Sorbet — Case Study · Noetic Creative
Noetic CreativeSorbet
Paid SocialCreativeData & Attribution

One event,
eight states.

Fragmented campaigns, an iOS blind spot and a conversion event nobody agreed on. We consolidated all three, and the applications followed.

−20%Acquisition cost
The Sorbet application as the customer completes it
The completed application · The conversion event everything was rebuilt around
ServicesPaid Social · Creative · Data & Attribution
IndustryFintech · Consumer lending
ClientPublished case study ↗
12% → 104%Booked applications
69% → 182%Approved applications
−20%Client acquisition cost
48%Return on ad spend
The Brief

Sorbet could generate leads. What it could not do was scale the thing that was working.

Before engaging with Noetic, Sorbet was struggling to scale its initial success generating leads from paid media. The pattern was familiar: something worked once, and nothing about the setup made it possible to find out why.

Learning what changes were needed, making those changes, and tracking performance over time is what allowed Sorbet to make its digital advertising predictable rather than lucky.

The Challenge

Fragmented across
eight states.

Three problems, each hiding the other two.

01

Eight states, eight campaigns

Spend and signal were split across eight state-level campaigns, so no single one ever gathered enough data for the platform to learn from.

02

An iOS14+ blind spot

Visibility problems post-iOS meant the numbers coming back were incomplete, and the optimisation was running on them anyway.

03

Overlapping ad sets

Audiences competed with each other, which inflated cost and muddied every read on what was actually performing.

We needed to consolidate the data before we could trust any of it.

The Strategy

Media, then data,
then content.

Three pillars, fixed in that order, because each one gated the next.

01

Media optimisation

The right audience targeting and segmentation, iterative testing, and compliance with ad platform policy — which in a credit category is not a footnote.

02

Data: custom events

What actually qualifies as a conversion? A form fill, or a completed application? We defined the event, then fed it back into the algorithm properly.

03

Content

Ad content had to be dynamic, unique and relevant to the viewer. The best performance came from creative that engaged before it asked.

A grid of Sorbet ad variants: different colours, hooks and offers tested against the same audience
Dynamic, not uniform · A sample of the variants running at once, each earning its place on performance rather than a brand-guideline default
The Creative

UGC did the
heavy lifting.

Sixty percent of total conversions came from creator content — and our approach to the credit category restrictions produced the best performing ad of the year.

Sorbet UGC still: a creator opening on the problem rather than the product
Problem first · The hook never opens on the offer
Sorbet UGC still: the application walked through on camera
Shown, not claimed · The application, walked through
Sorbet UGC still: a creator framing the outcome in plain terms
Plain terms · No category jargon, no small print voice
Sorbet UGC still: the best performing variant of the year
Best of 2023 · The variant the category restrictions produced
Sorbet · UGC still
Where the Conversions Came From

Creator content
carried it.

The split that decided where the creative budget went.

The credit category restricts what an ad can say. Creator content was how we said it anyway.

The Numbers Have the Floor

Predictable,
at last.

Booked applications

8.7× lift
Before
12%
After
104%

Approved applications

2.6× lift
Before
69%
After
182%
+27%

Approved applications from custom events alone.

−20%

Client acquisition cost after consolidation.

48%

Return on ad spend from a predictable system.

60%

Total conversions produced by UGC.

Eight campaigns learning separately will always lose to one learning properly.
Straight from the Source
The team at Noetic is incredibly capable. We were able to hit our KPIs and systematically increase performance goals, as well as media spend.
Cullen Bray Cullen BrayCMO, Sorbet
Martech Stack

The toolsbehind the work.

  • Meta Ads
  • Google Ads
  • GA4
  • GTM
  • HubSpot
Project Summary

Define the event, consolidate the signal, then let the creative work. In that order.

Today Sorbet runs a much smoother and more predictable lead generation programme on Meta, and the return on ad spend moves with decisions rather than with luck.

Other Case Studies