More acquisitions from the same spend — with 2 campaigns and 13+ variations.
COMPANY OVERVIEW
DrinkPrime, at a glance.
India's leading water RO rental solution, offering IoT-enabled RO+UV water purifiers on subscription starting at Rs.349 per month, with lifetime free maintenance and zero upfront cost. Built for the tens of millions of Indian households who need clean drinking water without the burden of ownership, DrinkPrime adapts to varying water conditions across cities using smart purification technology.
FOUNDED
2016
INDUSTRY
Water Purification (Subscription/Rental)
EMPLOYEES
280+
GEOGRAPHY
Pan India
use case
Landing page personalization for Google Search Ads
SCALE OF OPERATIONS
2.5 Lakh
Households served
72.1 Cr
FY25 Revenue (+54% YoY)
₹100 Cr
Target for FY26
₹349/month
Plans starting rate
Two paid search campaigns on Google Ads covering incentive-based keywords, region-based keywords, and budget-friendly search terms. Price/incentive intent, geo-local intent, and rental category intent — each representing a different buyer at a different stage of the subscription decision.
We quickly adopted Fibr AI to align our ads with personalized landing pages, achieving great results with minimal effort. We're excited to scale this across our campaigns for increased growth and improved team productivity.
A
Abhishek Bhushan
Senior Manager, Performance Marketing · Livspace
In a Nutshell
Every search was high-intent. The page it landed on wasn't built for it.
The growth team at one of India's fastest-growing water purifier rental platforms ran tightly built paid search campaigns on Google Ads, reaching households actively searching for water purifiers on rent across multiple cities. But every click, regardless of what they searched or where they were, landed on the same generic static page. With Fibr, the team matched every keyword to its own landing experience across 130+ variations in two campaigns. The result: 21% more leads from the same spend, with bounce rates falling 33% and time on page doubling.
THE CHALLENGE
DrinkPrime's paid campaigns ran on Google Ads with high ad CTR, targeting buyers through a mix of incentive-based keywords, region-specific searches, and rental category intent. The ad strategy was tightly built. The landing page was not.
Regardless of what a visitor searched, which city they were in, or which price hook brought them in, every click landed on the same generic static page: the same hero, the same CTA, the same message built for no one in particular.
The cost of mismatch
Form fill rate was stuck at 17.5%. Google Ad Quality Score was suffering. CPA was high. Bounce rates were high across all traffic segments. The team was slightly aware of the problem but did not have a mechanism to act on it at scale. Once Google Ads and Google Analytics were connected to Fibr, the full picture came into focus.
THE SOLUTION
One URL. A matched experience for every search that pointed to it.
Fibr AI approached the problem across two campaigns, each building on the learnings of the last: identifying the intent gaps, structuring personalization around keyword signals and device context, and deploying matched hero banner experiences within the guardrails of DrinkPrime's brand guidelines.
STEP 1
The Analysis: the hero banner was carrying the full burden
DrinkPrime's core message lived in the hero banner itself. The entire page conversion depended on whether the banner connected with the visitor's specific intent. Fibr's approach was therefore image-level personalization: generating contextually matched hero banners within DrinkPrime's brand guidelines, not just swapping copy.
STEP 2
The Signals: keyword intent and device
Personalization was structured around two signals already present in every campaign: keyword intent and device type. Three primary cohorts emerged: price and incentive intent (visitors responding to Rs.349/month hooks and zero-cost guarantees), geo-local intent (city-specific searches across Bangalore, Delhi, Hyderabad, Noida), and rental category intent (generic RO and water purifier rental searches).
STEP 3
The Setup: two campaigns, 130+ variations, live within days
Campaign 1 went live within one week, covering scoping, audit, campaign creation with 35 variants, review, and approval. Campaign 2 went live within 4 days. The only engineering involvement required was a Fibr script addition by DrinkPrime's engineering team. All 130+ variations were pre-reviewed for accuracy, hallucinations, and compliance. Sign-off was decisions, not corrections.
The New User Experience
Keyword intent, device context, and DrinkPrime's brand guidelines followed the visitor from the first fold, matching the hero banner, the CTA copy, and the value message to the search term they came from. Nothing was lost between the ad and the page.
The results
The pilot ran a controlled experiment comparing personalized landing pages against the original generic page. Personalized variants won across every metric measured.
FORM FILL RATE
↑11.6%
Creating a wider top of funnel without increasing spend or changing the ad strategy.
CONVERSION RATE
↑24%
More session-to-lead conversions across all keyword cohorts without increasing paid budget.
ADJUSTED CPL
Stable
Despite industry-wide CPC rises driven by festival seasonality, the adjusted CPL remained stable.
ADJUSTED CPA
↓20%
Consistently generating more qualified leads lowered the adjusted CPA across the pilot period.
The cost of mismatch
Form fill rates plateaued at ~13.5%, below industry benchmarks. Adjusted CPA stayed high across every city. Over 1.5 million visitors were being served an experience built for no one in particular.
Ongoing Campaign Results
After the pilot, the campaign rolled out across 23+ cities. The conversion uplift was consistent — every
high-volume city, every month the campaign ran, with quality score moving from 3 to 8 across ad groups
over 8 months.
KEY INSIGHTS
INSIGHT 01
Page relevance converts and
makes every dollar go further
When the landing page matched the
search, Google's quality score moved
from 3 to 8. That improved ad auction
performance, brought more suitable
visitors through, and lowered adjusted
CPA by 20%. The page improvement
paid back through the entire
campaign economics, not just the
form fill rate.
INSIGHT 02
What a buyer calls their need is
not semantic. It is a page brief
WiFi, Internet, Fibernet, and
Broadband are four different buyer
mental models at four different stages
of the decision. A page that treats
them as interchangeable gets the
argument wrong for all of them.
Matching the page to the category
was what unlocked the 32%
acquisition uplift.
INSIGHT 03
Regional content is the trust signal
that supports the connection
decision
In metros, Netflix is a strong signal. In
Tier 2/3 Southern cities, regional
OTTs like SUN TV resonate more.
Showing low-relevance content for a
region means the page fails the last
mile of conversion even when
everything else is right.
