> For the complete documentation index, see [llms.txt](https://docs.zigpoll.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.zigpoll.com/guides/how-to-test-product-names-and-creative.md). # How to Test Product Names, Packaging, and Ad Creative Naming and creative decisions are usually settled by whoever argues hardest in the meeting. They don't have to be. A twenty-minute survey with a hundred respondents will beat a week of internal debate, and it costs less than a single day of the ad spend you're about to commit. The trick is asking in a way that produces a decision rather than a set of opinions. *** ## Testing names ### Build it 1. Add a **Single Choice** question: **"Which name best fits a product that \[does the thing]?"** — list your candidates. 2. Add a **Long Answer** follow-up: **"Why that one?"** — this is where you learn a name is testing well for a reason you didn't intend. 3. Add a comprehension check with a **Short Answer** question: **"What would you expect this product to do?"** Show one name and ask what it means to them. Watch for order bias: whatever sits first in a list wins more often than it deserves. If the result is close, re-run it with the order reversed, or use the head-to-head method below, which sidesteps the problem entirely. That last question is the one people skip, and it's the most important. A name that everyone likes but nobody understands is worse than a boring name that's instantly clear. **Preference is secondary to comprehension.** *** ## Testing packaging, product design, and ad creative ### Build it 1. Add an **Image Choice** question with your candidate designs. Cap it at six — beyond that respondents skim rather than evaluate. 2. Set how many images can be selected and how many columns to display. 3. Add a **Long Answer** follow-up: **"What made you choose that one?"** 4. For a sharper read, add a **Ranked Choice** question instead of a single pick. Forcing a full ranking surfaces the *second* choice, which matters when your favorite and the winner are close. {% hint style="info" %} Ask "which would you **buy**," not "which do you **like**." Liking a design and paying for the product it wraps are different judgments, and only one of them predicts revenue. {% endhint %} *** ## Testing ad recall If you're running brand campaigns — TV, podcast, out-of-home, influencer — the hard question is whether anyone remembers the creative at all. 1. Add an **Image Choice** question showing several ad creatives, including at least one that is **not yours** (a competitor's, or a plausible fake). 2. Ask: **"Which of these ads have you seen in the last month?"** The decoy is what makes this work. Some respondents will claim to have seen everything, including the ad that doesn't exist. The gap between the recall rate for your creative and the recall rate for the decoy is your *real* recall — everything above the decoy line is signal, everything at or below it is noise. *** ## Running a true head-to-head For the cleanest read, don't show respondents the options side by side — show each person only one, and compare across groups. Side-by-side comparison measures which design wins an explicit beauty contest. Showing one at a time measures how each performs on its own, which is how customers will actually encounter it. Create a variant of the survey for each option and run them simultaneously as an [A/B test](/tutorials/a-b-testing-surveys.md). Then compare purchase intent (a **Range** question, 1–10) across the variants rather than comparing which picture people pointed at. *** ## Where to run it * **On-site to existing traffic** — Fast, and the respondents already know your brand. Best for packaging and design. * **Email to customers** — The right audience for a name or packaging change to an existing product; they have context nobody else does. * **A** [**public link**](/public-link.md) — Share it in a community or promote it to a cold audience when you need people who *don't* know your brand. For a name test, brand-blind respondents are the more honest sample. * [**Synthetic Research**](/synthetic-research.md) — Get a directional read from a target demographic in minutes when you need to narrow ten candidates down to three before spending real traffic on them. *** ## Reading the results * **A clear winner (over \~50% of a five-way choice)** — Go. The debate is settled. * **A close split** — The options are equivalent to customers, however different they look to you. Pick the one that's cheapest to produce and move on. A tie is a real result and it saves you a week. * **Preference and comprehension disagree** — Comprehension wins. Every time. Read the open-ends before you finalize. Z-GPT Insights clusters them into themes automatically, and this is exactly where you catch the name that half your audience finds charming and the other half finds unfortunate. {% hint style="warning" %} Don't test creative on your most loyal customers alone. They like everything you make, which is what makes them loyal — and it makes them a poor proxy for the strangers your ads need to reach. {% endhint %} *** ## Related {% content-ref url="/pages/uZzmwXRquTBAxLPBAgeA" %} [How to Test Your Messaging and Positioning](/guides/how-to-test-your-messaging-and-positioning.md) {% endcontent-ref %} {% content-ref url="/pages/enRwROsoKeKFUh7XMOOp" %} [How to Validate a New Product Idea](/guides/how-to-validate-a-new-product-idea.md) {% endcontent-ref %} {% content-ref url="/pages/qaUwTo0KoYMZQFwYZaVC" %} [A/B Testing Surveys](/tutorials/a-b-testing-surveys.md) {% endcontent-ref %} --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://docs.zigpoll.com/guides/how-to-test-product-names-and-creative.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.