How 6AM City Uses AI Workflows to Greatly Speed Up Newsletter Launches and Profitability

Launched in 2016 in Greenville, South Carolina, the local newsletter 6AM City has expanded to 31 U.S. markets

6AM City has developed a systematic approach to predicting market viability and accelerating expansion timelines. Using artificial intelligence and data analysis, the company has grown from a single Greenville newsletter to operations in 31 markets, reaching more than 1.5 million daily readers and generating about $10 million in annual revenue. Their process has vastly reduced typical market entry and profitability timelines. The business model has expanded beyond advertising to include deeper partnerships with advertisers, e-commerce, and software licensing.

6AM City deploys algorithmic precision. COO Ryan Heafy revealed their systematic approach in an interview with The Audiencers: "In terms of population density, we overlay our target demographics over the census data. Our readership is female leaning and wildly age diverse. Then we ask: Could we achieve 100,000 subscribers in this market? For us to be profitable and sustainable, we target a minimum of 50,000 subscribers."

Greenville, South Carolina, was 6AM City’s first city, and it took the company almost two years to go from zero to 50,000 subscribers and become profitable. In some of its newer cities, the local newsletter reaches the magic number of 50,000 subscribers after only three months. 6AM City is now active in 31 markets and reaches more than 1.5 million people daily.

For each of the markets, 6am City operates with about 3.5 headcount per city (2 editors and 1-2 people for marketing and sales). According to Heafy, it costs about $250-300k to launch a new city in year one, which nets out at zero. After two years a new newsletter is expected to generate a six- to low-seven-figure revenue.

The company engineered what Heafy calls a "pride of place" metric—an AI-powered assessment algorithm that analyzes inflows and outflows of people, retail spend, charitable giving, demographics, and more. It also includes a close look at social media sentiment around new businesses launching. This predictive modeling enables them to identify markets with the highest probability of success before investing a single dollar.

But the most striking metric is revenue acceleration. "We're seeing significantly rapid increases in audience development in all of our markets: more markets faster, faster growing audiences, faster revenue on day one. It used to take us six months to get business. Now we're generating revenue on day zero," Heafy said four weeks ago at The Newsletter Conference in New York.

Process Automation For Scaling: Behind 6AM City's rapid scaling lies a strategic technology approach. The company has developed systematic process documentation and training automation that eliminates the training burden on the rest of the team.

This isn't just about efficiency—it's about scaling human expertise through automated workflows. Their systems enable them to transform editorial teams from zero to operational in exactly two weeks, a timeline that would be impossible through traditional training methods.

Performance Analytics Engine: 6AM City deploys sophisticated analytics platforms that automatically optimize delivery timing, conduct A/B testing, and analyze subscriber behavior patterns. These systems provide real-time performance insights and automatic adjustments to maximize engagement across their growing portfolio of markets.

6AM City discovered that avoiding controversial topics might reduce enragement engagement but actually expands your addressable market. And it appeals to advertisers who don’t want to see their ads next to undesirable content. "We're designed to be a marketing engine for the cities. We curate the best of what's going on there about how you can become a better, more engaged and involved citizen, whether that means telling you quickly how you can join boards, nonprofits, participate in charity events, or where to go have a cocktail on a patio in downtown. It's more of that lifestyle approach," Heafy explained at the local business event Endeavor Greenville in January.

6AM City produces some original content and also leverages AI-enhanced content curation from local communities to burst out of filter bubbles. "I have friends that love food and wine festivals, but they're not part of the Hispanic community, so they never see the Hispanic Food Festival in their preference set. By not filtering the content down so much, we are opening up people's eyes on how they can join other elements of their community. And we're seeing great reward in that, and economic impact locally," Heafy illustrated in Greenville.

6AM City is able to build on its proven track record based on its tech stack by licensing the technology to other local news organisations, content creators and brands. The SaaS (Software as a Service) offerings include:

6AM City's AI-driven model represents more than operational efficiency—it's a fundamental reimagining of how local media can scale profitably. The flip side: The kind of approach that 6AM City excels at has undeniable value for local communities. But hard-hitting and fundamentally important local investigative journalism this is not. That requires more than streamlined operations and algorithms - more boots on the ground and building long-term relationships with trusted sources.

But the bigger disruption may come from their platform strategy. By developing AI tools for external creators, 6AM City is positioning itself to capture value from the broader creator economy while democratizing access to sophisticated local media technology.

It is not quite alone with this approach: The Canadian media company Village Media licenses its in-house AI technology to around 120 publishers across Canada and the US. I wrote about their strategy five weeks ago.

Based on 6AM City's AI-driven growth strategy, local news publishers should think about these key tactical approaches:

AI tools used for this newsletter:

Last week’s newsletter about the Wall Street Journal originally contained some errors. (Thank you, Tess Jeffers, for pointing those out to me.) The web version is now updated and corrected.

Apologies, that shouldn’t have happened. But it did. And so I reverse-engineered and improved my routine to avoid a repetition. The mistakes were human and procedural and not AI fabrication, therefore there is an actual lesson here in how to use AI tools more safely. Because my lesson, of course, can’t be to stay away from them.

The errors happened in two ways:

The information on the Narrativa website looks current but is actually outdated. WSJ doesn’t use Narrativa anymore. (That error had nothing to do with AI.)

The second mistake I made is maybe also a warning for others who use AI to parse audio from conferences.

Last week I used the free and limited version of Claude 4.0 Sonnet for the first time to write the first draft of the newsletter (I have since then upgraded to the Pro version). I thought I had built a fool-proof mini-RAG by allowing Perplexity and Claude to only use select sources. (I disallowed more than 20 other sources which I deemed outdated, irrelevant or redundant.) I ended up with:

The errors came in through the backdoor of the IJF panel transcription. Tess was of course not the only speaker on that panel. Otter.ai transcribes and separates speakers reliably and I had prompted Perplexity to query the transcription using only her active parts. Then I prompted Claude to write a first draft, using only the Perplexity answers, the Narrativa info and the transcription for more details. But I neglected to also prompt Claude to ignore the other speakers’ parts. So some details in last weeks newsletter were actually not about the WSJ and I missed that in proof-reading.

Learnings: