One of the first things I do when researching a new game idea is look at what comparable titles have sold. Not because I want to copy someone else's success, but because you can't plan a budget, set expectations, or evaluate a market without some sense of the money involved. The problem is that Valve doesn't publish sales numbers. So we're all stuck estimating.
The good news is that several methods exist for estimating Steam game sales, and they've gotten better over the years. The bad news is that none of them are precise, and each has blind spots that can lead you astray if you don't understand their limitations.
This guide compares the three most common approaches: the Boxleiter method, SteamSpy, and VGInsights. I'll walk through how each works, where each fails, and how to use them together to get the most reliable picture.
Why estimating Steam sales matters
If you're building an indie game, you're running a business whether you think of it that way or not. Estimating competitor sales isn't about envy -- it's about making smart decisions.
Competitive research
Before you commit to a genre, you should know what other games in that space are earning. If every roguelite colony sim from the last two years has 200+ reviews, the market has demand. If the top games have 30 reviews each, the audience might not be there -- or the genre might be so new that you'd be taking a bigger bet. Our indie game revenue data guide covers the benchmarks by genre, but you'll want to run estimates on specific titles to validate what you're seeing.
Market sizing
How big is the total addressable market for your type of game? You can answer this roughly by estimating the sales of the top 20-30 comparable titles and adding them up. If the whole genre does $10 million a year across all titles, you're fighting for a slice of $10 million -- not $100 million.
Financial planning
If you're taking out a loan, pitching a publisher, or deciding whether to quit your job, your revenue projections need to be grounded in something real. "I think my game will do well" isn't a plan. "Games comparable to mine typically earn $80K-$200K in their first year, and here's how I got that number" is a plan. Use our Revenue Calculator to build those projections once you've estimated your comparables.
Method 1: The Boxleiter method
The Boxleiter method is the most widely used approach for estimating Steam game sales. It's free, fast, and you can apply it to any game with visible reviews. Named after developer Jake Birkett, who refined earlier work by Simon Carless, it uses the publicly visible review count as a proxy for total sales.
How it works
The core formula is straightforward:
Estimated units sold = Number of reviews x Multiplier
The multiplier represents the ratio of total buyers to review-leavers. Most players never leave a review. The question is: how many buyers does each review represent?
- •Typical range: 20-60 sales per review
- •Common starting point: 30x for most indie games
- •Higher multipliers (40-60x): Casual audiences, very low price points, large non-English player bases, and free-to-play games where the barrier to "buying" is zero but the barrier to reviewing stays the same
- •Lower multipliers (20-30x): Niche genres with engaged communities, PC-first strategy or simulation games, higher price points where buyers feel more invested and review more often
To estimate revenue rather than just units, you extend the formula:
Estimated gross revenue = Reviews x Multiplier x Price
Worked example
Let's say you're researching a survival crafting game priced at $19.99 with 1,200 reviews.
- •Conservative (25x): 1,200 x 25 = 30,000 units. Revenue: 30,000 x $19.99 = $599,700 gross
- •Mid estimate (35x): 1,200 x 35 = 42,000 units. Revenue: 42,000 x $19.99 = $839,580 gross
- •Optimistic (50x): 1,200 x 50 = 60,000 units. Revenue: 60,000 x $19.99 = $1,199,400 gross
After Steam's 30% cut, the developer's net ranges from roughly $420K to $840K. That's a wide range, which is the Boxleiter method's biggest limitation: the multiplier is a guess, and small changes in the multiplier produce large changes in the estimate.
Strengths
- •Completely free and requires no tools beyond basic math
- •Works on any game with visible reviews
- •Quick to apply -- you can estimate a dozen competitors in ten minutes
- •Well-understood in the indie community, so other developers know what you mean when you reference it
Weaknesses
- •The multiplier varies significantly and you're never sure which value to use
- •Doesn't account for regional pricing (a $19.99 game might average $12 per sale once you factor in lower regional prices)
- •Bundle sales, deep discounts, and giveaways inflate the review count without contributing proportional revenue
- •Review bombing or abnormal review patterns distort the estimate
- •Free-to-play and free weekend promotions break the model entirely
Despite those limitations, the Boxleiter method remains the best starting point for anyone doing quick competitive research. Our Revenue Calculator automates this with genre-adjusted multipliers so you don't have to guess the ratio.
