TECH Signal 484
Revenue drops 24% and new subscribers fall as AI tools erode demand
↗
Illustration only Photo by Declan Sun on Unsplash
The author's subscription app experienced a 24% revenue drop and a steady decline in new sign-ups since February 2026, which he links to reduced public development, AI-based alternatives, and weaker marketing.
For engineers building subscription-based services, the case shows how a halt in public development and transparent metrics can reduce visibility and user trust. It also illustrates the competitive pressure when free or low-cost AI tools replicate core functionality, pushing paying users toward cheaper options. Understanding cancellation reasons and adapting marketing channels become essential levers to stem churn and stabilize recurring revenue.
Written by elseif from the cluster below · every claim links back to a sourceThe three things worth knowing
Revenue fell 24% and MRR dropped 12% compared to the February 2026 peak.
New subscriber count declined from 191 in January 2026 to 45 in August 2026, while MRR growth turned negative after April.
The author attributes the trend to stopped public updates, AI-driven bank-statement converters, and weaker outreach, and plans to improve cancellation follow-ups, content marketing, and app performance.
THE READ
What the cluster adds up to.
Revenue and Monthly Recurring Revenue are down significantly since the February 2026 peak, with revenue off 24% and MRR off 12%. New subscriber counts fell each month from 191 in January 2026 to 45 in August 2026. MRR growth turned negative starting in April 2026, showing losses of thousands of Hong Kong dollars each month. The author notes that the rate of cancellation remained roughly constant while the inflow of new users shrank.
He attributes part of the decline to stopping public development and no longer sharing revenue graphs on Twitter after September 2025. The author suspects that free AI chatbots are being used for bank-statement conversion, reducing the need to pay for his service. He also sees a competitor that may be outperforming his product because he built the app in public. In response, he plans to send follow-up emails based on Stripe cancellation reasons, experiment with content marketing, and try physical ads in Hong Kong as well as podcast and YouTube sponsorships.
Improving the speed of scanned-document processing and adding progress indicators are cited as product-level fixes derived from user support feedback. The author acknowledges that AI tools work well for users with few pages of statements, making his paid tier less attractive for that segment. He expresses reluctance to take a traditional job, citing dissatisfaction with managing AI agents and reviewing pull requests at companies. These factors together limit the effectiveness of simply increasing marketing spend without addressing the underlying value proposition.
Because only one source (the Hacker News post) presents this narrative, there is no external corroboration of the claimed causes or outcomes. The analysis therefore relies on the author’s own data tables and personal reflections as the primary evidence. Engineers should treat the story as an illustrative case rather than a validated industry trend. He notes that marketing experiments have not yet yielded responses, suggesting that outreach alone may not solve the problem.
Written by elseif from the cluster below · checked for specifics the sources never containedTHE CLUSTER