How AI Turned Numerology Into a Personalization Industry: Inside the Apps Generating Millions of "Personal" Number Reports

In July 2026, the AI image-and-video company Midjourney acquired Co-Star, the astrology app known for its blunt, algorithmically generated push notifications, bringing its roughly 4.3 million monthly active users and 24-person team into an AI lab's product portfolio for the first time. It was a small deal by Silicon Valley standards — neither company disclosed terms — but it made explicit something that had been building for close to a decade: the line between "AI company" and "personal spiritual guidance company" had effectively disappeared. Apps that generate horoscopes, numerology reports, birth charts, and compatibility scores at scale sit at the intersection of two real, well-documented trends — venture capital's interest in spiritual-wellness products, and the maturing of natural-language generation tools that can turn a date of birth into paragraphs of prose that read as though they were written just for you. This piece traces what's actually been reported about that industry: real funding rounds, real user numbers from real press coverage, and the honest, sometimes contradictory ways these companies describe their own technology. Some, like Co-Star and The Pattern, are openly built around algorithmic and AI-assisted content. Others, like Chani and Nebula, have staked their brand on explicitly rejecting AI-generated readings in favor of human astrologers — a split that itself says something important about where this market is headed. Numerology-specific venture funding data is thinner in the public record than astrology's, but numerology reports ride the same infrastructure and the same psychological mechanics as their zodiac-based cousins, which is why understanding the astrology-app boom is the clearest way to understand what's happening to numerology, too.

The clearest evidence that personalized spiritual content became a real business, not a novelty, is the money that flowed into it starting in the late 2010s. Co-Star, founded by Banu Guler and Ross Clark, raised a $750,000 pre-seed round and then a $5.2 million seed round in 2019 from investors including Maveron, Aspect, 14w, and Female Founder Fund, according to TechCrunch's reporting at the time. Two years later, Axios reported that Co-Star had raised a $15 million Series A led by Spark Capital, with Maveron and Female Founder Fund participating again — by that point the app had surpassed 20 million downloads and, per the same report, had been downloaded by roughly a quarter of American women aged 18 to 25.

Co-Star wasn't alone. Sanctuary, an app offering on-demand astrology and psychic readings from live practitioners (with one-on-one sessions priced between $10 and $60), raised a $3 million seed round led by BITKRAFT Ventures with participation from Greycroft, KEC Ventures, Azure Capital Partners, Advancit Capital, Gaingels, and Uncommon Denominator. Fortune reported in 2021 that investors were drawn less by astrology's scientific standing — which is essentially nil — and more by raw engagement numbers; one investor quoted in the piece praised Co-Star's founders for "rooting it in as much science as possible" even while acknowledging the underlying premise is pseudoscience. Futurism's reporting on the same trend noted that several astrology startups had "raised quite a bit of money from venture capitalists and other big-money investors," describing a wave of interest tied to the broader attention economy rather than any claim of predictive accuracy.

Not every major player took the venture-backed route, and the contrast is instructive. Nebula, operated by the Ukraine-founded company OBRIO, grew to roughly 30 million users worldwide and reportedly reached about $50 million in annual recurring revenue, according to a profile in The Next Web — repeatedly ranking as the top lifestyle app on the U.S., Canadian, and Australian App Store and Google Play charts. CHANI, the astrology app built by astrologer Chani Nicholas, took a different path entirely: according to reporting from Built by Foundry, the company took "zero venture capital," grew to over 2 million downloads and an estimated $14 million in annual revenue through subscription pricing alone, with roughly 50% year-over-year growth in both subscribers and revenue since its December 2020 launch.

The Pattern, founded by Lisa Donovan, took yet another path: it reached 15 million users, per Time's reporting, almost entirely through word of mouth after Donovan shared an early beta with 25 friends. What ties these companies together isn't a single business model — it's that each independently found a large, paying audience for individualized spiritual content delivered through an app, at a scale that would have been unthinkable for a newspaper horoscope column or a walk-in numerologist's storefront practice.

For years, the connection between "AI" and "astrology app" was implicit — companies used algorithms to calculate charts and, increasingly, natural-language tools to write up the results, without necessarily marketing themselves as AI companies. That changed in July 2026, when Midjourney — best known for AI image and video generation — acquired Co-Star outright. TechCrunch and Bloomberg both reported the deal, noting that terms were not disclosed by either side. What was disclosed: Co-Star's roughly 4.3 million monthly active users and its 24-person team, which Midjourney is reportedly bringing in specifically for their experience building a consumer-facing app, something Midjourney itself had never done (its core product has operated primarily through Discord).