Method 2: SteamSpy
SteamSpy, built by Sergey Galyonkin (who later went on to create the Epic Games Store), was the gold standard for Steam sales estimation from 2015 until Valve changed its privacy defaults in 2018. It still exists and still provides data, but its accuracy has taken a real hit.
How it works
SteamSpy originally scraped public Steam profile data to count how many accounts owned a given game. If 50,000 public profiles owned a game, and a known percentage of profiles were public, SteamSpy could extrapolate total ownership with reasonable confidence.
After Valve made game libraries private by default in April 2018, SteamSpy lost access to the majority of its data source. Galyonkin adapted the methodology to use a combination of remaining public profiles, player count data, review analysis, and statistical modeling. The site now shows ownership ranges (e.g., "200,000 - 500,000 owners") rather than point estimates.
What it tracks
SteamSpy provides several data points for each game:
- •Owners (range): Estimated total accounts that own the game
- •Players (range): Estimated accounts that have played the game in a given period
- •Median playtime: How long the median owner has played
- •Peak concurrent players: Historical CCU data
- •Review data: Total reviews and positive percentage
Accuracy limitations
Here's where I have to be honest: post-2018 SteamSpy data should be treated as directional, not precise. The ownership ranges are often extremely wide. Seeing "200,000 - 500,000 owners" tells you a game is moderately successful, but the difference between 200K and 500K units at $14.99 is over $3 million in gross revenue. That's not a rounding error.
SteamSpy also doesn't distinguish between paid purchases, free keys, bundle activations, and free weekend trials. A game that was in a Humble Bundle and had two free weekends might show a much higher ownership number than its actual paid sales would suggest.
For games with very low sales (under 20,000 owners), SteamSpy's estimates become especially unreliable because the sample size of remaining public profiles is too small for confident extrapolation.
Strengths
- •Free to use (with optional Patreon support for more detailed data)
- •Covers virtually every game on Steam
- •Provides useful supplemental data like playtime and player trends
- •Good for identifying broad ownership tiers (is this a 50K game or a 500K game?)
Weaknesses
- •Ownership ranges are very wide since the 2018 privacy changes
- •Cannot distinguish paid sales from free keys, bundles, or giveaways
- •Low-sales games have especially unreliable estimates
- •Data updates can lag behind reality
- •No revenue estimates -- you still need to multiply ownership by an assumed average sale price, which introduces another layer of uncertainty
Method 3: VGInsights
VGInsights is a paid analytics platform that aims to provide more accurate and granular Steam sales data than the free alternatives. It's used by publishers, investors, and market analysts, and increasingly by indie developers who want better data for business planning.
How it works
VGInsights uses a proprietary methodology that combines multiple data signals: review counts, player activity data, concurrent player numbers, wishlist tracking, historical pricing data, and machine learning models trained on known sales figures from developers who've shared their data privately or publicly.
The key difference from the Boxleiter method is that VGInsights doesn't rely on a single multiplier. Instead, it builds game-specific models that account for genre, price history, regional pricing, discount frequency, and other factors that affect the review-to-sales ratio.
What it provides
VGInsights offers considerably more detail than the free alternatives:
- •Revenue estimates: Both gross and net, accounting for regional pricing and discount history
- •Units sold estimates: With tighter confidence ranges than SteamSpy
- •Revenue over time: Monthly and weekly breakdowns showing launch spikes, sale events, and long-tail performance
- •Genre and tag analytics: Market-level data showing revenue by genre, trending tags, and saturation metrics
- •Comparable game analysis: Tools for finding and comparing games similar to yours
Accuracy claims
VGInsights claims their estimates are within 20-30% of actual figures for most games, and they've published case studies where developers confirmed their estimates against real Steamworks data. That's meaningfully better than the Boxleiter method's variance, but still not precise -- a 25% error on a game that earned $200K means the estimate could be off by $50K.
