Where should a subscription business invest next?
I examined when people first paid, where the purchase journey lost them, and what supported recurring value.
- My role
- End-to-end analytics
- Tools used
- SQL · Excel · Tableau
- Deliverable
- Analysis & recommendations
The problem
Leadership needed a connected view of subscription growth: more first purchases, a better purchase journey, and value worth renewing.
What I did
- Compared registration cohorts, funnel stages, benefit-specific demand, and payer value.
- Turned the evidence into recommendations for onboarding, offer clarity, and retention.
The first week was a key conversion opportunity.
61% of eventual first-year buyers purchased by day 7. That is a timing signal among buyers; only 4% of all registrants bought within the year.
An early window within a longer purchase journey
View data & definitions
100,000 new registrants, including 4,000 eventual first-year buyers. The plot shows days 0–90; day 0 is registration day. The table also includes the full-year checkpoint.
| Day after registration | Cumulative buyers | Of eventual buyers | Of all registrants |
|---|---|---|---|
| 0 | 1,689 | 42.2% | 1.69% |
| 3 | 2,216 | 55.4% | 2.22% |
| 7 | 2,449 | 61.2% | 2.45% |
| 14 | 2,730 | 68.3% | 2.73% |
| 30 | 2,986 | 74.7% | 2.99% |
| 60 | 3,330 | 83.3% | 3.33% |
| 90 | 3,462 | 86.6% | 3.46% |
| 365 | 4,000 | 100.0% | 4.00% |
Recommendation: Make early guidance useful at a real backup or photo-management moment, and judge acquisition quality beyond traffic.
How first purchases, renewals, and returning payers contribute
Different revenue streams need different growth strategies
View data & definitions
| Purchase type | Revenue share |
|---|---|
| Automatic renewal | 62% |
| First purchase | 24% |
| Returning payer | 14% |
Payment success improved while fewer users reached checkout.
Payment success rose from 80.0% to 86.0%. Checkout reach fell from 7.2% to 4.7%, and the overall active-user purchase rate edged down.
Read the whole funnel, not just the payment screen
View data & definitions
Daily-average counts over nine months. These are not monthly unique users. Panels use different percentage scales; stage rates multiply to the end-to-end rate.
| Month | Active users | Checkout visitors | Payers | Checkout reach | Purchase intent | Payment success | End-to-end |
|---|---|---|---|---|---|---|---|
| M1 | 450,000 | 32,400 | 1,840 | 7.2% | 7.1% | 80.0% | 0.41% |
| M2 | 458,750 | 31,195 | 1,833 | 6.8% | 7.3% | 80.5% | 0.40% |
| M3 | 467,500 | 29,920 | 1,822 | 6.4% | 7.5% | 81.2% | 0.39% |
| M4 | 476,250 | 29,051 | 1,752 | 6.1% | 7.4% | 81.5% | 0.37% |
| M5 | 485,000 | 28,130 | 1,852 | 5.8% | 8.0% | 82.3% | 0.38% |
| M6 | 493,750 | 27,156 | 1,871 | 5.5% | 8.3% | 83.0% | 0.38% |
| M7 | 502,500 | 26,130 | 1,883 | 5.2% | 8.6% | 83.8% | 0.37% |
| M8 | 511,250 | 25,051 | 1,871 | 4.9% | 8.8% | 84.9% | 0.37% |
| M9 | 520,000 | 24,440 | 1,913 | 4.7% | 9.1% | 86.0% | 0.37% |
Recommendation: Investigate the path into checkout and how benefits are explained, alongside payment-screen performance.
Different membership needs produced different kinds of value.
Device migration had the highest checkout conversion, at 25.5%. Storage expansion generated the most revenue per payer, at 34 CU. A single offer or audience rule would miss that distinction.
Separate willingness to purchase from value per payer
View data & definitions
Membership benefit entry points. Bubble area represents payer count. Reference lines show an illustrative standard entry at 16% conversion and 24 CU per payer; they are not universal targets. On narrow screens, use the numbered key below.
| Key | Benefit | Checkout → paid | Revenue / payer | Payers |
|---|---|---|---|---|
| 1 | Large-video backup | 11.8% | 18 CU | 850 |
| 2 | Private vault | 18.3% | 31 CU | 369 |
| 3 | Storage expansion | 23.0% | 34 CU | 580 |
| 4 | Fast backup | 14.9% | 14 CU | 340 |
| 5 | Ad-free browsing | 11.8% | 29 CU | 140 |
| 6 | Device migration | 25.5% | 12 CU | 390 |
| 7 | Original-quality export | 16.6% | 22 CU | 255 |
- 1Large-video backup
- 2Private vault
- 3Storage expansion
- 4Fast backup
- 5Ad-free browsing
- 6Device migration
- 7Original-quality export
Recommendation: Test benefit-specific entry points and make the monthly and annual choices easier to understand.