TechCrunch's coverage described Co-Star's existing content model as a mix — "both AI and human writing" used to produce horoscopes, compatibility assessments, and other advice drawn from astrological calculations. That hybrid description matters: it confirms, from the company's own disclosed practice, that AI-generated text is genuinely part of how a major personalization app produces its "personal" content at scale, rather than being purely a marketing claim.

The acquisition also signals where the industry may be headed next. Reporting on the deal noted that Midjourney has been building out divisions beyond its original image-generation product, and that Co-Star gives it both a distribution channel and a proven playbook for turning structured personal data (in this case, birth time and location) into engagement-driving personalized text — a capability that generalizes well beyond astrology to almost any personalization product, numerology included.

It's worth being precise about what wasn't reported, too: no financial terms, no roadmap for AI-authored content post-acquisition, and no confirmation of layoffs or product changes. The significance of the deal, as covered, is structural rather than technical — it is the first time a frontier AI lab has bought a mainstream personalized-astrology product, not a revelation about some new proprietary technology.

Despite the "AI" framing that now surrounds this category, the reported technology behind these apps is not mysterious, and it is not uniform across companies — which is itself an important, underreported nuance. At the calculation layer, astrology and numerology apps are doing genuinely deterministic math: converting a birth date, time, and location (for astrology) or a birth date and full name (for numerology) into fixed numerical outputs — planetary positions, a "life path number," an "expression number," and so on. That part is not AI at all; it's the same arithmetic astrologers and numerologists have used for centuries, just computed instantly instead of by hand.

The part that has changed is what happens after the calculation: turning a number or a chart position into paragraphs of prose that feel individually written. Outlets covering The Pattern have described it as an "AI-powered app," built around generating personality narratives — organized into categories like instincts, growth, and relationships — from birth data, explicitly avoiding traditional astrology terminology in favor of language that reads more like a personality-assessment tool such as Myers-Briggs. Co-Star's own disclosed model, per TechCrunch, blends algorithmically generated and human-written text for its horoscopes and compatibility reports.

But the industry is not unanimous on this point, and the disagreement is telling. Nebula's leadership has publicly stated they deliberately do not use AI to generate their core horoscope and reading content, saying in comments reported by The Next Web that they "do not believe that AI can equal humans in terms of empathy, understanding, good advice" — while confirming they do use AI for marketing, retention, and internal process automation. CHANI has gone further, building human authorship into its brand identity: according to Built by Foundry's reporting, the company states plainly that "all readings, horoscopes, and meditations are written and recorded by Chani Nicholas and a team of human astrologers, with no AI-generated content," explicitly positioning this against what it calls the "auto-generated horoscope slop" elsewhere in the category.

That split — some companies leaning into AI-generated text as a scaling tool, others treating "a real human wrote this" as their core competitive advantage — is arguably the most important, least-hyped fact about this industry. It means the honest answer to "is my personalized report AI-generated?" depends entirely on which specific app or service produced it, and increasingly, companies are choosing to advertise which side of that line they're on because consumers are starting to ask.

Numerology-specific companies have received far less venture and press attention than astrology apps, and verifiable, detailed technical disclosures about how sites like Numerologist.com generate individual reports are not well documented in mainstream press coverage. What can be said honestly is that numerology reports are built on the same underlying pattern as astrology apps: a fixed, checkable calculation (converting your birth date and name into a small set of numbers) followed by a natural-language write-up of what those numbers supposedly mean — the same two-step architecture reported for astrology apps, just with numbers standing in for planets.

The psychological mechanism that makes any of this work — human-written or AI-generated — has a name and a real research history: the Barnum effect, sometimes called the Forer effect after psychologist Bertram Forer, who first documented it in 1949. Forer gave 39 psychology students a personality test he called the "Diagnostic Interest Blank," then, a week later, handed every single student an identical personality description — one he had actually assembled from a newsstand astrology book — and asked them to rate its accuracy on a 0-to-5 scale. The average rating was 4.3 out of 5. Every student believed the description had been written specifically for them; none of them had received anything unique. Psychologist Paul Meehl later coined the term "Barnum effect" for the phenomenon, referencing showman P.T. Barnum's reputed skill at crafting appeal broad enough to flatter everyone.

Subsequent analysis of the effect has identified the specific ingredients that make it work: statements vague enough to be reinterpreted to fit almost anyone ("at times you are extroverted... while at other times you are introverted"), a belief that the analysis was individually generated, trust in the authority producing it, and a bias toward flattering rather than critical language. Those are precisely the ingredients an AI text-generation system is good at producing at scale and at near-zero marginal cost — which is exactly why natural-language generation has proven so well suited to this category. A well-tuned model doesn't need to know anything true or specific about you; it needs to produce prose that sounds specific while remaining broadly applicable, and current-generation language models do that comfortably.

