How to name a SaaS product

In December 2010, Patrick and John Collison's payments startup still didn't have a real name. Their internal codename was /dev/payments - a Unix directory path, chosen because the product was a developer-facing API and that seemed fine at the time. Except you can't register a company in Delaware with slashes in the name. So the legal entity became SLASHDEVSLASHFINANCE.

Not exactly what you'd put on a pitch deck.

Greg Brockman, their CTO, wrote a script. It scanned for available domain names - random nouns, each one checked against a registrar, each available one triggering an automated email to the owner. After hundreds of emails, he got a reply about stripe.com.

The team spent weeks debating two finalists: Stripe and PayDemon. Yes, PayDemon was a serious contender. They set a hard deadline - December 20, 2010. If they couldn't decide by then, Stripe would win by default. The deadline came, there was still no consensus, and the product launched as Stripe.

That's what naming actually looks like. Not a brand strategy workshop. A script, an arbitrary deadline, and a word that happened to feel right.

The question is what "felt right" actually means - and whether you can engineer more of it when you're staring at a blank domain search box.

Why SaaS naming is its own problem

Consumer brands have it easier in one specific way: they control the context. A sneaker company picks a name, buys billboard space, and builds the association over time. SaaS doesn't work like that.

Your name shows up in ways you don't control - Slack channels, pull request descriptions, invoices, browser bookmarks, developer documentation, the autocomplete when someone types your URL. It lives inside other products, other people's workflows. It gets said out loud in meetings by people who've never been to your website. That's a different burden than a consumer brand carries.

Then there's the market reality. Around 42,000 SaaS companies exist worldwide, with roughly 1,500 new ones launching every month. Whatever word you're considering, there's a decent chance someone already owns the .com domain, has a trademark in your category, or is building something similar with a confusingly similar name.

The constraints stack up fast: every good English word is probably registered as a .com, trademark conflicts can kill a launch, you're building something global so the name can't accidentally mean something bad in another language, and it needs to work in all those contexts I just described - subdomain, API prefix, Slack shorthand, invoice header.

So how do successful companies do it? Mostly by falling into one of four strategies - usually without explicitly choosing to.

The four naming strategies, with twelve real examples

Every SaaS name falls roughly into one camp. Knowing which camp you're in - and why - is more useful than any name generator.

Descriptive: tell them exactly what it is

The name communicates the product's function. No metaphor, no invention.

Descriptive names work best when there's a twist - a compound, a slight variation - that makes the description ownable. Pure dictionary descriptions like ProjectManager, InvoiceCreator, or SalesTracker are available as brandable names only in your dreams. Everything self-descriptive is either taken, generic, or legally undefendable.

Metaphorical: borrow an image from somewhere else

Take a word that means one thing in the world, and let it mean something different in your product context. The word's existing associations become part of your brand before you've spent a dollar on marketing.

The main advantage of metaphorical names over descriptive ones: they're almost always trademarkable in your category. Nobody had a prior claim on the word "Stripe" in financial software - it was just a word - until the Collisons made it a brand.

The trade-off is that you have to build the association. The word does nothing on its own. That's fine if you have scale and resources. It's harder if you're launching into a market that doesn't know you yet and can't afford brand campaigns.

Invented: make something up

Build a word from scratch, or fuse existing words into something new. Done well, you get a name that nobody else has, that's easy to trademark, and that you can define entirely from scratch.

The challenge with invented names: cold-start problem. A word that means nothing inherits no associations. You have to build them entirely. For developer-focused B2B SaaS - like Zapier or Vercel - this is actually less of a problem, because technical communities adopt new vocabulary quickly once a product gains traction. For consumer-facing or non-technical B2B products, the education cost is higher.

Compound and hybrid: fuse what exists

These sit between descriptive and invented. Two real words - or parts of words - fused into something new. Descriptive enough to explain itself, distinct enough to own.

Three questions that narrow the field

Nobody picks a naming strategy cleanly. You're usually choosing between a name that's available and one that feels right. But there are questions that help sort it.

How technical is your buyer?

