ProductStrip the small visual tells that make a product look unfinished
Software engineer turned curator and community builder
The room's roast of eBaza was not about features, it was about two small tells: emojis sprinkled through the interface and a colored left border on the cards that read as a default framework accent rather than a deliberate choice. Neither breaks anything, and that is precisely why they are dangerous. They are the quiet signals that whisper "template" or "AI draft" to a user who cannot articulate why a page feels unfinished. Polish at this stage is mostly subtraction, removing the stock flourishes that come free with the tools you built on, because a stray emoji and an unconsidered accent border are the fastest way to look like everyone else's first draft. Walk your own interface hunting for the things you did not decide on purpose, and delete them. The credibility you need in a payment or trust flow is built out of exactly these small, unglamorous removals.
Related advice
RaulManual curation is a moat, not a limitation
Algorithms can scrape job boards. Nobody can automate the judgment call of "is this company actually calm or just marketing itself as calm?" Raul manually researches every company in his directory, checking employee tenure, real policies, and community feedback. That process is slow, but it's the entire value proposition. What feels like a bottleneck is actually what customers are paying for.
RaulStart with a spreadsheet, not a SaaS
Raul's entire business started as a personal spreadsheet shared with friends. He didn't buy a domain, set up payments, or write a line of code until real people asked him to keep going. A spreadsheet forces you to do the work manually, which teaches you what the actual product is. If your idea can't survive as a spreadsheet first, it probably won't survive as a SaaS either.
RaulNever let a call to action lie about what it does
Raul's RCA flow had a button that promised to "see the price" and then, on click, just pushed the user deeper into the funnel instead of showing a price. That gap between what the button says and what it does is a small dishonesty, and each one teaches the user to distrust the next button you ask them to press. In a flow where trust is the currency, a CTA must do exactly what it claims. Either deliver what the label promises right there, or rewrite the label to match reality. A misleading button buys one extra click and costs you the user's belief in everything after it.
RaulProductize the recurring deadlines people already dread
eBaza's new shape is almost embarrassingly simple: a Romanian driver has three recurring obligations that all bite if you forget them, the RCA insurance, the ITP inspection, and the Rovinieta road tax, so eBaza tracks all three per vehicle and reminds you before each one expires. The product is not insurance, it is relief from forgetting, and the value is the avoided fine and the avoided last-minute scramble. There is a whole category hiding in the obligations people carry in their heads and resent: renewals, inspections, filings, subscriptions, anything with an expiry date and a penalty attached. Pull a set of those into one place, remind people reliably before the deadline, and you are selling peace of mind, which is one of the easiest things to charge for once the habit forms. Find the deadlines your users dread and miss, then become the thing that remembers for them.
Extracted from
Indie TM #11: An Intent Graph and the Hunt for Users