The lifetime deal that grew Pretty Prompt's subscription business

Charlie Day knew the AppSumo campaign could stress-test his systems. He didn't expect it to send his MRR up.

The seed round 

Before Charlie Day (no relation to the Always Sunny guy) said yes to AppSumo, he called past founders who'd already been through it. He wanted to skip past the pitch and get right to the real stuff.

What he heard surprised him. "You could almost see AppSumo as an alternative to seed investment," Charlie says. "The customers become like investors."

He and co-founder Ilai had built Pretty Prompt, a Chrome extension that improves your outputs from ChatGPT, Gemini, and Claude by rewriting your AI prompts in one click. They launched it in May 2025. 

A few months later, AppSumo reached out via LinkedIn. Charlie knew that Pretty Prompt worked. What he still needed was proof that people would pay for it at scale, and that the systems could handle heavy load.

5,200+ purchases and 170 reviews (4.86 out of five stars average) later, Charlie got a lesson about how lifetime deal buyers behave that he didn't see coming.

Running the numbers first

Pretty Prompt is an AI product, which means every prompt refinement costs money. Each one-click improvement triggers an API call to OpenAI, Anthropic, or Google. That’s a per-use cost that doesn't stop when the lifetime deal closes.

"The equation we had to work out," Charlie says, "was whether this was going to be really good for two or three months but ultimately bankrupt the company in six months?"

To figure this out, the team tracked token cost per user with enough precision to run a worst-case scenario: every user maxing their monthly credit allocation, every month, with no additional revenue coming in. Even under those conditions, the runway came out to about 18 months.

That was the number that gave them enough confidence to move.

In reality, usage looked nothing like the worst case. Credit consumption followed a familiar pattern: a small percentage of users accounted for most of the usage, while the majority stayed well below their limits. About half of Pretty Prompt's 20 most active users today are AppSumo buyers. The "nightmare scenario” users turned out to be a couple of enthusiasts, not a liability.

Here’s Charlie's advice for any AI founder running the same calculation: get your observability in order first. Know exactly what each user costs you at the token level. Then build your worst case from the actual numbers instead of your gut feeling.

The "collateral marketing" effect

While running his lifetime deal on AppSumo, Charlie didn’t expect the campaign to grow his subscription business at the same time.

He calls it the "collateral marketing" effect. AppSumo's social reach, like Facebook groups, ads, and organic posts, put Pretty Prompt in front of people who had no interest in a lifetime deal. Some people wanted unlimited usage or team plans. Some just preferred paying monthly. They found the regular subscription page and signed up there instead.

"If you're on AppSumo, you're probably not going to be buying monthly anyway," Charlie says. The two audiences barely overlapped. AppSumo reached people who wouldn't have found Pretty Prompt otherwise, and some of those people became paying subscribers on the spot.

Once the campaign ended, the subscriptions kept coming.

For weeks after the lifetime deal closed, Pretty Prompt saw a tailwind of new monthly and annual signups. Charlie's working theory was that AppSumo's marketing takes time to move through all the channels. By the time some people saw it, the deal had ended. So they signed up for what was still available.

Reviews were a similar surprise. Charlie expected maybe 10 or 20, and ended up with 170. Now they’re testimonials on the Pretty Prompt website.

Black Friday was the moment Charlie knew their strategy was working. He and Ilai had bookmarked the user count on their phones and checked it every morning. During Black Friday weekend, they went to bed and woke up to 2,000 new users. Then it happened again the next night. (By the way, checking your phone at 7 am for user numbers is either the sign of a great campaign or a concerning habit. Probably both.)

In the end, Charlie got his proof of scale and even higher MRR, while the 5,200 people who bought in got an AI refinement tool that keeps getting better. Not a bad “seed round,” if you ask us.

‍