The Future of AI-Powered Newsletters (and Why They're So Popular)

Full video transcript

All right, welcome everybody to today’s live webinar. We’re going to be recording today with Gar from Daily.ai, and we’re going to be covering the topic of newsletter AI-powered content curation and deliverability with Gar. I’ll let Gar give himself a little bit more of an intro with some of his experience, but I connected with Gar just a week and a half ago at a mastermind here in Austin, Texas, and Gar just blew me away with his understanding of how to scale and create these newsletter systems and what they’re doing at Daily.ai, and was kind enough to hop on a call with us and share a little bit about his experience. Gar, welcome in — I’m excited to dig into today’s topic with you.

Yeah, for sure, thanks for the intro, Darby. So I’m Gar, head of customer success at Daily.ai, and we are an AI-powered newsletter platform. My background is in email — surprise. I used to manage the campaigns for Nintendo, Sea Plus World Market, Oakley, and a handful of other Fortune 500 companies. Back then, before AI, we spent three or four weeks putting a newsletter together, with a full-time copywriter, full-time marketer, and full-time designer, and we could never figure out what to put in the newsletter. But now, with the power of AI, anyone can put together newsletters, and they’re extremely powerful for lead nurture.

Darby asked about how often people come to Daily.ai without a newsletter versus already having one they want to optimize, and invited attendees to share where they’re at. Gar estimated it’s roughly a 50-50 split: some customers have someone burnt out on managing newsletters manually and want to delegate to AI, while others already have a strong marketing stack — YouTube, Instagram, blog posts, podcasts — and want a newsletter to distribute and curate that content for their audience. Darby noted deliverability is the other half of the equation: there’s no point sending an email if it lands in spam or never reaches the inbox.

Gar’s high-level starting point: always be thinking about growing your email list, which is distinct from growing social followers because you own and control your list. List hygiene is critical — constant engagement, not purchasing leads, and not sending unsolicited spam, since ESPs like Gmail and Outlook are highly aware of poor-hygiene senders. On the worst-to-best scale of subscriber actions, spam reports are worst, followed by unsubscribes, then bounces, then inactivity; opens are good, clicks are better, and replies are best. Curating a list that scrubs out inactives and focuses on engaged subscribers is how you land in the inbox rather than spam.

On sending cadence, Gar said it depends on where you are in your email journey. If starting with a cold list, start small and scale up rather than blasting the full list, since a poor first impression with Gmail can tank deliverability. A preference center letting subscribers choose weekly versus daily frequency helps segment by engagement level.

On topic curation, Gar said newsletters should be conversational — even without AI, a best practice is designing newsletters that invite reader replies and feedback. At Daily.ai, they track which categories and topics users click to build a map of customer interests; a remodeling company client, for example, has content on landscaping, deck-building, and bathroom remodels, and click behavior reveals which promotions to follow up with (e.g., a sundeck discount). Darby reflected that a single newsletter might carry multiple topics, but tracking which segments click which topics can inform separate, more targeted follow-up campaigns.

Gar noted Daily.ai continuously optimizes a single newsletter per list for most customers, but for clients with genuinely distinct ICPs — like a marketplace serving both job seekers and employers — they’d run separate newsletters, since targeting multiple audiences at once effectively targets no one.

A viewer named Ron asked about having three ICPs (service professionals, marketing agencies, accounting) that share the same operational issues but need different SEO and newsletter titling. Gar explained Daily.ai’s AI writes copy explaining why curated content matters to a specific profile — showing an example from Joe Polish (founder of Genius Network), where content is reframed around his personal story of overcoming addiction and themes of leveraging time and people, and another from the remodeling company reframing third-party content as directly relevant to their audience. Darby added that this is about meeting each ICP where they are — the same article can matter for entirely different reasons to different segments, and it’s worth challenging yourself to find those different angles for each persona on your list, even manually.

