Published

How Zola Validated a Wedding Product for $40K

A Zola product lead and one engineer bottled a psychographic wedding-task splitter into a custom GPT — shipping in one month for $40K instead of six months and $125K.

~67%

Reduced vs. traditional dev cost

6–12 months

Implementation Time

$25K – $100K

Project Cost

the challenge

Zola’s annual First Look Report revealed that the single biggest surprise for newly married couples is the sheer volume of wedding decisions — and that in heterosexual partnerships, the overwhelming majority of those decisions fall on one partner. Zola wanted to address this imbalance but couldn’t easily spin up a new platform feature. They needed a low-cost, testable solution that could validate demand before committing engineering resources.

what they built

Jenny and one Zola team member built a custom GPT called 'Split the Decisions' in approximately one month for roughly $40,000. The tool asks both partners a series of questions — including psychographic prompts about strengths, concerns, and wedding vision — then equitably divides the full list of common wedding tasks between them based on individual profiles rather than gender roles. It generates a downloadable CSV assigning each task to a partner with links to relevant Zola articles and products. The GPT launched publicly, earned press coverage, and served as a low-risk prototype to gauge product-market fit before any deeper platform investment.

Jenny Nicholson and a single Zola team member approached the problem as a rapid prototype question: could they validate demand for a decision-splitting tool before committing engineering resources to a full platform feature? The answer was a custom ChatGPT GPT. The tool was designed around psychographic inputs: both partners answer questions about their strengths, concerns, and wedding vision. The GPT uses those profiles to divide the full list of common wedding tasks equitably between partners — assigning based on individual fit, not gender norms. Each task in the output CSV includes a link to the relevant Zola article or product, integrating discovery and commerce. Built using ChatGPT Team/Enterprise, the entire project took approximately one month and cost roughly $40,000 — compared to a traditional software build estimated at six months and $125,000. The GPT launched publicly and earned press coverage. The outcome served a dual purpose: solving a real user pain point and providing a low-risk signal about product-market fit before any deeper platform investment.

best fit for

Best for creative agencies, brand marketers, and digital product teams at consumer platforms who want to test AI-powered interactive experiences without full engineering investment — particularly teams sitting on consumer survey data and looking to turn it into an engaging, functional user tool quickly.

Ai ROLE

Not shared

impact

Cost Reduced by ~Two-Thirds

A project that previously would have required 6 months and $125,000 was completed in approximately one month for around $40,000 — reducing cost by roughly two-thirds and compressing the timeline by over 80%.

Earned Media and Proof-of-Concept Validation

The custom GPT launched publicly and generated meaningful press coverage, demonstrating that AI-powered tools can serve simultaneously as a marketing asset and a market research instrument — validating demand before any platform engineering commitment.

New Model for Idea-to-Launched Product

The project established a repeatable model for rapidly moving from consumer insight to a live, testable product at minimal risk — something Jenny described as previously impossible at this speed and cost.

implementation complexity

Not shared

Jenny Nicholson

Making Magic with Machines | ⚔️ Queen of Swords ⚔️ | Movement Strategy | Brilliant Failures

Queen of Swords

An experienced advertising creative helping agencies and brands move beyond a “faster, cheaper, and with fewer people” mindset.

Get an intro

Talk to this team

industry

Consumer Goods & CPG

Technology & Software

business organization

Product & Engineering

Marketing

AI TYpe

Conversational AI (Chatbot / Agent)

Decision Support & Scoring

value type

Revenue Growth

Customer Experience

Time Savings

other(firm) Case studies

frequently asked questions

How did Zola validate a new product for $40K using a custom GPT?

Zola, the wedding-planning company, built a custom GPT called 'Split the Decisions' with a product lead and one engineer instead of committing to a full platform build. The tool asks both partners psychographic questions, then divides wedding tasks equitably based on individual profiles rather than gender roles, outputting a CSV with links to relevant articles and products. It shipped in about one month for roughly $40,000 instead of an estimated six months and $125,000.

What AI tools did Zola use?

The team built a custom GPT using ChatGPT Team/Enterprise. The conversational tool collects psychographic inputs from both partners and uses those profiles to split tasks, with each output item linked to a relevant article or product.

What results did Zola achieve?

Three outcomes: cost reduced by roughly two-thirds and the timeline compressed by over 80%, completing in about one month for around $40,000 versus an estimated six months and $125,000; a public launch that earned press coverage and validated demand before any platform investment; and a new repeatable model for moving from consumer insight to a live, testable product.

How long did the project take?

Time to results was under a few weeks of active build, with the full project taking approximately one month — compared with an estimated six months for a traditional software build.

Who is this custom-GPT approach best for?

Creative agencies, brand marketers, and digital product teams at consumer platforms that want to test AI-powered interactive experiences without full engineering investment — particularly teams sitting on consumer survey data they want to turn into a functional user tool quickly.

Have a similar challenge?

Ask whether this would work for you, or describe what you're trying to solve.

TELL US WHAT YOU'RE EXPLORING