Neuro had 2 hours and a quiz Word doc. Octane AI built the rest

“We see a lot of traffic, but we need to convert a little better.”

–Johnny Grimal, Ecommerce Manager, Neuro

Urgency came with the end of year season. Neuro had several launches stacked before year-end, including its limited-edition Energy & Focus™ Sour Mints, and anything that took weeks to assemble wouldn't make the calendar.

Why a Quiz, and Why Now

Johnny had already seen the format work, and he knew the shape it needed to take.

“We saw that quizzes that offer 20% off or a little something afterward work really well. And especially if it's a small quiz, with like 5 questions, it's not that complicated.”

–Johnny Grimal, Ecommerce Manager, Neuro

The Sours campaign also had a sharper tone than the rest of the site, and whatever went live had to match it.

Phoenix Built the Quiz From One Document

Neuro's quiz was built in Phoenix, Octane AI's new quiz builder, during its closed beta, making it one of the first live quizzes on the new builder.

A Phoenix build opens with a choice between a blank canvas and Ask AI to Create. The second route takes what a brand already has and turns it into questions, pages, and rules that connect every answer to a product.

Johnny wrote the whole quiz in a document first, then handed it over.

“I created the quiz on a Word doc, basically on text, and then I attached that document into Phoenix, and then Phoenix built the whole thing, all the pages.”

–Johnny Grimal, Ecommerce Manager, Neuro

What caught him off guard was not the copy. It was the logic.

“It understood everything. Not only the text, but the logic between, like, okay, if you answer all the A's, you're going to get this answer.”

–Johnny Grimal, Ecommerce Manager, Neuro

That is the part of a quiz build that used to take the longest: creating every page, then wiring each answer to each outcome. With Phoenix, the whole build took about two hours across two work sessions: one to set it up, one to fine-tune and publish.

Why Neuro's Quiz Works: A Campaign-Led Strategy

1. The Questions Came From the Campaign, Not the AI

Johnny came in with the content already written, on purpose.

“I knew the campaign strategy, I knew the campaign tone of voice, I knew all of that, that maybe was some information that was still not there for Phoenix.”

–Johnny Grimal, Ecommerce Manager, Neuro

It's what Octane AI recommends to every brand starting a build: bring a first pass at the questions, then let Phoenix handle the pages, the logic, and the catalog mapping.

2. The Design Came From His Designer, Pasted In as Screenshots

The quiz is visually loud, and none of it was built by hand either. Phoenix takes any image as a styling instruction, a mockup, or a screenshot of your own storefront, and applies it across every page at once.

“I built a little mock-up, took screenshots, spoke with my designer, and she did mock-ups of how it should look. Then I pasted all those mock-ups in to say, hey, I need it to look like this. And it got it.”

–Johnny Grimal, Ecommerce Manager, Neuro

Asked how much of the build ran through the AI, Johnny was blunter than that: “I did everything on the AI in this quiz.”

3. Better Prompts Made the Difference

Day one went quickly, since Phoenix picked up his colors and images right away. Smaller adjustments were harder until he changed how he asked.

“I understood that I had to give better prompts. And with Leonardo [our account manager] as well, he helped me a lot with that. And then it ended up looking really good, so really happy about it.”

–Johnny Grimal, Ecommerce Manager, Neuro

4. Going Live Stopped Being a Project of Its Own

Before publishing, Phoenix maps the quiz as a diagram and walks the answer combinations a shopper could produce, flagging any path that ends without a recommendation and pre-writing the fix. On a dated campaign, a dead end found after launch means lost revenue within the first month.

Publishing itself happens inside the editor. “The process of publishing is so much easier as well,” Johnny says.

5. The Results Page Sold the Bundle, Not Just the Product

With Neuro's quiz, Phoenix handled something that would have required a developer in the old editor: variant selectors built directly into the results page. Rather than being sent to a product page to choose their options, shoppers could select their exact bundle, size, or format right there, without ever leaving the quiz experience.

In the previous editor, that kind of flexibility meant a custom dev build or an extra step that broke the flow. With Phoenix, it was part of the build from the start.

From One Launch Quiz to the Next

Neuro's story shows what changes when building a quiz stops being the bottleneck. Phoenix reads what a brand already has and builds the questions, the logic, the results page, and the design from it. Marketers keep control of the strategy, and Phoenix handles the build.

Phoenix is live, which means the next quiz you launch can start the way Neuro's did: with a document and a prompt. Book a demo, and we'll build a quiz with you.