In my experience, VGInsights tends to be more accurate for mid-tier games (1,000-50,000 units sold) than for either very small or very large titles. Very small games don't generate enough data signals for the model to work well, and very large games with complex pricing histories, multiple editions, and DLC create noise that's hard to filter.
Strengths
- •Most accurate of the three methods for most games
- •Accounts for regional pricing, discount history, and bundles
- •Provides revenue over time, not just lifetime totals
- •Market-level analytics useful for genre research and market sizing
- •Regularly validated against known data
Weaknesses
- •It costs money (subscription plans start around $15-30/month depending on tier)
- •Still an estimate -- 20-30% accuracy means significant uncertainty on individual titles
- •Accuracy drops for very small or very new games
- •The proprietary methodology means you can't fully verify how estimates are derived
- •Some features require higher-tier plans
Head-to-head comparison
Here's how the three methods stack up across the dimensions that matter most.
Accuracy
VGInsights leads here, with estimates generally within 20-30% of actual figures. The Boxleiter method can be within 20% if you get the multiplier right, but the multiplier itself is a guess, so real-world accuracy often ranges from 30-50% variance. SteamSpy falls last since the 2018 changes, with ownership ranges sometimes spanning a 2-3x difference between the low and high end.
Cost
Boxleiter method is free -- it's just math. SteamSpy is free for basic data, with more detail available through Patreon support. VGInsights requires a paid subscription, making it the most expensive option but still affordable relative to the decisions it informs.
Ease of use
Boxleiter method is the simplest: look up reviews, pick a multiplier, multiply. SteamSpy is also straightforward -- search for a game and read the data. VGInsights has a steeper learning curve due to its richer feature set, but the interface is well-designed.
Best use cases
- •Boxleiter method: Quick competitive scans, back-of-napkin estimates during brainstorming, conversations with other developers who use the same framework
- •SteamSpy: Identifying broad ownership tiers, tracking player trends and playtime data, supplementing other estimates with additional data points
- •VGInsights: Serious market research for business planning, publisher pitches, investor presentations, and any situation where you need the most defensible numbers available
How to triangulate estimates
No single method gives you a reliable number on its own. The smart approach is to use multiple methods and look for convergence.
Here's my process when researching a comparable title:
- 1.Start with the Boxleiter method using a conservative multiplier for the genre. This gives you a floor estimate. Our Revenue Calculator can do this automatically with genre-adjusted multipliers.
- 2.Check SteamSpy for the ownership range. If SteamSpy says 100K-200K owners and your Boxleiter estimate suggests 80K units, you're in the right ballpark. If SteamSpy says 500K-1M and your Boxleiter says 80K, something is off -- probably bundles or free promotions inflating SteamSpy's count.
- 3.Cross-reference with VGInsights if you have access. Their revenue estimate, which factors in regional pricing and discounts, often ends up lower than a naive Boxleiter calculation because the Boxleiter method assumes every unit sold at full price.
- 4.Look for developer disclosures. Some developers share their sales data publicly in postmortems, GDC talks, or social media posts. If you find one, use it to calibrate your estimates for similar games.
- 5.Average the estimates, weighting VGInsights highest if available, then Boxleiter, then SteamSpy. If all three roughly agree, you can have reasonable confidence. If they diverge wildly, dig into why before trusting any single number.
The goal isn't to land on a single precise figure. It's to define a credible range. "This game probably earned between $150K and $250K gross" is much more useful for planning than a false-precision point estimate of "$187,375."
Common pitfalls when estimating Steam sales
Even with multiple methods, several factors can throw off your estimates if you're not careful.
Regional pricing distortion
A game listed at $19.99 in the US might sell for the equivalent of $8-$12 in major markets like Brazil, Russia, Turkey, and parts of Southeast Asia. If 40% of a game's sales come from lower-priced regions, the actual average revenue per unit could be 25-35% below the US list price. The Boxleiter method completely ignores this. VGInsights tries to account for it, but even their adjustments are estimates. Our Steam pricing strategy guide covers regional pricing dynamics in detail.
