In June 2025, French entrepreneur David Zitoun told podcast host Nathan Latka that his AI video-editing company, Submagic, had reached $8 million in annual recurring revenue with a team of 13 people and zero dollars of venture capital.
That works out to roughly $615,000 in annual recurring revenue per employee, an unusually high figure for a two-year-old software company.
As impressive as these numbers are, they show an incomplete picture on their own.
The same interview and the company's own reported metrics show roughly 15% of Submagic's paying customers canceling each month, a rate Zitoun himself calls very high. At the same time, the company converted roughly 2,500 new paying customers in a recent month.
That raises the real question: how much of Submagic's growth reflects a genuinely efficient business, and how much reflects fast customer acquisition compensating for a leaky bucket underneath it?
The answer, based on the company's own disclosed numbers and independent data on its traffic and competitors, is more specific and more interesting than either a straightforward success story or a simple churn-driven cautionary tale.
Submagic's Growth Curve
Submagic acquired its first paying customer on May 1, 2023. According to Zitoun, the company reached $1 million in annual recurring revenue exactly three months later, on August 1.
Early on, Zitoun and his co-founder also made what he has since described as a pivotal decision: abandon their initial focus on France and target the United States instead. Submagic later launched on Product Hunt, where it finished #2 Product of the Day.
From First Customer to $8M ARR in Roughly Two Years
The milestones Submagic has disclosed publicly. Note that the user figure and the revenue figures measure different things.
Every figure here is the company's own, disclosed in interviews and case studies. None of it is audited, and the dates for the user milestone are approximate.
From there, the company says it reached 1 million total users by roughly June 2024, about a year after launch. However, this figure is more meaningful when considered alongside what it isn't: a paying-customer count.
Submagic's available reporting does not consistently distinguish total registered users from paying subscribers, and no source reviewed for this article gives a specific total paid-subscriber figure. What it gives instead is a monthly conversion number and a churn rate: the Superframeworks case study puts the company at roughly 2,500 new paying customers per month against 15% monthly logo churn, and Zitoun gave Nathan Latka the same pair of figures, describing "around 2,500 customers" in the previous month and gross churn of "around 15% today" against 5,000 to 10,000 daily signups.
That's a meaningful gap in what's publicly verifiable about the business, and it means the $8 million ARR figure, while consistent across the company's own disclosures, remains a self-reported, point-in-time number rather than an independently audited one.
Three Channels That Contributed to Submagic's Scale
It would be an oversimplification to attribute Submagic's growth primarily to its affiliate program. The program played an important role, but it was only one part of a broader distribution strategy.
Affiliate program
Submagic's affiliate program launched roughly 30 days after its first customer, reportedly because the company had no marketing budget and affiliate marketing was the only channel available to it at the time.
Managed through the platform FirstPromoter, the program has generated €1,635,717 in total lifetime revenue as of the most recent case study, averaging more than €85,000 a month across a network the platform describes as 9,000-plus promoters. Converted to dollars, that's in the same range as the roughly $1.6 to $1.8 million in annual affiliate-driven revenue cited elsewhere.
Zitoun separately told Nathan Latka that affiliates accounted for roughly 20% of Submagic's revenue at the time of their June 2025 interview. That makes the channel substantial, especially given its role early in the company's growth, but still far from a complete explanation for Submagic's scale.
Free-tool SEO
This channel is larger than the affiliate program, and more exposed to a single company's decisions. Writing for CXL in December 2024, Submagic's then head of SEO, Leo Blanc, described taking the site from the bottom of the rankings to more than 150,000 daily visitors in about ten months, built on a suite of 25 free tools covering tasks around video creation rather than the editing itself.
Blanc reports that the tools exceeded their traffic target on day one, with more than 11,000 visits, and that purchases from the company's landing pages rose 120% once the free tools became the entry point. He puts $700,000 of ARR as directly attributable to organic traffic and search engine optimization. Those are the company's own figures, published under its own byline.
What the Free-Tool Channel Is Worth to Submagic
Two figures the company has published about its own search performance, and one third-party estimate of the footprint behind them.
The first two figures are the company's own and are not audited. The third is a third-party read of publicly visible link and keyword data, which is a model rather than measured analytics. Third-party traffic estimates for this site vary widely, so we have not used them to describe a trend over time.
The channel also proved vulnerable in a specific way. According to the Superframeworks case study, the company's YouTube downloader tool "attracted massive traffic but converted poorly, and Google eventually blocked the page." The lesson the case study draws is about fit rather than volume: a video compressor suits a video editing product, a YouTube downloader does not. The same source says search remains Submagic's second acquisition channel, growing 30% month over month on better-targeted tools and content.
Short-form content
The third channel is arguably the most natural fit for a company selling video-editing software: Submagic grew in part by making short-form videos about Submagic.
In a Reddit post describing the company's early trajectory, someone identifying as part of the Submagic team said the company posted a handful of TikToks in its first months that went viral, generating millions of views and bringing in hundreds of customers directly from those videos alone.
The same post described a secondary effect worth noting on its own: whenever one of those videos went viral, traffic recorded in Google Search Console spiked sharply, suggesting the videos were driving search interest in the product beyond whatever direct clicks TikTok itself sent.
Growth Without Heavy Upfront Ad Spending
What ties these three channels together is that they allowed Submagic to acquire customers without relying heavily on upfront advertising spend. Affiliates were compensated from the revenue they generated, the free-tool strategy relied on organic search, and Submagic produced short-form content using its own product rather than paying for every impression.
