Payments · Revenue Systems · Analytics Leadership

Payments Revenue Operating System

I own and operate the measurement and decision infrastructure that drives revenue, margin, and cash flow across subscription payment systems. Across this work I have owned optimization across billions in annual TPV and a eight-figure subscription revenue base — directly influencing billthrough, net revenue, and EBITDA across global merchant programs. I influenced product, engineering, finance, and operations roadmaps across teams without direct reporting lines. The tools and frameworks on this portfolio are a working demonstration of that system — covering attribution, optimization, allocation, monitoring, control, and forecasting.

$100M+
Revenue recovered
Across merchants processing billions in annual payment volume
50K+
Subscribers recovered
From passive churn across major subscription platforms
60+
Markets
Carrier billing, wallets, and local payment methods
+2–10pp
Billthrough improvement
By market and decline reason — each 1pp is material revenue
+2–8bps
EBITDA improvement delivered
Across key merchant programs; higher in under-optimized markets

Payment optimization is one of the highest ROI levers in subscription businesses — often outperforming CAC reduction or pricing changes on a risk-adjusted basis, with payback measured in weeks rather than quarters.

Across this system, I have directly influenced $100M+ in annual revenue recovery and delivered +2–8 bps EBITDA improvement across key merchant programs, with higher impact in under-optimized markets — at merchants processing billions in annual payment volume. Improvements in billthrough directly increase NRR (Net Revenue Retention), extend customer lifetime, and reduce CAC payback periods.

Most subscription businesses measure authorization rate. Authorization rate tells you whether a single payment attempt succeeded. It does not tell you whether you got paid.

Billthrough — the rate at which subscribers are successfully billed across all retry attempts over a billing window — is the metric that maps to revenue. The delta between authorization rate and billthrough is where material revenue is lost, and it is invisible until you classify it.

At scale, a 1pp improvement in billthrough is not a metric improvement. It is a high-margin incremental EBITDA contribution, multiplied by subscriber LTV, compounded across every market where it applies.

At scale, payment inefficiencies represent a recurring revenue leakage equivalent to 2–8% of topline for most subscription businesses — with direct, measurable impact on EBITDA margin and operating cash flow. Most of it is recoverable. Every percentage point of billthrough improvement increases NRR, extends subscriber lifetime, and reduces the effective CAC payback period.

Revenue Leakage Sources

Unclassified declines retried incorrectly High
Retry timing misaligned to pay periods High
False positive declines from risk model Medium
PSP routing not optimized by decline type Medium
Identity state not linked to payment recovery Lower

We prioritized classification accuracy over retry timing early in the program — accepting slower initial recovery in exchange for a taxonomy that made every downstream decision faster and more defensible. That sequencing choice alone changed where the business invested for the following two years.

Layer Function Tool Financial Impact Typical Scale
Attribution Classify every decline as retryable, actionable, or terminal Decline Classifier → Eliminates undifferentiated retries; stops waste on terminal failures 1–3pp billthrough lift; $500K–$2M revenue recovery
Optimization Model revenue impact of billthrough changes net of retry fees Billthrough Simulator → Quantifies EBITDA impact; builds investment case for engineering resources 2–5pp modeled improvement; +20–80bps EBITDA
Allocation Score issuer performance; identify where failure is systemic Issuer Scorecard → Surfaces issuer-level negotiation levers; prioritizes relationship investment Issuer-specific +1–4pp auth rate; fee reduction
Monitoring Detect anomalies in authorization, settlement, and fraud signals Anomaly Detector → Catches revenue degradation before it compounds; triggers investigation Prevents $100K–$1M+ revenue erosion per incident
Control Self-serve payments data Q&A for non-technical stakeholders Payments Q&A → Consistent metric answers across product, finance, and operations 70% reduction in ad hoc data requests
Forecasting Model subscriber LTV and revenue retention curves by cohort Cohort Modeler → Connects billthrough improvements to NRR; values churn prevention in LTV terms 1pp billthrough = 8–15 months incremental LTV per sub

The Payments Stack Audit generates a ranked action list with estimated annual revenue return and investment level for each initiative. Representative outputs at a mid-market subscription business:

