HomeProjectsPaid Traffic Funnel Redesign

Education / Performance Marketing

Paid Traffic Funnel Redesign

Across three campaign variants, the funnel generated 1.04M reach, 14,885 purchases, and a blended $1.79 cost per result on $26.6K spend.

Year

2025

Role

Sr. Product Designer

Client

Zapiano

1.04M

Reach across all campaign variants

14,885

Purchases recorded across the test snapshot

$1.79

Blended cost per result

Overview

Structure

My Role

  • Sr. Product Designer
  • Design
  • Research
  • Strategy

Tools Used

  • Figma
  • ClickUp
  • Microsoft Copilot
  • Google Analytics
  • Meta Ads Manager

Timeline

2 to 3 months

Case Study

The problem

Zapiano is a Swiss online piano course platform for adults, founded by piano teacher Sven Haefliger. Meta campaigns were driving healthy reach (1.04M reach, 2.85M impressions), but paid landing conversion was stuck at 6% and most visitors bounced before any conversion event.

The existing page led with three Club subscription tiers side-by-side, dense feature blocks, and three competing CTAs. Cold traffic hit the highest-priced tier within the first scroll and left.

Context

Lift paid landing conversion and reduce CAC while working within two real constraints.

The site runs on Kajabi, so any redesign had to be modular and assemblable from existing content blocks. No custom engineering per variant.

The page had to serve two distinct user types Marketing identified in the ad funnel: complete beginners and adults returning to piano after years away. Same product, different motivations, same landing page.

The goal Marketing handed me was simple: more conversions, lower cost per acquisition, same ad spend.

Success Metrics

  • Increase paid landing conversion without increasing ad spend.
  • Reduce CAC enough to make scale more efficient.
  • Create a modular Kajabi page structure Marketing can keep testing post-launch.

6.0%

conversion rate on the original control campaign

$4.20

cost per purchase before the redesign and experiment

2,847

purchases generated by the baseline control variant

Methodology

Paid Funnel Redesign Sprint

1. Audit the funnel

2. Benchmark the category

3. Reframe the landing

4. Build modular components

5. A/B test social proof

Audit the funnel

Mapped Meta campaign data against Google Analytics flow reports to locate the drop-off. The pattern was clear: visitors scrolled past the hero, hit the EUR5220 Gold tier within seconds, and exited. The page was answering the wrong question first: is this about years of cost, or whether this is for me and how I start?

Funnel architecture | Before vs After

The strategic reframe: from a 3-tier upfront decision to a stepped, low-friction entry that mirrors daily practice

BeforeClub tier as paid entry

Cold paid traffic asked to choose a 1-year, 2-year, or 5-year commitment within seconds of landing.

Top of funnel

Meta Ad

~1.77M reach | 5.84M impressions across the test period

Landing page

Zapiano Club LP

All 3 tiers presented side-by-side. Dense feature blocks. Three competing CTAs.

EUR365 / EUR1044 / EUR5220 above the foldNo segmentationNo clear next step
6.00% conversion to purchase

Purchase decision

3-tier choice

Bronze (EUR365) | Silver (multi-year savings) | Gold (5-year commitment)

High cognitive loadHigh stakes

Outcome

Subscription (or bounce)

Most cold visitors bounced before this point.

Core problemUsers could not process the tier comparison on cold traffic. The page asked "how much will this cost over years?" instead of "is this for me, and how do I start?"
AfterStepped entry, then upsell

A EUR9 first action lowers the cost of saying yes. The big decision moves to a warmer point in the funnel.

Top of funnel

Meta Ad

Same campaigns, same spend, same audiences

Landing page

EUR9 Intro Course LP

Sven leads hero. Beginner / returner segmentation. Single CTA. Community proof in fold.

Founder credibilitySegmented copySticky CTA
25.00% conversion to purchase

First purchase

EUR9 Intro Course

One-time, transactional, low-friction. A clear first step that maps to daily practice.

Low cognitive loadHabit-forming

Warm upsell

PianoStarter EUR29/month

Recurring subscription offered after the user has experienced the product. Decision happens warm, not cold.

Core unlockThe biggest conversion lift came from changing what we asked visitors to decide, not from changing how we styled the buttons. The problem looked like UI. It was actually an offer problem.
Impact: 1.04M reach | 2.85M impressions | 14,885 purchases | $26,568.50 total spend | $1.79 blended cost per result

Benchmark the category

Before sketching anything, I audited five direct competitors against seven conversion heuristics that matter for cold paid traffic. The pattern was clear: Zapiano was the weakest landing in the category on every dimension that drives entry conversion.

