MWM
Contact salesMWM AI
Rankings & Market Intelligence

Revenue Estimation

Also known asRevenue ModelRevenue Estimate

Third-party modelled estimates of an app's total revenue (paid downloads + IAP + subscriptions) based on rank signals, panel data, and price tiers.

MWM data

State of April 2026

Apps that monetize at all

26.9%

Most active apps generate near-zero measurable revenue

Median monetizing-app revenue (30d)

$569

Mid-tier reality for apps that do monetize

Top-10% monetizing-app revenue

$30K

Strong performers — but still long-tail to the top

Top-1% monetizing-app revenue

$800K

Where most absolute revenue lives — power-law concentration

Key takeaways

  1. 01Apple and Google don't publish app-level revenue — every public number is a model output.
  2. 02Three inputs drive every credible estimator: chart rank, panel-data sampling, and calibration against apps with published financials.
  3. 03Different providers disagree by 20-50% on absolute revenue for the same app — directional accuracy is much higher than absolute.
  4. 04Use revenue estimates for trend / growth / cross-country comparison; don't treat single-app dollar figures as ground truth.

Apple and Google publish chart positions, download counts (Play Console — to the developer only), and category leaderboards. They do not publish app-level revenue. Every revenue number you see in a competitive-intelligence tool — MWM, Sensor Tower, data.ai (formerly App Annie), Apptopia — is a model output, not an observed value.

Three inputs drive every credible revenue estimator

  1. Chart rank as a continuous signal. Rank-to-download relationships are well-studied; combined with category-typical monetization (ARPU, IAP intensity), rank yields a first-order revenue estimate.
  2. Panel data — a sample of real users (typically 1-5 million worldwide on consented devices) whose actual app purchases are measured. The panel anchors the rank-based model to real revenue values for sampled apps.
  3. Public-financials calibration. Public companies (Roblox, Match Group, Bumble, Spotify) and apps that disclose revenue in investor decks / legal filings provide ground-truth anchors the model can calibrate against. The denser and more recent the anchor set, the tighter the estimates.

The accuracy reality: providers disagree by 20-50% on absolute revenue for the same app, even in the same month. Reasons: different panel composition, different rank-to-revenue calibration coefficients, different treatment of returns / refunds, different inclusion of subscription deferred revenue. Directional accuracy (is this app growing? is the trend positive?) is much higher — typically 80-95% accurate for month-over-month trend direction.

How to use revenue estimates responsibly

  • Use for trend ("is this app growing 30% YoY?"). High accuracy here.
  • Use for cross-market comparison ("how does this app compare in US vs Japan?"). Accuracy depends on panel depth per country.
  • Use for competitive benchmarking ("are we above or below the category median for our rank?"). Use rank-bucket aggregates, not single-app numbers.
  • Avoid for absolute dollar reporting to executives. Caveat that the number is modelled with ±30% uncertainty.
  • Avoid for legal / financial decisions (acquisition due diligence, deal pricing). Get the target company's actual financials.

A note on MWM's approach: we publish modelled revenue with explicit confidence bands rather than point estimates, and we anchor against an unusually large public-financials set. Even so, treat any single absolute number as ±20-30% with high frequency. Compare to other providers when stakes are high.

What apps actually earn — the long tail

Only 26.9% of apps monetize at all, and among those that do, the revenue distribution is a brutal long tail. The percentile table shows how far apart the median and the top decile sit in every category.

Mobile app revenue reality — the long tailDistribution of 30-day modelled revenue across apps with ≥100 downloads. The "$0 (no rev)" bucket dominates by count; most monetizers earn small amounts; the productive tail is steep and short.0125K250K375K500K$0 (no rev): 333,865<$100: 33,602$100-1K: 37,177$1K-10K: 30,328$10K-100K: 15,412$100K-1M: 5,342$1M+: 1,009Majority of catalog$0 (no rev)<$100$100-1K$1K-10K$10K-100K$100K-1M$1M+Monthly revenue (USD)
Mobile app revenue reality — the long tail — Apps with ≥100 downloads in last 30 days, MWM catalog, State of April 2026.

Monthly revenue percentiles by category (monetizing apps)

CategoryMedian monthly revenueTop-10% revenue
Social & Communication$1K$40K
Lifestyle & Well-being$754$20K
Productivity & Tools$593$20K
Media & Entertainment$581$30K
Game$463$40K
Education & Knowledge$421$10K

In most categories the median monetizing app earns four figures a month while the top decile earns five to six — an order-of-magnitude gap. Revenue "estimates" that quote averages hide this; the percentile shape is the honest view.

Quick answers

Where do app revenue estimates come from if Apple and Google don't publish them?

Every credible estimator combines three inputs: (1) chart rank as a continuous signal mapped to category-typical monetization, (2) panel data — sampling real-user app purchases on consented devices (typically 1-5M users worldwide), (3) calibration against apps with published financials (public companies, investor decks, legal filings). The blend produces a modelled estimate, not an observed value.

How accurate are mobile app revenue estimates?

**Directional accuracy** (growth direction, trend, growth rate) is typically 80-95% for established providers. **Absolute accuracy** (dollar figures) is much weaker — providers disagree by 20-50% on the same app, sometimes more for non-US markets or apps with unusual monetization. Use estimates for trend and competitive comparison, not for legal / financial decisions where the actual financials matter.

Why do Sensor Tower, data.ai, and Apptopia show different revenue numbers for the same app?

Different panels (which users they sample), different rank-to-revenue calibration coefficients, different anchor sets (which public financials they reference), different treatment of returns, refunds, and subscription deferred revenue. None is "right" in an absolute sense — each is a different model fit against partial data. For high-stakes analysis, look at multiple sources and use the spread to estimate uncertainty.

Can revenue estimates tell me how much my competitors earn?

Directionally yes, absolutely no. You can know with high confidence whether a competitor is growing or declining, whether they're outperforming category peers, and how their geographic split compares to yours. You cannot reliably know that "they made exactly $4.2M last month" — that number from any single provider carries ±20-30% uncertainty and could be materially different at a different provider.

Next reads

Back to glossary