Tony Dinh had a profitable business, a rejected half-million-dollar offer behind him, and roughly two weeks to find a buyer. Twitter was ending the free API access his product depended on, and the replacement cost would be $42,000 a month — three times his revenue. What happened next is a case study in how deadlines, not just prices, decide acquisition outcomes.

From $300 a Month to a $500,000 Offer

Tony Dinh, a software engineer from Da Nang, Vietnam, quit his job in September 2021 with seven years of engineering experience and two small products: Black Magic, a Twitter analytics tool making about $300 in monthly recurring revenue, and DevUtils, bringing in about $200 a month. He coded Black Magic himself — a product built around a Chrome extension called Magic Sidebar, according to Founded.org — and grew it into a business generating approximately $14,000 in monthly recurring revenue by early 2023, equivalent to $168,000 annualized, with profit margins near 90 percent. Then somebody offered to buy it for $500,000. He declined, as first told by AI & No-Code Exits.

The offer carried an important condition: only $100,000 was up front, with the remaining $400,000 staggered over time. In a July 2023 interview, Arvid Kahl described the offer that way, and Dinh did not correct the description. The payment schedule, guarantees, and conditions attached to the outstanding balance were never publicly disclosed. If the deferred payments depended on Black Magic’s future performance — as staggered acquisition payments sometimes do — Dinh would have retained significant exposure to the business even after selling it. And Black Magic’s future was about to become uncertain.

Twitter Turned a Profitable Business Into a Forced Sale

In February 2023, Twitter announced that free API access would end. Black Magic depended on that API; without it, the product’s core functionality couldn’t operate. Dinh wasn’t initially worried about paying — he said he would have been happy paying $100 a month, and his rule was to keep paying while the business remained profitable. Then, around March, a larger Twitter application company contacted him privately: enterprise API access would cost a minimum of $42,000 per month, with no free tier and no smaller plan, and existing free access would disappear in about two weeks.

Black Magic was generating $14,000 monthly. The proposed API bill alone would cost three times its revenue. His reaction, in his own words: “First of all, there is no way I can afford $42K per month.” He considered shutting down, which would have meant refunding customers and abandoning future income. The alternative was a fast sale — but as he wrote at the time, “If I decide to sell, I will have absolutely no negotiation leverage.” He contacted three potential buyers, received two offers, and chose Hypefury, a Twitter tool prepared to become a Twitter Enterprise customer and pay the $42,000 monthly fee itself. The acquisition valued Black Magic at $128,000 — less than one times annual recurring revenue — and moved rapidly through Escrow.com. Around March 30, Twitter cut off free API access, including Black Magic’s, but restored it soon after being informed of the acquisition.

Then Twitter announced a $5,000-a-month API plan. Dinh’s response: “I would have totally paid $5K/month if this had been announced one month earlier.” At $5,000, Black Magic might have remained viable. The lower-priced plan arrived too late to matter. Twitter changed his operating economics, but the deadline changed his negotiating economics: the business was still generating revenue, yet he had lost the ability to wait.

His Next Products Show the Other Side

While Black Magic approached its crisis, Dinh was already building. OpenAI released the ChatGPT API on March 1, 2023. He bought typingmind.com on March 2 and shipped the first version of TypingMind on March 6. Revenue reached $1,000 on March 7, $10,000 by March 10, and $22,700 across those first seven days — by his count, 99 percent of sales came from his Twitter audience. On March 11, TypingMind became Product Hunt’s top product of the day. By March 2024, it was generating more than $40,000 monthly. That same month, he described selling Xnapper, his macOS screenshot app earning about $4,000 per month from one-time purchases, for $150,000 — more than Black Magic’s sale price despite its smaller revenue.

The Numbers

What the Research Adds

Dinh’s own newsletter posts from April and May 2023 confirm the compressed timeline: he had “about two weeks to complete the transaction and transfer everything” and, given the time pressure, only managed to reach three potential buyers, receiving two offers. He also seriously considered simply shutting the business down. Founded.org independently records the Hypefury acquisition at $128,000 and notes that the product was eventually shut down. IndieHackers corroborates the broader arc: Black Magic grew from $300 in monthly recurring revenue, gave Dinh the confidence to quit his job, and was sold for $128,000 after Twitter’s API pricing change threatened its existence. What no public source discloses is the condition attached to the rejected offer’s $400,000 in deferred payments — leaving an open question about whether that money was guaranteed or would have evaporated when Twitter cut the API.

Why It Matters for Founders

• Headline price is not cash in hand. A $500,000 offer with $100,000 upfront and $400,000 in undisclosed staggered payments is a different transaction than $128,000 at closing — especially when the business faces existential platform risk. The payment terms determine whether the sale transfers risk to the buyer or leaves it with you.
• Deadlines destroy leverage. Dinh’s two-week window meant he could not run a broad process, could not negotiate terms, and accepted a price below annual revenue. Buyers who know your clock is ticking price accordingly.
• Platform dependency is a balance-sheet liability. Black Magic’s 90 percent margins meant nothing when infrastructure costs jumped from near zero to $42,000 monthly. The same risk faces any product built on a single provider’s API.
• Undisclosed earnout conditions are a red flag. If the rationale for deferred payments isn’t clear, assume part of the risk stays with you after the sale.

The Question to Ask Yourself

If your only infrastructure provider changed its terms tomorrow and gave you fourteen days to decide, would your position protect you — or would you be negotiating with no leverage at all? The answer isn’t in your revenue chart. It’s in the payment terms you’d accept today, and the runway you’ve built to say no.

Sources

• Tony Dinh Rejected $500K, Then Sold for $128K. Was He Wrong? — AI & No-Code Exits (original reporting)
• May 2023: I sold my 2 years old business for only $128K – Tony Dinh
• Apr 2023: I sold Black Magic – Tony Dinh's Newsletter
• Tony Dinh's Newsletter | Substack
• From a Regular Job to Earning $45K/Month as an Indie Hacker
• Black Magic: founders, history and what happened | Founded
• BlackMagic is shutting down – BlackMagic.so

AI Founder Stories is a weekly Feature Paper series. This feature builds on reporting first published by AI & No-Code Exits, with additional research by The Feature Paper. Browse the series at https://featurepaper.com/ai-founder-stories/.