Competitive benchmark

Adult piano learning category | 5 competitors evaluated across 7 UX and conversion heuristics

Strong
~Partial
Missing
Analysis: April 2025 | Method: competitive benchmark of paid traffic landing pages

Direct competitors

Simply Piano

hellosimply.com

Low-friction entry offer

Strong

Beginner vs returner path

Strong

Founder credibility upfront

Missing

Community proof on page

Partial

Pricing clarity

Strong

Mobile-first clarity

Strong

CTA visibility & clarity

Strong

Key takeaway

7-day free trial sets category baseline. Onboarding quiz segments by level.

Skoove

skoove.com

Low-friction entry offer

Strong

Beginner vs returner path

Partial

Founder credibility upfront

Missing

Community proof on page

Missing

Pricing clarity

Strong

Mobile-first clarity

Strong

CTA visibility & clarity

Strong

Key takeaway

Free tier as entry hook. Skips segmentation, treats all visitors as beginners.

Flowkey

flowkey.com

Low-friction entry offer

Strong

Beginner vs returner path

Strong

Founder credibility upfront

Missing

Community proof on page

Missing

Pricing clarity

Strong

Mobile-first clarity

Strong

CTA visibility & clarity

Strong

Key takeaway

Strong hero video. Beginner / intermediate / returner paths clearly labeled.

Yousician

yousician.com

Low-friction entry offer

Strong

Beginner vs returner path

Strong

Founder credibility upfront

Missing

Community proof on page

Partial

Pricing clarity

Partial

Mobile-first clarity

Strong

CTA visibility & clarity

Strong

Key takeaway

10M+ user social proof. Pricing transparency criticized in user reviews.

Pianote

pianote.com

Low-friction entry offer

Strong

Beginner vs returner path

Strong

Founder credibility upfront

Strong

Community proof on page

Strong

Pricing clarity

Strong

Mobile-first clarity

Strong

CTA visibility & clarity

Strong

Key takeaway

Category gold standard. Real teachers in hero, community is core value prop.

Zapiano

ZapianoBefore

zapiano.com (original LP)

Low-friction entry offer

Missing

Beginner vs returner path

Missing

Founder credibility upfront

Missing

Community proof on page

Partial

Pricing clarity

Missing

Mobile-first clarity

Partial

CTA visibility & clarity

Missing

Key takeaway

- High upfront cost

- Competing membership tiers

- No audience segmentation before users understood the value

ZapianoAfter

zapiano.com (redesign)

Low-friction entry offer

Strong

Beginner vs returner path

Strong

Founder credibility upfront

Strong

Community proof on page

Strong

Pricing clarity

Strong

Mobile-first clarity

Strong

CTA visibility & clarity

Strong

Key takeaway

€9 intro course as paid entry. Beginner / returner path. Sven leads hero. Community block in fold.

Score summary | Before: 0 strong, 2 partial, 5 missing of 7 dimensions. After: 7 strong of 7. Closed all 4 critical gaps versus Pianote category benchmark.

The redesign brief wrote itself: close the four critical gaps of entry friction, segmentation, founder visibility, and community proof while keeping what was already working, including risk reversal and content depth.

Reframe the landing

The Club tiers were the real conversion blocker, so I recommended killing them as the cold-traffic entry, leading with the €9 Introduction Course, and upselling to PianoStarter at €29/month after the intro.

Users did not need three options to compare, they needed one clear next thing to do. The team approved the change, and I redesigned the landing around the new entry:

- beginner vs returner segmentation

- one primary CTA for the €9 intro

- email gate below social proof

- community credibility in the fold instead of three competing pricing blocks

BeforeZapiano landing redesign reference

Build modular components

AfterZapiano landing redesign after applying benchmark and funnel improvements

The redesign combined a funnel audit, competitor benchmark, offer reframe, modular landing-page redesign, and post-launch testing. Reduce the cognitive cost of the first yes by leading with a single low-friction paid entry instead of multiple long-term commitments.

Make founder credibility, beginner-vs-returner relevance, and community proof visible early enough to help cold traffic orient before price friction hits. Design the page as a modular no-code system Marketing could keep testing inside Kajabi without needing engineering for every iteration.

