How InsightRaider Estimates Product Revenue
"How do you know how much a product makes?"
It's the question we get most often. And it's a fair one. Platforms like Gumroad, Systeme.io, and Whop don't publicly share revenue data. So how can we estimate it?
This article is our commitment to transparency. No black boxes. No hand-waving. Here's exactly how our methodology works, what signals we analyze, and what the limitations are.
The Inspiration: BrandSearch
Before building InsightRaider, we studied how existing companies estimate revenue for private businesses.
The gold standard is BrandSearch, a company valued at $110M that estimates revenue for Shopify stores and Amazon sellers. They've proven it's possible to achieve +/-20% accuracy by combining public signals with proprietary algorithms.
Our approach adapts their methodology for the infoproduct market. If you want to see how to apply these estimates in practice, check our 48-hour validation framework.
The Three Pillars of Our Methodology
Pillar 1: Public Signal Scraping
Digital product platforms expose more data than you'd think. We systematically collect:
Ranking Data
- Position in category bestseller lists
- Position in overall platform rankings
- Trending/hot product flags
Social Proof Metrics
- Number of customer reviews
- Average rating score
- Rating distribution (1-5 stars)
- Review velocity (new reviews over time)
Product Metadata
- Price points
- Creation/launch date
- Creator follower count
- Number of products from same creator
Why this matters:
There's a strong correlation between public signals and revenue. For example:
- Products in the top 10 of a category consistently outperform products ranked 11-50
- Review counts correlate with sales at roughly 2-5% (meaning 100 reviews = 2,000-5,000 customers)
- Rating scores above 4.5 correlate with 40% higher conversion rates
These aren't opinions. These are patterns extracted from 146,000+ products we track.
Pillar 2: Web Traffic Analysis
Revenue estimation requires understanding how many people see a product page. We analyze:
Traffic Volume
- Estimated monthly visitors to product pages
- Traffic trends over time
- Traffic sources (organic, social, paid)
Engagement Signals
- Time on page estimates
- Bounce rate indicators
- Return visitor patterns
We source traffic data from:
- Third-party analytics APIs (similar to SimilarWeb)
- Backlink analysis tools
- Social media API data
- Search engine ranking data
The conversion formula:
At its simplest:
Estimated Revenue = Traffic x Conversion Rate x Average Order Value
For digital products, industry conversion rates typically range from 1-4% depending on traffic temperature:
- Cold traffic: 1-2%
- Warm traffic (email, returning): 3-5%
- Hot traffic (referrals, affiliates): 5-10%
We calibrate our conversion rate assumptions based on traffic source mix.
Pillar 3: Cross-Validation Algorithms
This is where it gets sophisticated. We don't rely on a single estimation method -- we run four different models and triangulate the results.
Model A: Ranking-Based Estimation
We've built correlation maps between ranking position and revenue across thousands of products. A product ranked #3 in a category has predictable revenue ranges based on historical data.
Model B: Review-Based Estimation
Using the review-to-customer ratio (typically 2-5%), we estimate total customers from review counts. Multiply by price, adjust for refunds, and you've got a revenue estimate.
Model C: Traffic-Based Estimation
Pure math: traffic x conversion x price. We refine conversion assumptions based on product category, price point, and traffic source mix.
Model D: Machine Learning Ensemble
Our ML model is trained on products where creators have publicly shared their revenue numbers (from podcasts, tweets, Indie Hackers posts, etc.). This ground truth data lets us calibrate our other models.
Final estimate: We weight the four models based on data availability and confidence level. If we have strong traffic data, Model C gets more weight. If review counts are more reliable, Model B dominates.
Validation: How We Know It Works
We continuously validate our methodology against known data points.
Public revenue disclosures:
Creators like Pieter Levels, Tony Dinh, and Marc Louvion share their revenue publicly. We compare our estimates to their disclosures:
- Our estimates for known products are within +/-15-25% accuracy
- We update our models when discrepancies appear
- We track prediction accuracy over time to identify model drift
Creator feedback:
When creators using InsightRaider verify our estimates for their own products, we incorporate this feedback (anonymized) into our calibration process.
A/B testing:
We regularly test different model weightings against holdout datasets to optimize accuracy.
Accuracy Expectations
Let's be honest about what you can -- and can't -- expect:
What our estimates tell you:
- Order of magnitude (is this a $1k/month or $10k/month product?)
- Relative comparison (Product A likely earns more than Product B)
- Trend direction (is revenue growing, stable, or declining?)
- Market sizing (total addressable revenue in a niche -- see our digital product market size report for the full picture)
What our estimates don't tell you:
- Exact dollar amounts (we aim for +/-20%, not +/-1%)
- Net profit (we don't know expenses)
- Revenue from bundles, upsells, or external sources
- Currency exchange fluctuations
Factors that can reduce accuracy:
- Products with very few reviews (less signal = less accuracy)
- Very new products (not enough historical data)
- Products with unusual pricing (bundles, pay-what-you-want)
- Heavy promotional periods (temporary spikes)
What Makes Our Approach Different
We're not guessing.
Every estimate is backed by multiple data points and validated against known benchmarks. When confidence is low, we show it.
We show the work.
Unlike black-box tools, we explain our methodology publicly. You can evaluate whether our approach makes sense for your use case.
We continuously improve.
Our models are updated weekly based on new data, creator feedback, and market changes. Accuracy improves over time.
We're purpose-built for infoproducts.
Tools designed for Shopify or Amazon don't work well for digital products. The signals are different. The business models are different. We focus exclusively on the infoproduct market. Creators use our estimates to evaluate everything from profitable niches in 2026 to individual product performance.
How We Handle Edge Cases
Products with no reviews:
We rely more heavily on ranking and traffic data. Confidence levels are lower, and we flag these estimates accordingly.
Pay-what-you-want pricing:
We use average transaction values from similar products to estimate, but flag higher uncertainty.
Products on multiple platforms:
We currently estimate per-platform. Cross-platform aggregation is on our roadmap.
Bundles and upsells:
We estimate based on the primary product price. Revenue from upsells isn't captured, which means our estimates may be conservative for creators with sophisticated funnels.
The Bottom Line
Our revenue estimates aren't perfect -- no estimation can be. But they're accurate enough to answer the questions that matter:
- Is there money in this niche?
- What's the revenue ceiling for top products?
- Is this market growing or shrinking?
- How does my product compare to competitors? (For a complete approach to answering this question, see our competitor analysis framework.)
That's the information you need to make smart decisions about where to invest your time and creativity. Without it, you're guessing. And guessing is how 95% of creators fail.
We believe in transparency. If you have questions about our methodology, reach out anytime at contact (at) insightraider.com.
You just read how we estimate revenue. Now see it in action -- enter any niche and get estimated monthly revenue, trend data, and competitor benchmarks for the top products. Join 100 early adopters and put our methodology to work for your next product decision.
Got more questions? Check out our data-driven answers:
