Validation SOPs › Solution Validation
SOP 39 of 100 · Solution Validation
Solution Name Testing SOP
- Difficulty
- Beginner
- Time
- 2-3 days of naming, 1 week of testing
- Cost
- $100-300
- Best for
- Consumer products and apps where people hear, search and recommend the name
- Tools
- PickFu, Lyssna, Google Forms, Namechk, USPTO trademark search, EUIPO TMview, WIPO Global Brand Database
- Measure
- Spelling accuracy, Unaided recall, Association fit, Description match, Legal and domain clearance
A solution name testing SOP helps you pick a product name that customers can spell, remember and recommend, and it takes about a week and a half. You generate a long list, cut it with fast legal and domain checks, then test the shortlist with real people from your target group. A name gets judged out loud, in search bars and in group chats, so the tests copy those moments. Your own favourite name goes through the same tests as the rest.
What is solution name testing and when should you use it?
Solution name testing checks a shortlist of names against four practical questions: can people spell it after hearing it, do they remember it later, what does it make them think of, and can you own it legally and online. It uses small, cheap tests with target customers instead of a team vote.
Use this SOP:
- Before you buy a domain, print anything or launch a landing page. Renaming later costs you search history, links and recognition.
- When your product spreads by word of mouth, where people hear the name before they see it.
- When you plan to sell in several countries and need to know the name works in each language.
- When the team is split between favourites.
This SOP names one product or solution. The bigger question of how your brand sits against competitors belongs to brand positioning validation, and the promise that goes next to the name belongs to solution value proposition testing.
What do you need before you start?
- A naming brief: what the product does, for whom, the feeling it should carry, target countries and languages, and words to avoid.
- A list of your product's Nice classes. The Nice Classification sorts trademarks into 45 classes, 34 for goods and 11 for services. Software sits in class 9, financial services in class 36.
- Access to trademark databases: the USPTO trademark search for the US and TMview for the EU and many national offices.
- 20 to 50 target customers for tests, from your network of interviewees or a screened panel on PickFu or Lyssna.
- $100 to $300 for panel responses and a couple of domains.
How do you run it, step by step?
1. Write the naming brief
Put the brief on one page. Include three words for the feeling you want ("calm, shared, smart") and three you must avoid. The brief is how you judge names later, so write it before anyone suggests one.
2. Generate 30 to 50 names across types
Spread names across the trademark spectrum of distinctiveness from US trademark law: descriptive names ("Budget for Two") are the hardest to protect, suggestive names hint at the benefit, arbitrary names borrow an unrelated word, and invented names are the easiest to own and the hardest to remember. Aim for a mix.
3. Screen to five to eight names
Cut fast. For each name, check the .com or your country domain, social handles with Namechk, a trademark search in your Nice classes, the app stores and the first page of Google. Ask a native speaker of each target language whether the name means something unfortunate.
4. Run a spelling test
Read each name aloud to 20 people (on a call or as an audio clip in a survey) and ask them to type it. Count exact matches. A name people can't spell is a name they can't search for.
5. Run an association and fit test
Show each name alone and ask for the first three words that come to mind. Then show the product description and ask which name fits it best. Look for negative associations that appear more than once.
6. Run a recall test
Show the shortlist inside a short survey, then ask other questions for ten minutes. At the end, ask: "Which product names do you remember?" Unaided recall tells you which names stick without help.
7. Run a preference poll last
Only now ask which name people like best, and why. Preference comes last because liking a name says little about whether it works. Use it to break ties.
8. Clear the winner properly
Before you spend money on the name, have a trademark attorney run a full clearance search in your classes and countries. Your own searches catch obvious conflicts; an attorney catches similar marks and pending applications.
Template you can copy
NAMING BRIEF
Product does: __________ For: __________
Feeling (3 words): ___ ___ ___ Avoid (3 words): ___ ___ ___
Markets and languages: __________ Nice classes: __ __
SCREENING SHEET (one row per name)
Name | Type (descriptive / suggestive / arbitrary / invented) |
Domain free? | Handles free? | Trademark conflict in my classes? |
App store conflict? | Language check OK? | Keep Y/N
TEST SHEET (5-8 finalists)
Name | Spelled correctly after hearing (of 20) | First 3 words (top themes) |
Negative associations (count) | Picked as best fit for description (%) |
Unaided recall after 10 min (%) | Preference (%) | Notes
SURVEY QUESTIONS
1. [audio] Please type the name you just heard.
2. What are the first three words that come to mind for "[name]"?
3. Read this description. Which name fits it best?
4. (after other questions) Which product names do you remember from earlier?
5. Which name do you like best, and why?
Worked example: a budgeting app for couples
This example is illustrative, and all names are invented for it. A founder builds a shared budgeting app for couples. After screening 42 names, five remain. She tests them with 30 people in couples who share some expenses.
| Name | Spelling (of 20) | Unaided recall | Fit with description | Most repeated association |
|---|---|---|---|---|
| Kittywise | 16 (80%) | 63% | 47% | "shared pot", "money jar" |
| Pairpurse | 17 (85%) | 57% | 20% | 9 of 30 said "handbag" |
| Duetto | 11 (55%) | 47% | 17% | "music", "Italian" |
| Coupla | 8 (40%) | 30% | 10% | "couple of what?" |
| Halvsy | 7 (35%) | 37% | 6% | "half", hard to say |
Pairpurse was the easiest to spell, but nearly a third of people read it as a handbag. Kittywise led on recall and fit, and "kitty" already means a shared pot of money in everyday English. Her own search in classes 9 and 36 showed no conflicts, so she sent it to an attorney for full clearance before buying the domain.
What should you measure, and when do you stop?
| Metric | Pass signal | Warning signal |
|---|---|---|
| Spelling accuracy | 70%+ type it correctly after hearing it | Under 50% |
| Unaided recall | 50%+ remember it after 10 minutes | Under 30% |
| Association fit | Top associations match your brief | The same negative association from 2+ people |
| Description match | Picked as best fit by the largest group | Rarely matched to your product |
| Legal and domain clearance | No conflicts in your classes and markets, usable domain | Similar mark in your class, or only awkward domains |
Stop when one name passes every row. If none does, take the best two, generate variants around them and test again with a new group. Never skip the legal row for a name that wins everything else.
Which mistakes should you avoid?
- Voting inside the team. You know the product, so every name makes sense to you. Test with strangers.
- Falling for a descriptive name. It explains itself today and is hard to protect tomorrow.
- Buying the domain first. Domains are cheap, and attachment to them is expensive. Test before you buy.
- Skipping the language check. A name that sounds fine in English can be a joke in your second market. Geographic market validation covers the wider country checks.
- Treating your own search as legal clearance. It catches identical marks. An attorney catches the similar ones that cause trouble.
How can AI speed this up?
AI assistants generate 50 names in seconds; feed them your brief and ask for each type on the spectrum of distinctiveness. The free F/MS business name generator gives you a starting list too. AI can also check a shortlist for meanings in other languages, but confirm with a native speaker. The human check: real people do the spelling and recall tests, and a trademark professional clears the winner. Once you pick, the digital marketing module helps you make the name findable in search.
Where does this SOP fit in your validation plan?
Name testing sits late in Solution Validation, once you know what the solution is and what it promises.
- Before: settle the promise with solution value proposition testing.
- Next: put the name on a page with the landing page test SOP, and use the online survey validation SOP to recruit and screen respondents.
- Later: check brand fit against competitors with brand positioning validation and tune copy with marketing message testing.
See every Solution Validation SOP.
FAQ
How many people do I need to test a product name?
20 for the spelling test and 30 to 50 for association, fit and recall. That is enough to see clear failures, which is what you are looking for.
What makes a good startup product name?
People can spell it after hearing it once, remember it later, link it to the right feeling and you can own it as a trademark and a domain.
Should I check trademarks myself?
Yes, as a first screen in the USPTO database, TMview or the WIPO Global Brand Database. Before you commit, pay a trademark attorney for a full clearance search.
Do I need a .com domain?
Not always. A country domain or a clear variant works if customers in your market find it easily. Test the spelling of the full domain the same way you test the name.