This is compounded by a separate, well-established finding in the broader personalization and engagement literature: content that is presented as individually tailored — even when the personalization is superficial — tends to be processed more attentively and rated as more relevant than the identical content presented as generic. Marketing and human-computer-interaction researchers have studied this repeatedly in contexts far removed from astrology, from personalized advertising to recommendation systems, generally finding that the perception of personalization, not necessarily its actual depth, is what drives increased attention and engagement. Combine that general finding with the Barnum effect's century of specific evidence about vague-but-flattering personality content, and you get a fairly complete explanation for why an algorithmically assembled numerology report showing up in your phone notifications at 8 a.m. can feel uncannily accurate.

None of this means the people who find these apps meaningful are being irrational or foolish. The Time reporting on Gen Z's adoption of personalized astrology apps points to a real, separately documented driver: Pew Research data showing younger generations moving away from organized religion, alongside quotes from users and founders describing these apps functioning as accessible, low-stakes tools for self-reflection during a period, especially through the pandemic, when many people were seeking exactly that kind of structure. A tool doesn't have to be scientifically valid to provide genuine comfort, a prompt for self-reflection, or a shared language for talking about identity with friends — which is a different question from whether its specific numerical claims are individually accurate.

The fairest way to close this out is to hold two facts side by side without collapsing them into each other. First: the underlying calculations behind numerology and astrology reports — life path numbers, planetary transits, compatibility scores — are simple, deterministic, and, from a scientific standpoint, not predictive of anything about your personality or future. Second: natural-language generation and skilled human writing can both take that simple calculation and produce text that feels genuinely individual, meaningful, and even useful to the person reading it. These two facts are not in conflict, and an honest essay about this industry has to say so plainly instead of picking one and ignoring the other.

The market itself is already having this argument, and its behavior is a useful signal. Some of the best-funded, highest-download products in the category — Co-Star, The Pattern — have built their scale explicitly around AI-assisted or algorithmic content generation, and investors have rewarded that approach with real capital, as the funding figures above document. Other prominent players — most visibly Chani Nicholas's CHANI app — have built a competing business model around the opposite claim, explicitly marketing human authorship as a trust signal against a category increasingly associated with "auto-generated horoscope slop," in the company's own words as reported by Built by Foundry. Both models are commercially viable right now, which suggests the market hasn't settled on whether AI-generated personalization is a selling point or a liability — consumers seem to want both options available, depending on what they're looking for.

What should make a reader skeptical is any claim of individual predictive accuracy — the idea that a numerology report has correctly identified something specific and unique about you that a differently-dated stranger would not also recognize in their own report. What shouldn't necessarily make a reader dismissive is the broader practice of using these apps as a structured journaling prompt, a conversation starter, or a low-stakes ritual of self-reflection — uses that don't depend on the underlying numerology being true in any falsifiable sense.

The Midjourney-Co-Star acquisition is a useful marker for where this goes next: as AI labs increasingly see "personal" content generation as a core consumer product category rather than a novelty, the pressure to disclose — clearly and specifically — how much of any given report is calculated, how much is AI-written, and how much is human-edited will likely grow. Readers of numerology and astrology content are well served by asking that question directly of any app or service they use, the same way this piece has tried to answer it about the industry's most visible players.

• Co-Star raises $5 million to bring its astrology app to Android (TechCrunch, 2019)
• Astrology app Co-Star raises $15 million in new funding (Axios, 2021)
• Midjourney acquired the astrology app Co-Star (TechCrunch, 2026)
• AI Startup Midjourney Buys Astrology App Co-Star and Is Building Its Own Apps (Bloomberg, 2026)
• Midjourney acquired the astrology app Co-Star (Yahoo Finance UK syndication of TechCrunch, 2026)
• Why Personalized Astrology Apps Are Appealing to Gen Z (Time)
• An astrology startup gets venture capital funding (Fortune, 2021)
• Venture capital firms are pouring money into astrology (Futurism)
• Ukrainian startup behind astrology app Nebula thriving despite war (The Next Web)
• How Chani Nicholas Built the $14M CHANI Astrology App (Built by Foundry)
• The Pattern astrology app offers hyper-personalised insights about your personality (Options, The Edge)
• Barnum effect (Wikipedia, summarizing Forer 1949 and Meehl 1956)