Bundle and key sales
Games that have appeared in Humble Bundle, Fanatical bundles, or other key-selling platforms may have tens of thousands of additional owners who paid $1-$3 per game as part of a package. These owners still leave reviews and show up in SteamSpy ownership counts, but their revenue contribution is a fraction of a full-price sale. If you're estimating a game that's been in multiple bundles, you should discount your revenue estimate significantly.
Free-to-play and free weekends
Free-to-play games break the Boxleiter method because the "price" variable is zero. For F2P games, you'd need to estimate average revenue per user (ARPU) from in-game purchases, which is a completely different analysis. Similarly, games that have run free weekends or been given away temporarily will have inflated ownership and review counts relative to their actual paid sales.
Early Access distortion
Games in Early Access accumulate reviews over a longer period than traditional launches. A game that's been in Early Access for two years might have 500 reviews that accumulated slowly, while a game that launched fully and got 500 reviews in its first month had a much more concentrated sales pattern. The total sales might be similar, but the revenue timing and the effective multiplier can differ.
DLC and edition complexity
Games with multiple editions (Standard, Deluxe, Collector's) and extensive DLC catalogs make revenue estimation harder. The base game's review count tells you about base game ownership, but DLC revenue can equal or exceed base game revenue for some titles. Neither the Boxleiter method nor SteamSpy captures DLC revenue well.
Survivorship bias
When you look at comparable games, you're usually looking at games that succeeded enough to be visible. The games that failed hard and sold 50 copies are much harder to find in search results. This means your "comparable" set is naturally biased toward the more successful end of the spectrum, which can inflate your expectations.
Frequently asked questions
Which Steam sales estimation method should I use if I can only pick one?
The Boxleiter method is the best single method for most indie developers. It's free, fast, and widely understood. Use a conservative multiplier (25-30x) for the most defensible estimates, and always present your results as a range rather than a single number. If you're making a significant business decision -- quitting a job, signing a publishing deal, taking on debt -- invest in VGInsights for more reliable data. Our Revenue Calculator automates the Boxleiter method with genre-adjusted multipliers.
How accurate is the Boxleiter method for estimating Steam game sales?
In practical terms, the Boxleiter method is typically accurate within 30-50% of actual revenue if you choose a reasonable multiplier for your genre. The main source of error is the multiplier itself, which can vary from 20x to 60x depending on the game. Regional pricing is the second-biggest source of error, since the method assumes every unit sold at the listed USD price. For a more refined approach, use multiple estimation methods and look for convergence.
Can I estimate a Steam game's revenue if it's free-to-play?
Not with the standard Boxleiter method, since the price variable is zero. For free-to-play games, you'd need to estimate total players (using SteamSpy or SteamDB concurrent player data) and multiply by an assumed average revenue per user from in-game purchases. Typical ARPU for indie F2P games ranges from $0.50 to $3.00, but this varies enormously by game type and monetization design. VGInsights provides some F2P revenue estimates, but they're less reliable than their paid-game estimates.
How do I estimate sales for a game that hasn't launched yet?
You can't directly estimate sales for an unreleased game, but you can build projections by estimating sales of comparable games that have already launched. Identify 5-10 games similar to yours in genre, scope, and production quality. Estimate each one's revenue using the methods in this guide. Then assess where your game is likely to fall within that range based on your wishlist count, marketing effort, and store page quality. This won't give you a prediction, but it gives you a grounded range for planning. Check our indie game revenue data for genre benchmarks to supplement your comparables.
Ready to estimate revenue for your game or your competitors? Run your numbers through the Revenue Calculator with genre-adjusted Boxleiter multipliers, then read our indie game revenue data guide for full benchmarks by genre and our Steam pricing strategy breakdown to make sure your price point is optimized.
Browse our genre-specific optimization guides for strategies tailored to your game type, and check the Steam Page Leaderboard to see how top games optimize their store pages.