Compare benefit discovery and click-through
A small audience can still show strong intent
View data & definitions
| Key | Benefit | Exposure / eligible active users | Entry click-through |
|---|---|---|---|
| 1 | Large-video backup | 24.0% | 30.0% |
| 2 | Private vault | 4.2% | 48.0% |
| 3 | Storage expansion | 6.0% | 42.0% |
| 4 | Fast backup | 9.5% | 24.0% |
| 5 | Ad-free browsing | 8.5% | 14.0% |
| 6 | Device migration | 3.0% | 51.0% |
| 7 | Original-quality export | 4.8% | 32.0% |
See the proposed pricing communication concept
One benefit set. Two understandable choices.
Billed monthly.
Renews at 9 CU per month.
Cancel future renewal anytime.
6.50 CU per month equivalent.
Billed 78 CU once per year.
Renews at 78 CU per year.
Design concept, not a tested pricing change. CU denotes currency units.
Evaluate a change with net revenue per eligible user, including conversion, cannibalization, refunds, and costs. Annual-plan take-up alone is not the objective.
Retention work began at the first payment.
Auto-renewal opt-outs clustered early. Opting out stops a future automatic payment; it does not mean the customer immediately loses access or has already churned.
Understand when users opt out—and what that event means
View data & definitions
8,000 new monthly subscribers observed for 90 days. Both panels use the original cohort as denominator. The daily panel is not a risk-set hazard rate; opt-out is not immediate churn.
| Day after first payment | Daily opt-outs / initial cohort | Cumulative opt-outs / initial cohort |
|---|---|---|
| 0 | 17.45% | 17.4% |
| 1 | 1.19% | 18.6% |
| 7 | 0.91% | 25.5% |
| 14 | 0.76% | 31.4% |
| 30 | 4.41% | 50.3% |
| 60 | 1.94% | 60.4% |
| 90 | 0.09% | 65.0% |
Recommendation: Understand early opt-outs and different payer needs before choosing a blanket retention incentive.
Explore 12-month payer value and payment segments
Measure value over a defined horizon
View data & definitions
| Follow-up month | Cumulative revenue / initial payer | Payers in month |
|---|---|---|
| 1 | 9.00 CU | 10,000 |
| 2 | 15.30 CU | 7,000 |
| 3 | 20.84 CU | 6,152 |
| 4 | 25.61 CU | 5,306 |
| 5 | 29.66 CU | 4,500 |
| 6 | 33.24 CU | 3,971 |
| 7 | 36.34 CU | 3,448 |
| 8 | 38.97 CU | 2,919 |
| 9 | 41.07 CU | 2,340 |
| 10 | 42.69 CU | 1,800 |
| 11 | 44.31 CU | 1,800 |
| 12 | 45.93 CU | 1,800 |
A value view should retain the full original cohort
View data & definitions
| Observed payments | Initial payers | Share |
|---|---|---|
| One payment | 3,000 | 30% |
| 2–4 payments | 2,500 | 25% |
| 5–9 payments | 2,700 | 27% |
| Full year | 1,800 | 18% |
Match the growth lever to the part of the journey that needs it.
- First purchase: connect the offer to a useful photo-management need at the right moment.
- Conversion: improve checkout reach and offer clarity, as well as payment success.
- Recurring value: understand opt-outs and payer needs before selecting retention tactics.
What I delivered: A unified monetization diagnosis and recommendations. The report does not claim measured implementation uplift.
Methods, definitions & source notes
Original tools: SQL, Excel, and Tableau. The examples use separate registration, funnel, new-subscriber, and payer-value populations; their rates should not be combined.
- The first-purchase curve is conditional on buying within a year, rather than all registrants.
- Funnel counts are daily averages across nine months, not monthly unique users. The four panels use different scales.
- Opt-out is distinct from immediate churn. The value chart uses a separate group of initial monthly-plan payers and a fixed 12-month revenue horizon, not lifetime profit.
- Benefit audiences may overlap; channel mix and self-selection can affect comparisons.
The English presentation and vector charts are newly produced. Original employer reports, screenshots, SQL, and customer data remain private.
Download chart data