Zipify Labs is a fast-moving team inside Zipify, a Shopify app maker. It was built together with Master of Code Global.
Most store quizzes end with an email capture form. Zipify core expertise helps thousands of merchants to sell more (or earn more) through different upsell mechanics. The team had a idea: create a quiz builder that ends in a sale.
So we built PerfectMatch, a Shopify app where AI creates a personalized quiz in 15–30 seconds. Merchants publish it in a few clicks. Over 5 months, quizzes generated over 900 orders and $75K in sales across 12 stores.
Zipify builds Shopify apps, including OneClickUpsell, Zipify Pages, and ShopRank AI. Master of Code Global has been its development partner for over a decade.
During Zipify’s AI transformation, we set up Zipify Labs, a small & fast-moving team that tests new product ideas. For PerfectMatch, the brief was a quiz that sells products and that a merchant can launch alone.
Most Shopify quizzes follow one script. A long list of questions. Then an email request. The result arrives in the inbox. Many shoppers leave before the end.
Launching a quiz is slow, too. Before shoppers see it, a merchant has to build it. A copywriter writes the questions. A colleague chooses the products. Stakeholders approve the brand voice. A developer places the widget. In the best case, this takes one to two days. Often it takes a week.
“Our solution had to let a merchant launch a quiz that drives sales almost in one click.”
— Oleksandr Fesenko, Product Lead, Zipify Labs
Why it’s a challenge
Few quiz products aim at the sale, so there was little to get inspiration from. Shopify themes add friction as well. Placing a quiz on a page usually meant watching tutorials, pasting IDs into code snippets, or hiring help.
We built PerfectMatch around one rule. If a step doesn’t help the shopper or the merchant, cut it.
Take a barbecue sauce store. A shopper picks “backyard barbecue” and answers two more questions. The quiz then shows two sauces with a one-line reason for each. There is no email gate.
“Anyone can figure it out, because it’s a guided experience. You don’t touch any code. You just press ‘add to this page.’”
— Oleksandr Fesenko, Product Lead, Zipify Labs
We first tested PerfectMatch on two large brands in the Zipify portfolio. Then we ran a private whitelist. We set up each early store by hand and fixed what tripped people up before the wider market saw it. After that, the app went public on the Shopify App Store.
Two obstacles shaped the build.
Speed. The first versions took about a minute to write a quiz. That is far too long for a merchant waiting on a screen. We benchmarked a wide range of language models for cost, quality, and speed. We also set the default quiz at two to three questions, with five as the maximum. Gemini 2.5 Flash cut the wait sharply while maintaining high quality. Faster models exist, but their quality fell short, so we kept this setup.
Simplicity. We redesigned the merchant flow several times. Each round used feedback from early users about where they struggled to go live.
“The main challenge was making it simple, because we set ourselves the goal that it must be the easiest.”
— Oleksandr Fesenko, Product Lead, Zipify Labs
Zipify Labs now owns the app and runs it as its own product.
Master of Code Global has built Zipify’s products for over a decade. We know how to ship inside Shopify’s rules. We turn new ideas into products merchants can launch without help.
We pick AI models by benchmark, not by habit. For PerfectMatch, that cut quiz generation from about a minute to 15–30 seconds without a drop in quality.
We test on live traffic and open in stages. We start with a private group, fix what appears, and then go public. The same approach works for any business that sells a catalog of ten or more products, on Shopify or elsewhere.
Want the same results as Zipify? Contact Master of Code Global.