Customer Conversion Simulation
Use research and public benchmarks across paid, referral, influencer, ecommerce, SaaS, app, and consultation funnels to forecast conversion ranges before spending.
Written and reviewed by Eric Han · Co-founder & CTO, reopt
Before spending on advertising, someone asks: “How many conversions will this budget buy?” The usual answer draws on a past campaign or an average found online. Less attention goes to whether the rate uses sessions or users as its denominator, or how the source defines a customer.
This handbook collects published research and data on typical conversion rates by role, product, and channel, then explains how to turn those figures into forecast ranges rather than single numbers.
Who this is for
Performance marketers, growth leads, founders, and RevOps teams who need evidence-based conversion assumptions for budgets and channel selection. Averages and ratios are enough to use the calculators. Research terms are explained in How to Read the Evidence.
Three Things to Know First
| Finding | Figures | Implication |
|---|---|---|
| The same metric differs across datasets | Overall ecommerce conversion: 1.3%, 1.4%, and 1.97% | Align denominators and samples before comparing |
| Platform-reported conversions differ from actual effects | Meta last-click estimates were 1.2–2.0 times randomized estimates for large US advertisers in 2019–20; vendor experiments also show the opposite direction | Apply an incremental adjustment range instead of using benchmark CVR directly |
| Published effects can be inflated | An estimate from about 6,700 retail and travel experiments placed 90% of effects within ±1.2% (Qubit customers, grade B). Public-sector studies found 8.7 percentage points in journals versus 1.4 in operational RCTs | Do not put case-study lifts straight into a plan |
Most role-specific conversion evidence is missing
We found no solid public aggregates for free-to-paid or win rates by job role. This book leaves those gaps visible in the Conversion Atlas by Role and explains how to fill them through internal experiments. Where a simulation requires an input, use a wide interval explicitly labeled as judgment. We also found no official industry benchmarks from Naver or Kakao, Korean SaaS trial-conversion benchmarks, or stage-by-stage conversion benchmarks for early startups. Available Korean figures are collected in the Korean Conversion Atlas.
Handbook Structure
Short on time? Find figures for your situation in the Conversion Atlas; for consultation and application funnels, start with Consultation and Photo-Upload Forms, and for local figures, the Korean Conversion Atlas. Use How to Read the Evidence to judge reliability, choose concentration k from the table in Conversion Rate Intervals, and follow the Worked Scenarios.
Contents
Reading Guide
Ch1. How to Read the Evidence
Evidence grades A–D, aligned denominators, rules for benchmarks, and a glossary.
Ch2. Conversion Rate Intervals
Choose concentration using mean × k and simulate both expected and actual counts.
Conversion Atlas
Ch3. Conversion Atlas by Channel
Conversion rates and costs for search, Meta, TikTok, LinkedIn, email, referrals, and consultation forms.
Ch4. Conversion Atlas by Product
Conversion ranges for ecommerce categories, SaaS models, app subscriptions, AI products, and services.
Ch5. Conversion Atlas by Role
Figures and gaps in public evidence for CMOs, developers, HR, finance, and founders.
Ch6. Korean Conversion Atlas
Where Korean figures are available, possible inputs, values to avoid, and missing evidence.
Paid Marketing
Ch7. Paid Channel Baselines
Industry CTR, CPC, CVR, and CPL tables, and the state of Korean channel data.
Ch8. Reported and Incremental Conversions
Adjustment multipliers from randomized experiments at eBay, Meta, and Google.
Ch9. Creative, Retargeting, and Saturation
Evidence on targeting fit, retargeting timing, frequency, and diminishing returns.
Referral Marketing
Ch10. Referral Funnel Metrics
Sharing rates, invitation conversion, time to first referral, and denominator differences.
Ch11. Referral Experiments and Limits to Spread
Referred-customer value, incentive experiments, and why K rarely exceeds 1.
Partners and Influencers
Ch12. Influencer Marketing Conversion
The source of the $5.78 claim, code attribution versus incrementality, follower size, disclosures, and live commerce.
Ch13. Partnerships, Collaborations, and Sponsorships
Affiliate incrementality, B2B partner win rates, podcasts, newsletters, and brand collaborations.
Ecommerce
Ch14. Ecommerce Conversion Baselines
Overall, category, and device conversion rates, plus Korean statistics.
Ch15. Ecommerce Funnels and Carts
Add-to-cart and checkout completion rates, and what cart abandonment really means.
Ch16. Effect Sizes of Ecommerce Levers
Speed, reviews, delivery, discounts, and realistic A/B test effects.
SaaS
Ch17. Self-Serve SaaS Conversion
Freemium, trials, credit-card requirements, activation, and randomized experiments.
Ch18. Early-Stage SaaS Conversion
Retention, growth, first revenue, and limits of benchmark discounts from prelaunch to $1 million ARR.
Ch19. Sales-Led Conversion
Lead-to-opportunity-to-win conversion, lead-source differences, and cold-email stage rates.
Apps
Ch20. Mobile App Acquisition and Retention
Store conversion, Apple Ads costs, retention, permissions, and new-app revenue distributions.
Ch21. App Subscription Conversion
Download-to-trial-to-paid conversion, paywalls, AI apps, and Korean regulatory effects.
Ch22. Desktop App Conversion
Downloadable software funnels, Steam wishlists and refunds, and installation warnings.
Consultation and Application Forms
Building Forecasts
Ch24. Seven Worked Scenarios
Calculate D2C, B2B SaaS, referral, app, outbound, Korean ecommerce, and consultation-form scenarios.
Ch25. Expert Forecasts and Calibration
Reference-class forecasting, small expert ensembles, pilot sizes, and updates.
Ch26. Choosing a Forecasting Model
Requirements and limits of Bass, BG/NBD, agent-based models, and MMM.
Ch27. LLM Synthetic Customers
Evidence on what synthetic customers predict well or poorly, including replication in Korea.
Korean Case Studies
Ch28. Korean Cases: Advertising and Commerce
Naver log analysis, ad-platform cases, store conversion and repeat purchases, and live commerce.
Ch29. Korean Cases: Subscriptions, B2B, Messaging, and Referrals
Millie's Library filings, company-reported experiments, B2B, email, AlimTalk, referrals, and partnerships.
Ch30. Korean Cases: Consultations, Finance, and Public Experiments
Hospital no-shows, insurance, lending, consultation cases, and Korean randomized experiments.