⚙ MACHINE MODE. Plain-text semantic view for AI crawlers, screen readers, and structured data consumers. Click HUMAN in the footer to return to the visual experience.

Jem · CapitaLand

Singapore · Jan 2025 → present

SEO + Google SEM Retail · Mall Western Singapore GEO · AI Search

From 13 to 7.
A mall on page one.

How Kaliber built two parallel discovery engines for Jem. Organic content that 10x'd the keyword footprint in seven months, and Google SEM that delivered at 78% below projected cost-per-click.

10×

Organic keyword footprint

1,210 → 11,920 in pos 1–20 · 7 months

13×

Google AI Mode citations

~30 → 403 · Oct 2025 → Mar 2026

−78%

SEM CPC vs projected

$0.33 delivered · $1.50 projected

Scroll to read the case study

Where we started

A mall of this calibre.
Stuck on page two.

Average organic ranking position at engagement start. Page-2 territory for most queries Jem should have owned.

1,210

Keywords ranking in positions 1–20. For a 250-tenant mall anchoring Western Singapore, a thin footprint.

Active SEM spend. No baseline data on what paid search could deliver for footfall in the Jurong East catchment.

The mall economics

Footfall starts
with a search.

Mid-to-mid-plus mall. Anchored by IKEA Jurong, FairPrice Xtra, Shaw Theatres, Don Don Donki, H&M, Kopitiam, UNIQLO. Jem is the crown jewel of Western Singapore. But a brand presence doesn't translate into search presence on its own.

580k sqft retail across six levels

250 tenants

1 direct MRT interchange

West Singapore residents searching for dining, shopping, things to do. Jem wasn't the answer Google gave them
Industry projections suggested $1.50 CPC was reasonable for footfall-oriented mall SEM. That became the bar to beat
Two parallel engines needed: organic content for cumulative authority, paid search for immediate visibility

The brief

Western Singapore malls compete hard on footfall. Jem needed to show up in Google, and show up efficiently. The job: turn search into the most efficient acquisition channel in the marketing mix.

The challenge

Two engines.
One discovery system.

Most mall marketing teams choose one. SEO for the long game. SEM for the quarter. Jem needed both. Running in parallel, sharing the same intent map, compounding off each other.

SEO needed to expand beyond branded queries into local lifestyle clusters: "jurong east food", "malls in west singapore", "things to do jurong east"
SEM needed a structure that matched the way people actually shop the mall. By intent moment (dining, shopping, things to do, parking) not by generic keywords
Both needed to be ready for AI-driven discovery. Answer-shaped content that could be cited by Google AI Mode and AI Overviews, not just blue-link content

Branded search is table stakes. The growth lives in non-branded local intent, and that's where Jem had no presence.

AI Max unlocks inventory you never paid for. Lifestyle, beauty, entertainment, categories Jem hosts but had never been visible in.

Display CPM beats keyword auction for awareness objectives, by orders of magnitude when the structure is right.

The approach

Intent-mapped.
Structurally efficient.

Kaliber built Jem's search system around four intent moments, the way people actually decide which mall to visit. Each moment got matching organic content and a matching paid campaign. The structure made the algorithm's job easier and made the cost-per-result lower.

Map the intent moments. Not the keywords

Four moments shape how people search for a mall: Dining, Things To Do, Shopping, Parking. Each got a dedicated SEM campaign and a matching content cluster on the site. No campaign was structured around a generic keyword. Every campaign answered a specific question someone in the Jurong East catchment was already asking.

Use AI Max to discover inventory the algorithm finds for you

Things To Do (AI Max): entertainment, lifestyle, beauty, inventory Jem hosts but had never paid for. AI Max found it.
Dining (AI Max): dining intent in the Jurong East catchment. Answer-shaped pages that fed back into AI Mode citations
Shopping (AI Max): retail queries with intent to buy that day, not browse

PMax for impression-per-dollar awareness

"ParkforLess" PMax delivered impressions at $5.21 CPM. Vs Shopping at $113 CPM
The lesson: for awareness objectives, display inventory beats keyword auction by orders of magnitude when the audience signal is strong
Brand-level reach scaled into seasonal events without bidding up keyword inventory

Content that compounds into AI citations

Halal Food Guide, Cafes & Desserts, Things To Do guides, answer-shaped content that Google AI Mode could retrieve and cite
Each article fed both classic SEO ranking signals and the citation graph AI engines now use
Result: rank improvements visible inside 7 months, AI Mode citations 13x in 5 months. The same content powering both

How it ran

Two engines, one
intent map.

Organic content and paid search drew from the same intent moments. So the work in one channel made the work in the other channel cheaper. Both are still running.

Organic search · SEO

Content compounding

Average ranking moved from 13.3 to 7.63 in seven months, keyword footprint multiplied 10x.