On building out segment profiles, Gar recommended starting broad, personalizing the send (e.g., “Hey guys, it’s Darby, in this newsletter I’ll cover the latest trending AI tools”), and closing with an invitation to reply with feedback — direct replies are more valuable than click-tracking. Segmentation reveals itself over time as clear audience populations emerge.

Darby asked about practical, low-tech ways to track feedback if someone doesn’t have a full platform — using Google Sheets and ChatGPT. Gar noted that an overwhelming number of newsletter replies is actually a rare, good problem to have (industry-standard open rates are around 20%; Daily.ai’s best-performing newsletters hit 40-50%+, with maybe a 0.5-1% reply rate). For tracking responses, free tools like Google Sheets work, and Zapier can automate zapping responses into a sheet, which can then be uploaded to ChatGPT to summarize and extract insights. Darby mentioned using Make.com similarly for automations, and stressed being intentional about capturing responses so they don’t get lost in the inbox.

On tool choice, Gar said there’s no single right platform — Beehiiv and Substack are powerful DIY options with strong templates if you want to be hands-on, while Daily.ai leans on AI to curate and write the newsletter as a full-service solution. Darby shared that Gen AI University currently uses Dive (mentioned when they first met) and has experimented with ActiveCampaign and other tools, and that investing in the right consulting help early can prevent deliverability problems down the line.

Darby described his own newsletter-writing process: mostly written by him with Claude’s help, often starting from a morning-walk voice recording that gets transcribed and then rewritten/fine-tuned with Claude or ChatGPT — typically taking 30-60 minutes. He noted the value of finding leveraged services versus doing everything yourself, and the reality of missing sends some weeks due to competing priorities.

On where newsletters are headed, Gar said newsletters will keep getting smarter with AI’s ability to unpack large datasets — what content is clicked, time spent, and by whom — enabling automated sequential marketing (e.g., a coaching company’s audience clicking on time-management content could automatically be targeted with related affiliate offers or tools). Gar advised prioritizing content creation (podcasts, YouTube) over newsletter production directly if resources are limited, using the newsletter as a summarization and delivery mechanism that links back to the full content — citing Joe Polish as someone who invests time in podcasts/YouTube rather than newsletter-building directly.

A participant, Bruce, asked about warm-up sending — segmenting a list by recent activity (e.g., last 7 days) before sending broadly to build sender reputation with Gmail/Yahoo, and what list size makes that worthwhile. Gar said critical mass starts around 2,000 subscribers, with segments of 500; below 500, providers generally let sends through. He described this as one piece of deliverability, alongside newsletter content practices: compressing images, avoiding overtly promotional copy that trips Gmail/Outlook’s promotions-tab detection, and keeping a roughly five-to-one ratio of informational to promotional content cards (Gar aims for one call-to-action per newsletter against five or six value-add cards). On design philosophy, Gar said he prefers aesthetically designed newsletters over plain text, though both formats can successfully reach the inbox — he named Morning Brew, Substack, and Beehiiv sends he personally subscribes to as examples that are both good-looking and valuable.

On dark-mode rendering, Gar explained that email clients (iPhone, Android, Windows, Mac, various browsers) all render emails differently, and video can’t be embedded in email at all. A common pitfall is setting background colors, since dark mode/light mode color inversion can make text illegible — best practice is to avoid custom background colors.

A viewer asked for clarification on “informational cards” — Gar explained these are the non-promotional content blocks used to dilute promotional content and avoid the promotions folder, showing an example where a video-of-a-job card and a waterproofing promotion card are sandwiched among other non-promotional content.

Ron asked about a LinkedIn strategy: posting carousels that link to a LinkedIn-native newsletter (to avoid the algorithm penalty for external links) with that newsletter then linking out to his website. Neither Gar nor Darby had direct experience with LinkedIn’s newsletter feature, but Darby noted LinkedIn suppresses posts that try to drive traffic off-platform, so routing through an on-platform newsletter first is a plausible workaround worth testing; he also noted people commonly add a newsletter subscribe link directly to their LinkedIn profile.

