper - Use the Right Model", earns a line if you pay for AI by usage: send summaries and data extraction to a cheaper model, and compare the cost of a finished result rather than the price per token. On a flat chat subscription, it earns nothing. The evening email that day, "The One Sentence that Becomes a Video Ad", launches an AdsCreator feature and earns nothing unless you are shopping for a video ad tool. Over two weeks, any line the daily earns will come from its morning editions, since every evening email in that window sold something.

When to pick differently

  • Your customers find you through Google or AI answers. Put Growth Memo in the method slot. It is written for senior marketing leaders, so expect to scale its frameworks down to one person; the free Monday essays show whether $15 a month is worth it to you. If you are weighing paying someone for this work instead, AI Marketing Agency: Three Offers and Five Questions to Ask explains what agencies sell as visibility in AI answers.
  • You want the news the day it happens. The AI Marketing Newsletter is the only daily of the four. Read the morning edition and expect the evening one to sell something.
  • You would rather listen. Almost Timely links a video and an audio version of each issue.

Subscribe to two today, one per job, and open a note dated four weeks from now. Each issue gets its line or nothing. On that date, unsubscribe from whichever has no lines.

Did this article help?

Get the best articles, carefully selected to save you time.

Read next

An AI marketing agency is a marketing firm that builds its offer around AI, as the tool it works with or the channel it targets. The label covers three different purchases: done-for-you marketing where AI speeds up the staff, automations built to keep running in your accounts, and work to get your business cited in ChatGPT or Google's AI answers. Ask what you keep if you stop paying to see which one is on offer.

ManyChat and Chatfuel both automate Instagram, Facebook and WhatsApp conversations, but they bill for different things. ManyChat charges by Active Contacts, the distinct people who interact in a billing month: 25 free, then paid tiers up to 7,500. Chatfuel charges a flat $20 a month with unlimited contacts and caps AI usage instead. Your comment volume and your channels decide which costs less.

Featured

A pivot is often just the polite word we use with investors when the first company is dead and we have decided to build another one. And that is fine. Not because failure is noble, but because luck needs exposure: every market you enter, every product you ship and every channel you test is one more surface where something unexpected can land.

Build articles

AI agents integration is the work of giving a model tools it can call to read from and act in other apps: your Stripe account, a customer's Gmail, a CRM. You wire it in one of three ways: a tool you write around the app's API, the app's own MCP server, or a platform that stores OAuth tokens for your users. Whose account the agent acts in comes first, because it decides whether you need an OAuth app of your own.

D-RAG, in the AI sense, is Differentiable Retrieval-Augmented Generation: a method published at EMNLP 2025 that trains the retriever and the generator of a knowledge-graph question-answering system together, by letting the generator's answer loss flow back into the retriever. The other results for this query are a headwear spelling, a rapper, and a 2024 LinkedIn post that reused the acronym for something else.

Projects

Brands

The essentials, by email.

What works, what does not, what I would do differently. Sent when I have something useful to say.