The store directory became Jem's organic engine, 20,562 monthly clicks at peak (55.4% of all organic traffic). Lifestyle articles compounded behind it: Halal Food Guide +17% MoM, Cafes & Desserts +155% MoM.

+43%

Avg ranking improvement
13.3 → 7.63

37k

Monthly clicks (Mar '26)
peak 52k in Dec '25

Paid search · SEM

Algorithmic discovery

SEM delivered at $0.33 CPC. 78% below the $1.50 industry projection. AI Max unlocked inventory Jem had never paid for.

"Things To Do" delivered +259% impressions in a single month while CPC fell 37%. PMax delivered impressions at $5.21–$7.58 CPM, a different cost structure to keyword auction entirely.

$0.33

CPC delivered
vs $1.50 projected

135k

Monthly impressions
on $2k budget · Feb '26

Keyword footprint. Positions 1–20

Date

Keywords ranking

Change

Aug 2025 (engagement start)

1,210

baseline

Jan 2026

8,012

6.6×

Feb 2026

9,210

7.6×

Mar 2026

11,920

9.9×

The results

A mall of this calibre,
finally where it belongs.

Measured across organic search, Google AI Mode citations, AI Overviews, and paid search efficiency. All numbers from March 2026 monthly report.

AI Visibility. Ahrefs Brand Radar

Google AI Mode
403AI Mode citations · March 2026 record high
~30 (Oct '25) → 403 · 13× in 5 months

Google AI Overviews

AI Overview citations · March 2026

5 (Sep '24) → 49 · trending upward

ChatGPT

ChatGPT citations · March 2026

peaked at ~135 (Jan '26) · freshness-driven

January 2026 snapshot

200+

Multi-platform citation moment

AIO 50 · ChatGPT 112 · Perplexity 19 · Gemini 14 · Copilot 5

Organic search. March 2026

+885%

Keywords pos 1–20
1,210 → 11,920 · 7 months

37,096

Monthly organic clicks
+1.56% MoM · peak 52k in Dec '25

1.26M

Page impressions
+5.8% MoM · 85.6% of YTD target

SEM efficiency. Feb 2026

$0.33

CPC delivered
vs $1.50 projected. 78% more efficient

+259%

"Things To Do" impressions
single month · CPC fell 37%

$5.21

PMax CPM (lowest)
vs Shopping at $113 CPM

Where the traffic landed

20,562

/store-directory clicks
+7.2% · 55.4% of all organic

41.12%

CTR on "jem food"
5,946 monthly clicks · position 1

+155%

Cafes & Desserts article
MoM growth · article-led demand

Four reasons this worked

Transferable principles
for discovery-led retail.

Not Jem-specific. The structural reasons two parallel discovery engines outperformed projections. And why the same approach works for any destination retail brand competing on local search.

Map intent moments, not keywords

Dining, Things To Do, Shopping, Parking, that's how people decide which mall to visit. Build campaign structure around the moment, not the keyword. The match rate to actual buyer behaviour is higher and the algorithm rewards it.

AI Max finds inventory you'd never bid on

Jem hosts beauty, lifestyle, entertainment, categories never explicitly bid on. AI Max discovered the inventory and the audience. The result: +259% impressions in a single month while CPC dropped 37%.

Match channel cost structure to objective

Awareness ≠ keyword auction. PMax delivered impressions at $5.21 CPM versus Shopping at $113 CPM. Different objective, different inventory, different math. Don't pay click-rates for impression goals.

Build SEO content that AI engines can cite

Halal Food Guide, Cafes & Desserts. Answer-shaped articles that ranked classically *and* fed AI Mode citations. The same investment compounds across two systems. AI Mode citations went from ~30 to 403 in five months, without a single article written specifically for AI.

[Quote pending. Ally Tan, Marketing, Jem. Suggested angle: efficiency or discovery, "we're now showing up for searches we didn't know existed."]

Jem · CapitaLand

Services used

What we worked on.

Google Ads → GEO & AI Search → Intent-mapped SEM architecture AI Max campaigns PMax brand & seasonal Local-intent SEO content AI citation tracking

Discovery-led retail

Are you visible
where local buyers look first?

Most retail brands compete on broad keywords. The growth is in intent moments, and most malls aren't competing for them at all.

Get a free discovery audit

Google Premier Partner US$500M+ managed 90% from referral

From 13 to 7.A mall on page one.

A mall of this calibre.Stuck on page two.

Footfall startswith a search.

Two engines.One discovery system.

Intent-mapped.Structurally efficient.

Two engines, oneintent map.

A mall of this calibre,finally where it belongs.

Transferable principlesfor discovery-led retail.

What we worked on.

From 13 to 7.
A mall on page one.

A mall of this calibre.
Stuck on page two.

Footfall starts
with a search.

Two engines.
One discovery system.

Intent-mapped.
Structurally efficient.

Two engines, one
intent map.

A mall of this calibre,
finally where it belongs.

Transferable principles
for discovery-led retail.