In kicking off the new JumboMouse site, I’m refocusing my blogging around the hardcore mechanics of defining, launching and building companies. And one of the first topics every entrepreneur faces is naming a company. While much of the process is incredibly subjective, I’ll provide some tips, tricks and tools that have been useful for me over the last 10 years.
First off, you’ve got to know what market you're taking on – and how mature it is. The further along a market is, the tighter (and more differentiating) your positioning should be – and therefore your name should reflect. For something as still-wide-open as search was when Google launched in 1998, a meaningless name worked easily enough. But I’d contend there’s an inverse relationship between the simplicity of your functionality and the specificity of your name. Search is a top-level concept that could be called virtually anything – from an adjective of joy (Yahoo) to a late-to-the-game random sound (Bing) to a made up name like Google.
At the opposite end of the spectrum, if you’re trying to convey a specific solution to a well-defined market, you’re much better off with a literal brand like SalesForce, Evernote or Illustrator. When you’re pointed at a clearly defined group of users (and even more when your solution is B2B focused) there’s plenty of reason to craft something straightforward and easy-to-understand. If it’s an enterprise-level solution, don’t name it something an IT executive will be embarrassed to utter in an all-staff meeting (think, FlyingSquirrel). The slightest out-of-place brand could be the difference between mad success and mediocrity – assuming the technology is killer.
For more emerging markets like we see with many of today’s hottest and most disruptive startups (think Twitter, Turntable, Quora, Hipmunk etc.) a hipster-friendly name will go a long way to cementing your web 3.0 positioning – and most importantly, getting your first 100K users. And there’s little penalty for deriving new words from existing ones – in fact, that’s often the norm.
The rest of the psychology, sociology and gut feel of creating a name should be owned by you and your core team. Don’t listen to junk science from other industries like telecommunications, automotive or pharmaceuticals that all have multimillion-dollar naming agencies. Make it authentic, and ensure you’re behind it 1000%. Remember, it’s the story of YOUR company – and shame on anyone who answers “because the domain was available.”
But let’s be real, the availability of the domain is among the most critical aspects of a name. And while we’re talking marketing channels, you should lock-up matching usernames on Twitter and Facebook – and get to work tweeting and driving enough likes to secure your vanity URL at Facebook. The rules of the road for URLs are pretty wide open, but I’ll share a few specifics I live by:
- Avoid purposeful misspellings of known words such as Klout. Until they reach widespread adoption, every single intro for every person in the company goes like “it’s like clout, only with a K.” And then there’s the issue of users misspelling your domain on direct visits. Unless you can own the correct spelling of an existing word, my opinion is move on.
- If you’re going to make up a word, focus on phonetically simple words and think about Scrabble points. In general, the higher the derived Scrabble score, the more unique the name – Zynga being a great example. Also, avoid homophones and homographs at all costs. In fact, stay away from anything in the blue, yellow or the overlaps in this Venn diagram on Wikipedia. Simplicity and clarity are the keys.
- Alternate TLD (top level domains) from non-US countries can be good options but only if they work in two ultra-specific ways: a) it’s an existing word that spans the dot, i.e., mixtu.re; or 2) if it’s rock-star short and the characters to the left of the dot are a known word, i.e., mix.ee.
- If you’re going to use an alternate TLD, register a matching .com like get<brand>.com or go<brand>.com and redirect that to your site. If you’re crazy lucky enough to hit Twitter-ish popularity, you want something a decreasingly web-savvy group of users can comprehend and reach via direct address.
- If there’s political unrest where your country’s domain is registered, go ahead and renew for 5 years at the first sign of trouble. Most problems (like the well-publicized issue for letter.ly) come when a domain expires. Once you’re registered, it’s normally set and forget.
- Base your name on one word that best represents your market position – ideally a word that’s less than five letters. From there, you can explore derivations like -rama, -ology, -ed, -ous, -ly, and hundreds more. For example, one of my startups is GARKD, which is based on the single word ‘gark’. It took a lot of cycles to get to that one based on the complexity of the technology behind the company, but spend the time – it’s well worth it!