On adding read time and a table of contents, Gar said he personally likes both — expectation-setting through a read-time estimate and previewing what’s covered upfront (similar to Morning Brew’s structure) helps hook readers into scrolling further.

Asked how to grow a newsletter from zero, Gar said the best time to ask for an opt-in is right after delivering value — after a purchase, a podcast appearance, or a useful YouTube tutorial. He cited the remodeling company’s full bathroom-remodel video documentation as an example that builds enough trust to convert viewers into subscribers. Darby added that consistency and congruency of messaging across podcast, YouTube, and newsletter content increases the likelihood someone converts, and asked about strategic partnerships with aligned audiences. Gar suggested using webinars (like the one they were running) as a value exchange to capture emails, combined with paid traffic (e.g., Facebook ads) to the webinar funnel; he noted Daily.ai got its own start doing interviews within the Jasper AI community a few years earlier. Darby added that YouTube has been a strong anchor channel for Gen AI University because content has longer shelf life there than on feeds like Instagram, and video builds a more personal connection with the audience.

On distributing via other social channels, Darby and Gar agreed that platforms like LinkedIn or Instagram (including newsletter or broadcast-channel features) are useful for driving awareness and traffic back to your own list, but the email list itself remains the owned, controllable asset — social followers and their visibility into your content are ultimately controlled by the platform, not you.

On formatting, a viewer named Ron shared a tip: a designer had converted a newsletter to PDF, which rendered illegibly across devices; Gar confirmed PDFs don’t render consistently across computer, tablet, and mobile, and that HTML is the reliable format — Daily.ai tried a PDF approach briefly and it didn’t translate well. Ron also shared his own workflow for generating newsletter and post topics: uploading a detailed ideal client profile (as a PDF or docx) to ChatGPT and prompting it to generate five topics each around pains, problems, and pursuits, then three sample titles per topic and a synopsis — describing the output as roughly 60-80% usable before a rewrite pass, enabling a month’s worth of content in three to four hours. Darby emphasized how valuable having even a basic ICP reference document is for this kind of prompting workflow.

Wrapping up, Darby reflected that regardless of where someone is in their newsletter journey — starting from scratch, optimizing an existing one, or stepping back from day-to-day management — owning and segmenting an email list remains valuable even as social algorithms shift. He named his own biggest takeaway as the importance of both segmentation/engagement tracking and framing curated content in a way that resonates with each specific audience segment, since the same content can completely miss the mark if the audience or framing is mismatched.

A participant, Julie, asked about sending to a very cold 18,000-person list — how many emails to send at a time when restarting. Gar recommended not exceeding 500 per send, and running the full list through a validation tool like ZeroBounce or NeverBounce first to scrub bounces and spam traps, then sending in batches of 500 to different segments per day. He emphasized that continually adding higher-quality new leads is more valuable than trying to revive an old, possibly irrelevant cold list, and named Mail Gun, ZeroBounce, and NeverBounce as the three tools that come to mind for identifying dead or risky addresses — while cautioning that a “clean” list from a validator still isn’t proven engaged; you only find out once you actually send to it, and never-opted-in contacts can still report spam even after cleaning.

On whether it’s acceptable to include a “feel free to share this newsletter” call to action, Gar confirmed it’s fine, since a forwarded newsletter isn’t sent from your own domain and won’t affect your sender reputation. On double opt-ins, Gar described the topic as controversial and said he isn’t an expert on the legal requirements across jurisdictions, noting that many websites treat entering an email as opting in without a confirmation click, and ultimately gave no definitive recommendation either way.

Closing out, Darby thanked Gar for sharing his insights on curating and optimizing newsletters for deliverability, and Gar shared that he’d dropped his email in the Zoom chat for anyone wanting to follow up. Darby thanked everyone for attending live or watching the recording, encouraged viewers to subscribe to Gen AI University’s newsletter, and previewed the upcoming Scale with AI Summit for November 5-7, 2024 — a free virtual event featuring more interviews and experts in this same vein.