Before building mani, we talked to 200 founders about their marketing challenges. We expected them to say: I need more creative volume, I need faster production, I need lower costs. They said those things. But the most frequent and most emotional response was something we did not expect: I need creative that looks like my brand, not like AI.
The fear of looking like AI was the dominant theme. Founders had tried ChatGPT for copy and Midjourney for images. The output was impressive on first viewing and disappointing on closer inspection. The copy was smooth but generic. The images were striking but unmistakably AI-generated. When they used this output in their marketing, their audience responded poorly. Comments like "this looks like AI" and "is this real?" appeared frequently. The association between AI-generated aesthetics and low effort was already established in their audience's mind.
This insight shaped mani's core architecture. Brand DNA is not just a nice-to-have feature. It is the answer to the primary customer concern. When generation starts with brand identity rather than a generic model, the output looks like the brand, not like AI. The distinction is subtle but important: the same AI model produces different output when it is constrained by brand parameters versus when it is unconstrained. Constrained output feels intentional. Unconstrained output feels generated.
The second most common request was not a feature. It was a feeling: I want to feel in control. Founders felt uncomfortable delegating creative decisions to AI because they feared losing their brand's personality. They wanted AI to do the production work while they retained the judgment calls. This maps directly to the swipe-to-approve workflow. The AI generates. The founder curates. The founder's taste is the quality gate. This division of labor gives founders the control they need without requiring the production effort they want to avoid.
The third insight was about consistency, not quality. When we asked founders whether they would prefer 10 perfect ads or 50 good-enough ads, 74% chose 50 good-enough ads. The reasoning was practical: they had learned through experience that ads fatigue quickly, so volume matters more than individual quality. They did not need each ad to be a portfolio piece. They needed enough ads to rotate fresh creative continuously. This validated the volume-oriented design of mani: the daily queue, the high generation count, the emphasis on quantity within a quality floor.
The fourth insight was about time, specifically where in their day marketing fits. Most founders described marketing as a guilt-laden obligation: they knew they should do it, they felt bad when they did not, and they dreaded the time it consumed when they did. The emotional weight was disproportionate to the time investment. A 30-minute marketing session felt like 2 hours because of the anxiety surrounding it. This is why the 15-minute daily queue became central to the product. The goal is not just to save time. It is to make marketing feel lightweight enough that the anxiety disappears.
The fifth insight was about pricing psychology. Most founders did not want the cheapest tool. They wanted the most predictable tool. They had been burned by tools with usage-based pricing that generated surprise bills. They wanted to know exactly what they would pay every month, with no per-generation surcharges, no credit overages, and no hidden fees. Flat monthly pricing with generous limits was strongly preferred over per-generation or credit-based pricing.
The sixth insight was unexpected: founders wanted to see what the tool generated for other brands (with permission). They wanted a gallery of examples, not to copy, but to calibrate expectations. "Show me what mani generated for a SaaS brand, a DTC brand, and an agency." The gallery answers the question "what will I get?" more effectively than any feature description or screenshot. We built a public examples page based on this insight.
The meta-lesson from the research is that founders' stated needs (volume, speed, cost) are real but secondary. Their primary needs are emotional: feeling in control, avoiding the AI stigma, reducing marketing anxiety, and trusting the pricing. A product that addresses only the stated needs (here are your cheap, fast, voluminous ads) but ignores the emotional needs (but they look like AI and I feel out of control) will fail. The emotional needs must be addressed structurally, in the product design, not cosmetically, in the marketing copy.
Mani was built from these customer insights, not from our assumptions about what founders need. Every major product decision, from Brand DNA to the swipe interface to flat pricing to the examples gallery, traces back to something a founder told us during those 200 conversations. We did not always agree with what they said. But we always listened.
The follow-up research we conducted 90 days after launch validated several of these insights and added one new one: customers want to see improvement over time. They do not just want a tool that generates good creative on day one. They want a tool that generates better creative on day 90. The expectation of improvement is not about new features. It is about the engine learning their preferences and their audience's responses. A tool that is static, equally good on day 1 and day 90, feels stale. A tool that improves visibly feels alive.