- If you’re going to embrace texting vernacular by using GARKD instead of GARKED, you’d better own both and redirect the -ED. Following launch, I expect Clicky to show me that a fair percentage of my direct traffic is redirected to -D from -ED. Don’t out-trick yourself by preventing users from finding you at the exact moment you need them the most!
- And for the love of little puppies, PLEASE run a Google search on your new name before buying domains. You need to see who owns the first 2-3 pages of results, and whether a similar name already exists in close proximity to your market (the worst case scenario) or if an aggressive SEO buyer has a product in a similar space (almost as bad as a direct competitor having a similar name).
So once you develop the core of the name, it can be mind-numbing exercise to drive out the final derivation. And if you’re using a non-US domain, be prepared for a longer registration process (up to a month in some cases) and more fees (average seems to be ~$70/year, but varies widely). I’ll also share from experience that allowing random name availability to get you away from your core concept is nothing but a distraction. Pick one word (like ‘mix’ in my case) and stick to it. Once that’s locked in, here’s a set of tools to grind out your final name:
- Name Generator: I recently saw a tweet from Mark Suster about a great URL site called Panabee. It allows you to combine two words (or just use one as I prefer), and it applies a bunch of naming logic out ahead of a domain lookup function. Things like appending ‘go’ or ‘get’ to a word as I discussed above. It gives you literally 35-40 derivations of a word, and an instant status on domain availability. It’s hot!
- Scrabble WordFinder: I like A2ZWordFinder the best given it allows virtually unlimited use of # to wildcard letters. Take your core word and start by adding 4-5 wildcards. You can also control whether the core word is locked at the beginning or end of the candidate words – or even drive the exact placement inside the word. And finally, I prefer to sort the list alphabetically, but my most important filter here is length. The downside of this approach is you only get Scrabble-cleared words, which are often taken – so get creative with your own conjugations too.
- Text Editor: I use Word because I’m a kid of the 90s, but anything that can do Find/Replace All will suffice. Once the candidate list is built, copy and paste it into a new document. Then run a Replace All function to convert commas to hard returns. If you’re searching alternate TLD (like .ly or .ee), you can increase the complexity of the Replace to create the list with the appropriate syntax on the URLs. Or if you’re looking for just .com, then change the syntax to append that on each line. (Pro tip: only tackle one extension at a time given you may need to programmatically modify hundreds of lines.)
- Bulk Domain Lookup: Again, there are lots of options but I prefer the tool at 101 Domain given they support the widest list of country-specific TLDs. Once my candidate list is generated, I slam a maximum of ~500 domains at a time into their bulk search tool. I’ve found anything over ~500 creates some timeout errors. Just make sure the syntax is dead on, and there’s one domain per line in your final file.
- Country-Specific WHOIS: If you’re considering non-US domains, there’s one more critical step. Once you’ve shaken down what 101 Domain tells you is available, you still need to hit a domain-specific WHOIS to make sure everything’s cool. I found out the hard way that 101 Domain works hard to provide the most recent status, but things get weirder the more obscure your extension is. You can Google for these country-specific IANA pages, but in general the following link works when plugging the county-specific extension in the URL where I show **: http://www.iana.org/domains/root/db/**.html
The last bit of advice I’d offer is to have your own trusted inner circle to reality-check you during the naming process. For some, that’s your startup co-founders and for others it’s a small group of people who know you (and startups) inside out. Much of the experimentation and grinding process is best done solo (and immersively in my case), but having some kind of external view on your thinking is critical.
And if you have the opportunity (and predisposition), do some light focus group testing. I’m not a fan of wasting time in the whipsaw of medium-sized focus groups, so keep it small and quick with real customer targets. You’re looking for genuinely negative responses based on attributes you haven’t previously considered, so I’d ignore any ill-formed feedback or slight discrepancies. Selling your product to customers is Job One!
Good luck in the trenches, and I’d love to hear about any tools you use to crank out company names!