A/B test social proof

Variant A used physical community proof through annual Zapiano member meetup photos in Switzerland and Germany. Variant B used digital community proof through mobile app screenshots, member feed views, and in-product social proof.

A/B testing comparison

Same offer, same audience, two different social proof treatments.

Social proof treatment

Variant A

Physical community proof

Variant B

Digital community proof

Why it mattered

The test isolated proof format, not the offer.

Creative used

Variant A

Annual Zapiano member meetup photos in Switzerland and Germany

Variant B

Mobile app screenshots, member feed views, and in-product social proof

Why it mattered

Variant B made the product itself visible much earlier.

Audience signal

Variant A

Warm, authentic, but implied in-person participation

Variant B

Immediate, tangible preview for cold, mobile-first visitors

Why it mattered

Cold paid traffic responded better to product tangibility than community warmth.

Conversion outcome

Variant A

Lower-performing variant

Variant B

Winning variantWinner

Why it mattered

Variant B won decisively and became the rollout direction.

Once both variants had statistically meaningful data, I used Copilot to compare scroll depth, time-to-CTA, and segment behavior across variants. It compressed the slowest part of the testing cycle, synthesis, without replacing the design judgment that came before it.

Results & Impact

$26,568.50

Amount Spent

~5.0x

Return on Ad Spend (ROAS)

14,885

Purchases

~$0.25

Cost per Click (CPC)

$1.79

Cost per Result (CPR)

1.04M

Reach

2.85M

Impressions

Enrollment Behavior Insights

  • Variant B won decisively because mobile app screenshots gave cold visitors a tangible preview of what they were buying.
  • The meetup photos in Variant A were warmer and more authentic, but they implicitly required showing up in person, a commitment cold visitors were not ready to make.
  • The lift did not come from visual polish alone. It came from reordering the decision the page asked users to make.

Opportunities and Next Steps

  • Personalized hero copy based on ad creative so beginner ads route to a beginner hero and returner ads route to a returner hero.
  • A pre-paywall qualifier quiz with 3-4 questions on goals and level to lift intro-to-subscription conversion further down the funnel.
  • AI-generated copy variants for headline A/B testing at scale.

Projected Improvements

  • Conservative 1-month scenario: with a 5,500 to 6,000 media budget and using the blended dashboard CPR of $1.79 rather than the winning-variant low of $0.85, the campaign would project roughly 3,070 to 3,350 intro-course purchases.
  • At a €9 entry ticket, that implies about €27.6K to €30.2K in top-of-funnel revenue before any PianoStarter upsell or downstream LTV is counted.
  • Using the preserved dashboard delivery ratios as a baseline, that same budget range would conservatively translate to roughly 215K to 235K people reached and about 590K to 645K impressions over one month.
  • Primary markets worked in this setup: Germany, Switzerland, Sweden, and the Netherlands.
  • Recommended targeting mix for a conservative next-month run: broad Advantage+ / algorithmic cold audiences for scale, piano-learning and adult-beginner interest clusters for control, plus warm retargeting from site visitors or CRM / first-party lists when available.
  • Estimated audience size by targeting type: broad Advantage+ cold audiences in these four markets can usually support a combined reachable pool in the low millions, interest-based pools are typically narrower but still large enough for monthly testing, and first-party / retargeting pools depend on site traffic and database volume.

Conservative 1-Month Scenario

Budget

$5,500

Reach

~216K

Purchases

~3,073

Revenue

~€27.7K

Budget

$6,000

Reach

~236K

Purchases

~3,352

Revenue

~€30.2K

Learnings

The bottleneck was the offer, not the UI. The Club tier setup was not a UI problem. It was an offer problem dressed up as a layout problem. The biggest conversion lift came from changing what visitors were being asked to decide.

Cold traffic needs a low-cognitive first step. On cold paid traffic, the landing page's first job is not to sell everything. It is to lower the cognitive cost of saying yes. The EUR9 intro offer did that far better than a multi-tier subscription choice.

Proof type matters as much as proof volume. Variant B won because the mobile app screenshots gave cold visitors a tangible preview of what they were buying. The meetup photos in Variant A were warmer, but they implied a commitment visitors were not ready to make yet.

Tools

FigmaClickUpMicrosoft CopilotGoogle AnalyticsMeta Ads Manager

Tags

GrowthPaid AcquisitionExperimentationLanding Pages

More work

Other Projects