How to design naming research for the real world

Naming research can be a hot topic for clients who want to know that their name is going to say or do all the things they want and need it to do. However, “testing” or “validating” names is tricky and should be done with great caution and care.

So, when research is called for, we must tread carefully. Here are some techniques to help make naming research as helpful as possible – without having it kill all the good names.

Great names aren’t built for the logical mind, so naming research shouldn’t be built that way either

Ownable brand names need to work at an emotional, intuitive, non-logical level that taps into heuristics or metaphors, giving a gut understanding of what a thing is or what a company does without being overly accurate or descriptive. In behavioral economics, you would ascribe the lightening-quick understanding and projection onto a brand name to what Daniel Kahneman calls the “System 1” thought process. These kinds of names garner understanding at a gut level, without necessarily eliciting understanding at a rational, accurate, or logical level. Think Google, Nest, Nike, Patagonia, Amazon, Zulily, or Etsy. Not one of them says what the company does, yet each is utterly emotionally evocative.

We want names to resonate with System 1. This is how most high-profile names (and, more likely to be trademarked, suggestive or even fanciful) are received in the real world. You see a name on a package, in an ad or on a website, and your System 1 immediately forms an opinion about that name based on heuristic information. When you first saw Google your brain might have quickly run through a burst of images or associations such as “googly eyes,” “fun sounding,” “big, enormous, maybe even infinite,” etc. And these heuristics taken from different parts of the name didn’t really tell you what Google is, but it gave you some good ideas about what Google may be about. That’s how most great high-profile names work.

And yet, name research generally asks that respondents engage and turn on their “System 2” thought process to evaluate a name. System 2 is described by Daniel Kahneman as the more logic-based thinking process of our brains. If a group of people were asked to tell someone what kind of company you think Google is and what it does, chances are each person’s System 2 would have some wildly different and “off” answers. Because System 2 wants to understand things rationally and logically, it’s not a great evaluator of a high-profile name. In fact, in this fictional Google name research study, had respondents seen the name “Google” and an alternate name like “Info Search,” and told that one of those names would be used for a box on the Internet where you can type anything and it will call up all available and searchable information from the entire Internet, respondents would likely say (or at least their System 2 would say) that the name “Info Search” is far superior to that crazy sounding gobbledygook of a name “Google.”

Our System 2 loves descriptive names that say exactly what something is. Yet, our System 1 (the one judging all the high-profile names in the real world) lights up when a name hits an emotional chord.

But, what if you need to check against a disaster with the name or want customers to help give you insight into which name choice will help you the most as you go to market and grow?

Qualitative research is the preferred method for researching names.
Qualitative can never achieve the standards or numbers of being a true “test.” And self-reported opinions aren’t really a test of a name. It’s best to think of qualitative as being a resource for guiding decision-making, not as a contest to identify a winner. Chances are, your respondents will like the most descriptive option available to them, but they can also tell you that one of your fanciful options makes them feel calm, in control, and reassured. If those are the emotions you want to anchor your brand to, then that name is a great choice! 

When done right, qualitative can help you hit a homerun without resorting to boring, overly descriptive names. Below are three methods to help your qualitative naming research give you the best information. Remember there are no results – just information from which you must make decisions.

1. Timing of the research
Try to time your naming qualitative research when you have developed a few name options that you think might work, don’t wait until you’ve completed the creative process. What you find out in research can and should be fodder for zeroing in on the right tone, style and concept for the final name. Be sure that these options are distinct enough for a customer to see the differences, and that you include a broad selection of names to represent the breadth of the exploration. If you’re exploring all the way up through coined names, but you only research real English words and camelbacks, then your research might not be helpful to your ultimate decision-making.

2. Question flow
When running qualitative, you’ll want to flow the naming discussion as follows:

Before any category or brand conversation is had, show the names unbranded and unaided. If it’s a portfolio name, don’t show the Primary Brand. And don’t give any hints – verbal or visual – about what the name is for. Just ask for associations: words, images, phrases, feelings that come up when respondents see the name. 

Next, dive into a conversation about the category, differentiated capabilities and benefits offered by the product or brand. Revisit those same names and ask how respondent opinions have shifted (if at all) and why. This is checking the names in an aided way – now respondents generally know what the name is for, and they will be gauging for fit. However, they have already committed to the feeling feedback from the unaided portion of the conversation, so you will get fit information within the context of their System 1 reactions – perfect!

Lastly, you can reveal the Primary Brand name, if there is one, and gauge for changes in associations or opinions. It can be helpful to conduct this line of questioning about competitors, too. For example, what if this name were from Amazon? From Google? From Microsoft? How do those Primary Brands change the respondent’s opinion about the name and why?

3. How you share and show the names
As mentioned above, first share the name as a simple, black and white piece of text. No extra information, no design. Next, share it as a logo or wordmark, and even (ideally) in a situation the customer is likely to encounter the name – for example, on a picture of the packaging, or a picture of street signage. Separating out the name from the name in situation can help you understand the difference between whether the name is conjuring up great associations, or if a name has issues or needs help from a tagline (or just a more descriptive name) or other context when in certain situations.

Quantitative research
Sometimes, you need a projectable data set to feel confident – or make your decision maker or product maker or budget holder confident – with a naming choice. Quantitative research is true testing, and you can come out with a winner. However, it still must be executed well to give you real-world results, and not just respondents picking the most descriptive (and therefore easiest to understand and relate to WHAT you’re selling) name available.

1. Design of the quantitative
First, make your survey monadic, meaning one set of stimulus. In this case, one name for each cell of respondents. Cell 1 will only see and react to name 1. Cell 2 will only see and react to name 2. And so forth. You’ll need a projectable, stable sample size for each cell, so this can get pricey depending on how finely you need to cut the data or how many names you want to test. Why design it monadically? Because it best duplicates the real world, and so the results are the most reliable. And, yes, they can radically differ from results of comparative questioning.

2. Flow of the quant
The same flow from qualitative can be used here. First, show the name unaided and unbranded, with no context given, and ask respondents to share or click on or rank the attributes and tonal words they most associate with the name. But remember! The more suggestive your naming strategy, the less you should be testing for attribute fit, and the more you just want to gauge general feeling and tonal associations. Next, show a definition or explanation of what you’re naming and then ask respondents to rank the name for fit. Lastly, show the name branded (if applicable).

3. When to bring in comparison
You can bring in comparison, but it’s best to do after the monadic testing is completed –for example, the first ¾ of the survey will be focused on a single name, and the last ¼ may show all respondents a list of all the possible names and ask them to rank them. You can also do sequential monadic – which simply means that you switch up the order of your names for different cells, to flush out order bias from your data.

Naming breakthroughs happen when we use research to find out what a name evokes and how it works (or doesn’t work), not whether the name summarizes what the product or company is or does.

Make sure your research is designed to give you real-world inputs, not staged insights that would never be set up or even necessary in the real-world use of your name, and you’re more likely to get breakthrough, emotionally evocative names that work hard for you and for your customers.

If you're in need of naming or naming research contact Northbound for a consultation.

Learn from us

Sign up to receive monthly insights right to your inbox
Get on this list

Explore more resources

We solve business challenges with brand.