At some point you will spend three hours naming a product that took you three months to build. You will call this unproductive. You will be wrong. Naming is one of the most genuinely constrained design problems in software — not because the stakes are cosmically high, but because the constraints stack on top of each other in ways that are easy to underestimate until you are sitting there refreshing a domain registrar at midnight, watching every promising combination cost $2,400 on the aftermarket.
The builders among us tend to believe that naming is a marketing problem — which is to say, someone else's problem. Build the thing well, the thinking goes, and a name is just a label. You can always rebrand. This is approximately true if you are a late-stage company with resources for a rebranding campaign. It is approximately false if you are a small team or a solo founder who is also the marketer and the support desk. The name you launch with will follow you into App Store reviews, into customer support emails, into the mouths of people recommending you to their colleagues. Getting it wrong is recoverable. Getting it right from the start is better.
The technical layer is brutally narrow
The English-language dictionary of short, pleasant, available .com domains is effectively exhausted. A single-word .com that sounds like a product name was registered somewhere between 1998 and 2004. What remains is compound words, invented words, misspellings, or foreign words that happen to be pronounceable in English — and even those categories are thinner than they look. Type a promising word into any registrar and the available version has seventeen hyphens or ends in .io at a price that reflects the fact that the current owner knows exactly what it is worth.
The App Store adds another layer. Apple's name collision policy does not require exact matches — names that are "confusingly similar" to existing apps can result in rejection. Search your shortlist against the store before you print a single business card. The Android Play Store is more permissive but has its own duplicate-name considerations. A name that passes every other test can still fail at submission if someone got there first.
Trademark search comes last in most founders' workflows and should come second. The USPTO's TESS database is free, searchable, and will tell you within twenty minutes whether your name is already registered in your class of goods. International trademark databases are less convenient but worth checking if you plan to operate outside the domestic market. A cease-and-desist that arrives after you have already printed packaging is one of the more expensive kinds of mail you can receive.
The marketing layer is deceptively subjective
Assume you have cleared the technical hurdles. The name is available as a .com, it passes the App Store search, it is clear in TESS. Now you have to ask whether the name actually works as a name.
The most underrated test is the phone test: say the name out loud in a noisy room and ask whether someone across the table could spell it correctly without being shown. Names that fail this test — that require "with a K, not a C" or "no, it's two T's" — will cost you a measurable fraction of word-of-mouth referrals forever. If the spelling is ambiguous, some of those referrals become dead ends. This is not a fatal flaw, but it is a tax you pay on every conversation about your product for as long as it exists.
The implication test is subtler. A name implies something even when you do not intend it to. "Slack" implies looseness, informality, room to breathe — which maps well onto a product designed to feel lighter than email. "Stripe" implies a clean, precise cut through something — appropriate for a payment API. Neither of these was accidental. Consider what your name implies before someone else points out what it actually sounds like, because someone will, and it will be in a review.
The phone test: say the name out loud in a noisy room and ask whether someone across the table could spell it correctly without being shown.
Memorability correlates loosely with brevity but more strongly with distinctiveness. A short name that sounds like every other name in its category is not memorable. This is why invented words — Kodak, Xerox, Spotify — punch above their weight. They have no competitors in semantic space.
Famous placeholders you already know
Several products now considered to have excellent names launched under names that were, charitably, temporary. Twitter was called "Status" in its early internal pitch. Instagram launched as Burbn — a name that reflected its original purpose as a check-in app and that the founders wisely killed before anyone outside the team had heard it. Amazon was briefly called Cadabra, until Jeff Bezos's lawyer pointed out it sounded like "cadaver" over the phone. Google was BackRub. Yahoo was "Jerry and David's Guide to the World Wide Web." These are not cautionary tales so much as evidence that the right name can be found after the product exists, and sometimes the product has to exist before you know what it actually is.
The pattern holds: placeholder names persist until the product's identity sharpens. When you are not sure what your product is, you are not sure what to call it. The name hunt often forces a useful clarification, because answering "what should we call this?" requires answering "what is this, exactly?" Those are not always the same question, but they are connected.
The psychological constraint nobody talks about
There is a third layer to naming that gets almost no coverage in the "how to name your startup" genre, which is the psychological cost of committing publicly. Once the name is on the App Store, on the domain, on the invoice template, on the business cards in the drawer — you are that name. Changing it later is not technically impossible, but it is a rebrand, and rebrands are expensive in time, money, and the attention of customers who have just finished learning the first one.
This creates a kind of paralysis that can look, from the outside, like indecisiveness. It is not. It is the rational response to understanding that a name is a public commitment, and public commitments have weight. The founder who spends an entire afternoon on this is not wasting time. They are trying to get right a decision that will shape every piece of copywriting, every App Store description, every email signature for the foreseeable future.
The practical solution is to set a deadline. Give yourself a week, work the constraints methodically — technical clearance first, marketing tests second, gut check third — and then choose. The name you land on after a disciplined week will almost certainly be better than the name you land on after a paralyzed month. Not because time pressure produces wisdom, but because the constraints do not change, and circular loops through the same list do not generate new options. At some point you have enough information to decide, and deciding is better than not deciding, even if the name turns out to be imperfect. Most of them are.