How to test a product name
Test a product name by checking what people can do with it, not whether they like it: can they recall it the next day, spell it after hearing it, say it confidently, and guess roughly what it's for? Run the same tests on two or three finalists side by side, with people who resemble your customers.
By Phil Martin at Incyworks Ltd · Published · 7 min read
Why this matters
Once you're down to two or three names, it's tempting to ask friends which one they like and go with the most votes. That's usually the weakest test you can run. People judge unfamiliar names harshly, they say what they think you want to hear, and taste tells you little about whether a name will work. Good naming tests measure what people can do with a name.
What naming tests can and can't tell you
Simple tests are good at finding problems: names people can't spell, can't say, forget overnight or read the wrong way. They're poor at predicting which name will “win” in the market, because a name gains meaning from the product and the marketing around it. Use tests to rule names out, not to crown a winner on a show of hands.
Why “do you like this name?” is a weak test
Almost every new name feels odd at first, and familiar-sounding names get an unfair advantage. Liking is also shaped by the order you show names in and by who asks. If you do ask about preference, ask last, and treat it as one signal among several.
The recall test
Mention your finalists in passing, without writing them down. The next day, ask which names people remember. Names that survive a night's sleep are the memorable ones.
The spelling test
Say each name once and ask people to type it. Count the different spellings you get back. Every wrong spelling is a customer who can't find you or an email that goes astray.
The pronunciation test
Show each name written down and ask people to read it aloud. If you hear several versions, people will hesitate to say it to others, and word of mouth suffers.
The association test
Ask “What's the first thing that comes to mind?” You're listening for useful images and feelings, and for anything negative, embarrassing or linked to a competitor.
The comprehension test
Ask “What do you think this does?” You don't need people to guess exactly. You do want to avoid names that point them firmly in the wrong direction.
The comparison test
Show the finalists together and ask which best fits a one-sentence description of your product. Rotate the order between people so the first name doesn't always win. This is more useful than rating names one at a time; see how to compare product names.
The small-audience test
Ten to fifteen people who resemble your customers will reveal most serious problems. Friends and family are convenient but too kind, and colleagues already know too much. If you can, include people from each market and language you'll sell in.
What to do with conflicting feedback
You will get contradictions. Weigh problems above preferences: one person who can't spell a name is a signal; one person who doesn't like it is an opinion. If a name fails a practical test with several people, drop it. If the remaining names split opinion, go back to your brief and choose the one that does its job best.
Worked example
Two finalists go to twelve people. Name A gets more “I like it” votes, but five people spell it three different ways and only four remember it the next day. Name B is less loved at first, but nine remember it and everyone spells it correctly. Name B is the stronger name; the liking will come with use.
Common mistakes
- Testing only with friends and family.
- Showing names in the same order every time.
- Asking leading questions, such as “Isn't this one catchy?”
- Testing names before checking availability, then losing the winner to a trade mark search.
Checklist
- Availability checked first, so you only test usable names.
- Ten to fifteen people who resemble your customers.
- Recall, spelling and pronunciation tests run for every finalist.
- Associations and comprehension asked as open questions.
- Names compared side by side, in rotating order.
- Problems weighed above preferences.
How NameClicked can help
NameClicked can't replace testing with real people, but it can make sure you only test strong names. Its assessment flags likely spelling and pronunciation problems, unwanted meanings and similarity to competitors before you show anyone. Check your finalists first, then use these tests to choose between them. For the criteria behind it, see how to evaluate a product name.