Hello Jasmine Replaces Its Wix Site — and Doubles Its Page-One Keywords

• Client:Hello Jasmine
• Industry:Taiwanese tea house — boba, popcorn chicken, bento
• Locations:3 (Chinatown, Lincoln Park, Lincoln Square — Chicago)
• Point of Sale:Square
• INFI products:SITES, ORDER (self-ordering kiosk), GROW

Overview

Hello Jasmine brews loose-leaf tea and slow-cooks its pearls in house across three Chicago neighborhoods. The brand is well known locally: a little over two thousand Google reviews across its three storefronts, rated 4.5 to 4.8 stars.

Its website did not match. The site ran on Wix, and it was carrying about half a megabyte of HTML on the home page alone. In June 2026 the same domain moved to INFI SITES. Because the domain never changed, everything below is a genuine before-and-after on the same address, the same brand and the same audience.

The challenge

The old site was not ugly. It was heavy, and it told Google almost nothing.

The home page shipped roughly 507 KB of HTML, about sixty script tags with around ten external script files, and twenty-two external stylesheets. Structured data amounted to two thin blocks on the home page — a bare business listing and a website record — with no opening hours, no coordinates, and no menu markup anywhere on the site. Interior pages such as the location and story pages carried no structured data at all.

Google's own field measurement of real Chrome users rated the site SLOW on mobile in the last full window before the move.

Speed: measured on real visitors, not a lab

The most useful number here is not a Lighthouse score. It is Google's Chrome User Experience report — the speed real people on real phones actually experienced on this domain, before and after.

Mobile field data (75th percentile)	Last window on Wix	First full window on INFI	Change
Largest contentful paint	4,051 ms	1,754 ms	−57%
First contentful paint	3,646 ms	1,028 ms	−72%
Cumulative layout shift	0.00	0.00	held
Google's verdict	SLOW	FAST	—

Lab testing agrees. The home page scores 98 on mobile and 99 on desktop in Lighthouse, with a perfect 100 for SEO and best practices; the menu page scores 91 on mobile and 99 on desktop. Those are not one lucky run: INFI's own page-speed cron has recorded 109 independent runs on this domain over three and a half months, averaging 98 to 99 on mobile and 100 on desktop.

Findability: the search surface roughly tripled

Comparing two matched 82-day windows in Google Search Console — the last stretch on Wix against the equivalent stretch on INFI, same property, same domain:

Google Search Console	Before	After	Change
Impressions	35,167	55,942	+59.1%
Distinct queries the site appears for	1,175	3,101	+163.9%
Queries ranking on page one	668	1,303	+95.1%
Queries ranking in the top three	230	450	+95.7%
Clicks from non-brand searches	92	215	+133.7%
Impressions on non-brand searches	9,014	21,935	+143.3%
Home page clicks	1,665	2,542	+52.7%

One honest note, because it explains the shape of the result. Hello Jasmine's brand searches were already won — people searching “hello jasmine” were landing on the site at position one or two before the rebuild, and there was nowhere for those to climb. So the total click count did not jump. What changed is everything the brand was not being found for: nearly three times as many distinct queries, twice as many of them on page one, and non-brand clicks up 134%.

Part of that new surface simply did not exist before. The site's journal — ten posts published between June and September, each written around a real local search opportunity — went from nothing to 10,580 impressions. The best-performing post now sits at position 4.3 for its topic.

Keyword movement

Two different things happened, and they are worth separating.

Location searches improved in both rank and traffic. Each store got its own page, with its own hours, coordinates and live menu:

Query	Position before	Position after	Clicks
hello jasmine chinatown	2.57	1.55	92 → 292
hello jasmine lincoln square	3.76	2.38	65 → 176
hello jasmine lincoln park	4.18	3.20	48 → 60
hello jasmine	1.93	1.30	already at the top

Category searches moved from nowhere into contention. These are the searches from people who do not yet know the brand — the hardest and most valuable kind to win:

Query	Position before	Position after
jasmine bubble tea	25.8 (page three)	3.9
chicago boba	37.7 (page four)	8.8
boba chicago	42.4 (page five)	11.9
bubble tea chicago	49.4 (page five)	14.5
best boba in chinatown chicago	17.0	8.8

Being on page one of “chicago boba” is not the same as owning it, and we are not going to pretend otherwise: those terms are still converting in ones and twos. But a brand that was invisible on page four for its own category is now in the race, and two category terms are already producing clicks — “taiwanese food chicago” went from 2 clicks to 16, and “taiwanese restaurant chicago” from 3 to 13.

What is not in this case study

Two things, deliberately.

No paid advertising. Every number above is organic. INFI has not run a single Google Ads campaign for Hello Jasmine — the paid layer is still completely untouched, which is worth knowing when you read the category rankings above: that is what the free half of the channel produced on its own.

No online-ordering claim. Hello Jasmine's online ordering runs on a separate provider, and it ran there before the rebuild too. Nothing about their ordering changed when the website did, so nothing about the website's results can be credited to it. What INFI runs in store is the self-ordering kiosk, which predates this website by well over a year — that story is told separately in how Hello Jasmine improved their tech stack with INFI kiosks.

If your site is on a builder that is fast to launch and slow to load, the questions worth asking are the ones this project answered: what does Google's field data say about your real visitors, how many of your menu items does Google actually understand, and how many searches are you appearing for that are not your own name. Book a demo and we will pull those three answers for your own domain before you commit to anything.