Wanna See
Under the Hood?
Explore the Full Technical Analysis, methodology, code, and reproducibility. The complete project on GitHub
Showz – Marketing Cohort Analysis
Are We Really Sure Our Marketing Dollars Are Working?​
Marketing dashboards often look healthy on the surface — traffic is coming in, orders are being placed, and acquisition seems active.
But...
• Which channels bring high-quality users
The Reality Behind User Behavior
How engaged are users once they arrive?
At first glance, traffic volume looked healthy. However, user behavior told a more nuanced story.
I observed that:
• Session durations were heavily right-skewed
• Most users placed only one order
• Revenue distribution showed a classic long-tail pattern
What this means:
While the platform successfully attracts users, repeat engagement is limited. A small group of high-value users drives a disproportionate share of revenue — a common but risky growth pattern.
Business takeaway:
Growth based purely on acquisition volume may be overstated. Retention and user quality require closer attention.
When Do Users Actually Convert?
Next, I examined how quickly users move from first visit to first purchase.
Two key patterns emerged:
-
A large share of conversions happens very early
-
Conversion performance varies significantly by acquisition source
What this means
Users tend to make their purchase decision quickly. Channels that fail to convert users early are unlikely to recover value later.
Business takeaway
Early-funnel optimization is critical. Marketing teams should prioritize channels that demonstrate strong early conversion efficiency.
The Retention Reality Check
Cohort analysis revealed the most important insight of the study.
Across both monthly and weekly cohorts:
• Retention drops sharply after the first period
• Very few users remain active long-term
• Order volume fluctuates but does not translate into strong cohort stickiness
The business is effective at generating first purchases but struggles to build lasting customer relationships.
Without improvements in retention, scaling acquisition spend may produce diminishing returns.
Not All Channels Are Created Equal
When comparing acquisition sources, performance differences became clear.
Some channels delivered:
-
Higher average orders per user
-
Stronger average order value
-
More efficient acquisition costs
Others generated volume but weaker downstream value.
What this means
Looking only at traffic or conversion volume would mask meaningful differences in customer quality.
Business takeaway
Marketing budget allocation should be guided by unit economics, not just top-line conversions.
The Bottom Line: LTV vs CAC
Finally, I evaluated true marketing efficiency by comparing lifetime value against acquisition cost...
This is where the real story emerged:
• Only a subset of channels generated positive ROMI
• Some high-spend sources were underperforming
• Desktop users showed higher LTV than touch users
• Revenue remains concentrated among a small user segment
Not all growth is profitable growth. Several channels may be scaling inefficiently.
Reallocating budget toward high-ROMI channels could significantly improve marketing efficiency without increasing total spend.
Strategic Recommendations
1. Rebalance marketing investment
• Scale channels with positive ROMI
• Audit or reduce spend on underperforming sources
2. Focus on early conversion optimization
• Most purchase decisions happen quickly
• Improve landing page and onboarding performance
3. Prioritize retention initiatives
• Encourage second purchase behavior
• Introduce lifecycle marketing (email, remarketing, loyalty)
4. Segment by device experience
• Desktop users currently deliver higher value
• Investigate friction in touch/mobile journey
Wanna See Under
the Hood?
Explore the Full Technical Analysis methodology, code, and reproducibility.
The complete project on GitHub
What I've been doing since I graduated... until my transition into Data Analytics & the tripleten bootcamp
EXPLORE LCTB AGENCY