Initiative Est. Investment Est. Annual Return ROI Multiple Payback EBITDA Impact Effort
Retry logic optimization — decline taxonomy + timing $20K–$50K $800K–$2M 20–40× <1 month +30–80bps margin Low–Medium
Billing configuration audit — PSP settings, retry caps $20K $400K–$1M 20–50× <1 month +15–40bps margin Low
Routing intelligence — PSP selection by decline type $150K–$300K $500K–$1.5M 3–10× 2–4 months +20–60bps margin Medium–High
Signal enrichment — tokenization + 3DS optimization $150K $300K–$800K 2–5× 3–6 months +12–30bps margin Medium
False decline reduction — issuer data sharing $300K–$500K $600K–$2M 2–4× 6–12 months +25–80bps margin High

Operating under constrained engineering capacity, I prioritized initiatives based on ROI and payback thresholds — making explicit trade-offs between higher-return but lower-effort interventions and structural changes requiring significant engineering investment. The sequencing decisions were as consequential as the initiatives themselves.

Returns are illustrative ranges for a subscription business processing $100M–$500M annually. Investment estimates reflect engineering effort level: Low = ~$20K (1–2 engineer sprints), Medium = ~$150K (one quarter, small team), High = ~$500K (multi-quarter cross-functional program). The Payments Stack Audit generates return figures calibrated to your specific volume, ATV, and optimization maturity.

This system is not a reporting layer. It is an operating cadence — a set of decisions made at defined frequencies by specific stakeholders, each with a measurable revenue consequence.

Decision Frequency Who Uses It What It Drives Revenue Impact
Retry logic update — timing, suppression rules, attempt caps Monthly Payments Analytics, Engineering Billthrough rate improvement; fee cost reduction +$500K–$2M/yr
Decline taxonomy refresh — reclassify new failure codes Monthly Payments Analytics, Operations Accurate retry targeting; reduced terminal decline waste +$200K–$800K/yr
PSP routing change — reallocate volume by decline segment Quarterly Payments Analytics, Product, Finance Authorization rate improvement; net revenue per transaction +$300K–$1.5M/yr
Issuer performance review — identify underperformers Quarterly Payments Analytics, C-Suite Escalation to PSP account team; issuer-level relationship action +$100K–$500K/yr
PSP contract / fee renegotiation — backed by billthrough attribution data Annual Finance, Payments Analytics, Legal Fee reduction; SLA enforcement; concentration risk management +$200K–$1M/yr
Payment expansion / market entry decision Annual Product, Finance, C-Suite Method coverage; market billthrough baseline; LTV model by market Strategic
On metrics
Billthrough is the metric. Authorization rate is an input. The gap between them is where revenue is lost and where the work lives.
On diagnosis
Most revenue loss is invisible because it is uncategorized. Decline taxonomy is infrastructure, not analysis. You cannot optimize what you cannot classify.
On investment
Retry programs have a P&L. Gross recovery minus fee cost minus incremental churn risk is what actually hits EBITDA. Both sides of the equation matter.
On AI in analytics
The analytics team that owns the AI instruction layer owns the measurement standards for the entire organization. Canonical definitions become enforceable at the point of generation.
"The goal is not better reporting. It is building systems where the correct decision becomes the default — and where the financial impact of every intervention is visible before engineering resources are committed."
Andrea Register · Payments Analytics Leadership
Director, Business Analytics (Payments & Revenue)
Boku, Inc. · Denver · 2021–2025 · 4 years
Led global payments analytics across 60+ markets and 200+ local payment methods. Owned billthrough, authorization rate, and net revenue performance across major subscription merchants. Built the global decline taxonomy, established canonical KPI frameworks, and ran the cross-functional program that recovered $100M+ in revenue and 50,000+ subscribers from passive churn. Defined instrumentation standards and data quality practices across the payments data stack. Managed a team of analysts, data scientists, and technical PMs.
Head of Business Analytics
Boku, Inc. · 2019–2021 · 2 years
Built the BI function from early-stage tooling. Established canonical KPI definitions, metric governance, and self-serve analytics infrastructure. Reduced ad hoc request volume by 70% and increased dashboard adoption to 50% of the organization.
Product Manager, Checkout, Internal Tools & Payment Operations
Boku, Inc. · 2016–2019 · 3 years
Revenue recovery product management. 90% YoY revenue growth in managed accounts. Built internal tooling for payment operations. Contributed to IPO readiness analytics and reporting infrastructure.
Run the Payments Stack Audit → Open Billthrough Simulator Full Portfolio Get in Touch