AI website builder for small businesses

Problem

The builder market was crowded and hard to stand out in, and the wizard itself was the weak link: with too many questions people never reached the end, and the drop-offs showed many of them were coming from a phone, where the wizard did not work at all.

Hypothesis

If a wizard assembles three personalised site options in minutes, from the smallest possible number of questions, the number of customers will double, because people don't want to create a website, they want to edit one already made for them.

Solution

A short quiz with minimum questions and prewritten multiple-choice answers, then the wizard generates three personalized site options in minutes. Pick one; everything stays editable. Fully functional on mobile.

Problem

Every builder on the market put configuration on the left and the preview on the right. The flow was unclear, users got lost between the two halves, and templates could only be changed as far as the panel allowed.

Hypothesis

If editing moves inside the template itself, with hover settings on every element, users will understand the builder without a manual, and templates can become fully customisable.

Solution

Hover settings on every element replaced the rigid split layout: users change what they see, right where they see it, and templates became fully customizable.

Problem

To sell online a small business owner has to enable the store and set up product and collection pages, cart, checkout, wishlist, payment and shipping methods, discounts, coupons and promotions, and do it without a developer and without reading documentation.

Hypothesis

If the whole store (pages, cart, checkout, payment, shipping, discounts) is configured inside the same builder, a small business can start selling without a developer.

Solution

A full store inside the builder: product pages with flexible block sequences, collection pages with filter and sidebar settings, cart and wishlist, payment configuration.

Problem

Research on checkout flows showed the same thing everywhere: the longer the procedure, the more buyers fall out of it mid-way.

Hypothesis

If payment and shipping details pull from the site itself, the buyer only has to confirm, and more of them will reach the end.

Solution

Research showed long checkouts kill conversion, so payment and shipping details pull from the site automatically, and the buyer confirms instead of typing.

Problem

Shipping had to work for any kind of business, with fees depending on the country and the region of the delivery address, and coupons had to be enabled in a few clicks, not configured.

Hypothesis

If fees are set by country and region and coupons switch on in a few clicks, configuring a store stops being a developer's job.

Solution

Shipping fees by country and region for any business type; coupons enabled in a few clicks for any product or collection; promo bundles with suggested similar products.

Problem

Once the site was live, the owner faced a dozen sections with no obvious place to start.

Hypothesis

If the site's status and every tool sit in one control centre, a small business runs its site without an admin.

Solution

The site owner's control center: site status, quick access to every tool, and the settings that keep a small business running without an admin.

Problem

Promotion lived in separate services: mailing in one, analytics in another, links in a third.

Hypothesis

If lists, campaigns and their analytics live inside the builder, the owner does not have to assemble a stack of three subscriptions.

Solution

Mailing lists and email campaigns with built-in performance analytics, plus integrations and social links: promotion without leaving the builder.

Problem

Images and documents were uploaded one by one straight into a page, so the same file existed in five copies.

Hypothesis

If there is one library for the whole site, content gets reused instead of uploaded again.

Solution

Upload, store and organize content in an integrated file manager: images, documents and media ready to drop into any page.

Problem

To run a blog a business went to a separate platform, and the site and the blog stopped being one product.

Hypothesis

If publishing happens in the same editor as the site, the blog stays part of the site instead of living next door.

Solution

Blogs, articles and a dynamic content hub: publishing built into the same editor as the site itself.

Problem

Service businesses took bookings over DMs and the phone, and the site played no part in it.

Hypothesis

If appointments are taken on the site itself, the business stops losing people between tabs.

Solution

Scheduling and booking for service businesses: appointments taken directly through the site.

Problem

Final consumers rarely leave reviews on their own, and businesses lose potential customers over the missing feedback.

Hypothesis

If the system asks for feedback right after a completed purchase, the business gets the reviews customers never leave on their own.

Solution

Consumers rarely leave reviews on their own, so the system gathers feedback automatically after completed transactions and showcases it on the site.

Problem

Every business segment had to find something fitting on day one, so the product needed an enormous template base, but in 2021 there was no AI generation to produce it. Every palette, font pairing and style variation was configured by hand and linked to UI elements.

Hypothesis

If there are 2000 templates built on 300 vetted font and palette pairs, every business segment finds a fitting starting point on day one.

Solution

Category analysis of business websites defined 2000 template themes; a rapid-creation tool and 50 freelance creators produced them on 300 hand-configured font and palette pairs. I built 50+ templates myself.