Technical buyers - developers, engineers, data teams - pick up new vocabulary fast. Linear, Vercel, and Zapier built large audiences with names that meant nothing at launch. If you're selling to non-technical buyers (SMB owners, teachers, HR teams), a name that does more explanatory work upfront saves you marketing budget later.

How crowded is the naming landscape in your category?

When every competitor has a forgettable descriptive name, a metaphorical or invented name stands out immediately. In 2010, project management tools had Basecamp, ActiveCollab, and Teamwork - all compound or descriptive. Asana stood out the moment someone heard it. If your category already has distinctive brands, descriptive might actually help - it signals clearly what you do in a sea of clever-sounding names.

How much will the product change?

A descriptive name anchors you to the current function. If the product is likely to expand - from email tool to marketing platform, from payments to financial infrastructure - pick something with room. Notion isn't trapped in "notes." Stripe isn't caged by "payments." Canva has gone from design tool to full creative platform without the name ever getting in the way.

The domain reality check

Of 550 SaaS companies analyzed, 95% use .com domains. That's still the baseline expectation for buyers and investors - it signals permanence and establishment.

But something is shifting. Around 33% of VC-backed startups now launch with alternative TLDs, up from just 15% in 2014. The companies pulling this off aren't doing it by accident - they're doing it with strong enough products that the domain stops mattering.

Modern TLDs that hold up:

• .app - Linear.app works because Linear's community is loyal and the product earned that trust
• .io - still carries developer credibility; Customer.io, Framer.io, and dozens of others built real businesses here
• .ai - commands premium pricing (five-figure domain sales are common now), but signals AI products clearly
• .so - Notion started at Notion.so, then eventually bought notion.com as resources allowed

The pattern in every successful non-.com launch: the product was good enough that the domain stopped being the main story. You can't plan to be that good. What you can do is pick a name where the .com is obtainable - even if it requires a small modifier or a direct approach to the current owner.

Three things that will date your name

Generic suffixes (-ly, -ify, -hub, -stack) are everywhere. Expensify, Calendly, Recurly - all coined when those suffixes felt fresh. Now they read as "SaaS from 2015." Not disqualifying, but not a signal you want to lead with either.

Deliberate misspellings (Lyft, Fiverr, Tumblr) made sense at the time - they guaranteed domain availability and looked distinctive. They've since become their own cliche. They also create friction in word-of-mouth: "is it spelled normally?" is a question you don't want to answer in every pitch.

The most consequential mistake, though, is picking a name that locks you into your current scope. A purely descriptive name anchors you to the product's function right now. When you expand - and if you build something successful, you will - the name either becomes misleading or requires a full rebrand. A rebrand costs $50K-$500K minimum, and means resetting whatever recognition you'd built. Better to have a name with room before you need it.

How the process actually goes

The strategy is the hard part. After that, the process is fairly mechanical.

Generate 50-100 candidates. Fifty minimum before you evaluate anything. Use a thesaurus, dig into etymology, explore metaphors from adjacent industries, try portmanteaus. Most will be bad. Good - that means you're generating enough volume.

Filter for availability first. Check .com (and any TLDs you'd consider), the USPTO trademark database, and app store availability. This cuts 80% of the list before any taste-based debate.

Tell five people your top three names in casual conversation - "I'm thinking of calling my product X." Not as a formal test, just in passing. You'll immediately feel the difference between names that land and names that need explaining.

Then decide. Analysis paralysis is a real risk in naming because there's no objectively correct answer. Stripe got its name by deadline. A good name launched beats a perfect name still being debated six months from now.

References and sources

• Smart Branding - 500+ SaaS companies domain name analysis - 95% .com adoption rate, naming patterns across 550 companies
• Startup Archive - Stripe naming story - PayDemon vs Stripe decision process
• Smart Branding - How did Stripe get its name? - Greg Brockman's domain-hunting script and the deadline story
• Quartz - Slack is secretly an acronym - Slack backronym origin and Stewart Butterfield's internal message
• namemy.app - SaaS naming guide - analysis of 100 successful SaaS companies, 82% single-word stat
• Backlinko - SaaS statistics 2026 - 42,000 SaaS companies, 1,500/month new startups
• Dynadot - Best domain extensions for startups 2025 - 33.1% of VC startups now use alternative TLDs