A product with no paying customers cannot be priced like a proven cash-flowing business. It can still be attractive when the buyer is acquiring a working shortcut: a polished product, a useful codebase, a brand, distribution assets, or a wedge into a market they already understand.
The mistake is presenting a pre-revenue project as if a forecast were revenue. Be direct about what exists today and make the upside legible without promising a result. A buyer should be able to separate the delivered asset from the experiments that remain.