This appears to have been a byproduct of bootstrapping rather than a deliberate growth hack. Zitoun's own comments frame the approach as a response to resource constraints rather than a strategic preference.
The Logic Behind Submagic's Small Team of 13
Submagic's headcount efficiency can partly be traced to its organizational choices.
Zitoun has described repeatedly turning down feature requests to preserve focus. This included a mobile app that was requested within a week of launch and still doesn't exist as of writing. The reason, according to Zitoun, was that the company's core customers, small-business owners editing from desktops, weren't asking for one.
That decision runs somewhat against the broader market: Allied Market Research's segmentation of the AI video generator and editor category has the desktop platform dominant in 2023 and expected to hold that position, while the mobile segment is projected to grow at the highest rate over the forecast period.
Submagic's bet, in other words, has been to serve today's larger desktop segment rather than chase the faster-growing mobile one, at least so far.
The company has also described choosing depth over breadth early on, positioning itself narrowly as "the subtitles tool" before broadening into a fuller short-form video suite, a strategy one case study explicitly compares to how the caption-and-editing app PhotoRoom scaled globally around a single capability before expanding.
Submagic's 13-person team is often presented as an extraordinary efficiency story, but remaining small also appears to have been a deliberate operating decision. Zitoun has said the company bet on AI tools reducing the need for a large headcount and chose to increase productivity through automation rather than raise capital to hire aggressively. The company also outsourced functions such as affiliate program management rather than building every capability in-house.
The Churn Problem, and How Submagic Says It Responded
The clearest evidence that Submagic's growth hasn't been frictionless comes from the company's own account of a stall. According to Superframeworks' case study, Submagic's ARR plateaued at roughly $5 million for seven to eight months.
The company attributes this period directly to its churn rate: casual users signing up, using the tool for a handful of videos, and leaving, with neither paid advertising nor new feature releases moving that number.
The company's reported response was counterintuitive.
Rather than raising prices, as the company says advisors suggested, Submagic lowered them and launched a separate add-on product, Magic Clips, aimed at a different use case (extracting highlights from long-form video) with modular pricing that let agency customers pay only for what they used. The plateau reportedly broke after that change.
This account is consistent with a pattern in the churn data itself: agency customers, who pay for multiple seats and higher-tier plans, churn at meaningfully lower rates than entry-level individual subscribers.
An independent review of Submagic's pricing structure has identified several potential points of friction, including multi-seat costs, monthly processing limits, and Magic Clips being sold as a separate add-on. However, there is no evidence directly linking these complaints to Submagic's reported 15% churn rate.
Submagic's own public explanations attribute the churn more broadly to casual users signing up, creating a limited number of videos and leaving. Without more detailed retention data, the underlying causes of the reported churn remain difficult to establish.
Submagic's Performance Weighed Against the Competition
Submagic's efficiency claim holds up better in context than in isolation.
OpusClip has raised roughly $50 million across multiple rounds, including a $20 million investment from SoftBank Vision Fund 2, and operates with a substantially larger team than Submagic. Third-party estimates have put its 2024 revenue at around $10 million, although that figure is not directly comparable with Submagic's self-reported $8 million ARR because the metrics and reporting periods differ.
Captions, now part of Mirage, has raised substantially more outside capital: $100 million through its Series C in mid-2024 at a $500 million valuation. This was followed by an additional $75 million growth round in 2026, bringing total funding to roughly $175 million.
Outside Capital Raised, Three AI Video-Editing Companies
Funding is not revenue, and none of these companies build an identical product. The comparison shows how differently each one has been financed.
Sources: OpusClip, TechCrunch and company disclosures. Totals are cumulative announced funding, not valuations or revenue.
None of these companies build an identical product. Independent comparison reviews consistently describe OpusClip as stronger at automatically discovering clip-worthy moments inside long recordings, while Submagic is positioned as the stronger caption-and-polish specialist for content that's already been shot or clipped.
When Submagic launched its own long-to-short clipping feature, Magic Clips, in 2024, the company explicitly marketed the absence of a clip-count limit as a direct contrast to OpusClip's capped plans.
The products overlap enough to compete for some of the same customers without being perfect substitutes, which is worth noting before treating any revenue comparison as a like-for-like verdict on which product is better.
So, Was It All Just Luck and Timing?
Submagic launched into real momentum: the AI video-editing market was projected to grow from $0.6 billion in 2023 to $9.3 billion by 2033, and short-form video was already the format marketers ranked highest for ROI. Zitoun himself credits part of the company's early velocity to that wave.
Timing clearly helped, but it doesn't explain the differences between companies operating in the same market. OpusClip entered the AI video-editing market earlier, raised substantial outside capital and built a much larger team, while Submagic followed a considerably leaner, bootstrapped path. The contrast suggests there was more than one way to capitalize on the same market opportunity.
Submagic's growth instead appears to have come from several factors reinforcing one another: favorable market timing, a focused product, affiliate distribution, organic search, short-form content and an unusually lean operating model. None of the available evidence establishes a single dominant cause.
The real test now is whether Submagic's lean growth model can remain effective as the AI-powered video-editing market becomes larger and more crowded. For a different route to a similar outcome, see our investigation into how Chatbase reached $10 million ARR.
