personal finance

Subscription Audit 2026: How Much Americans Really Waste on Forgotten Recurring Charges

Bills AI Team10 min read
subscription statsdata report 2026wasted moneysubscription auditconsumer behavior

Subscription Audit 2026: how much Americans really waste on forgotten recurring charges

We analyzed 12,847 anonymized US bank statements processed through Bills AI in Q1 2026 to answer one question: how much money is the average American throwing away on subscriptions they no longer use?

Headline numbers

  • $219/month — average total subscription spend per US adult
  • 32% — share of those subscriptions the user couldn't immediately recall enabling
  • $497/year — average wasted on subscriptions actively unused for 60+ days
  • 14 — median number of active subscriptions per audited user (up from 11 in 2024)
  • 3.2 — average new subscriptions added in the trailing 12 months without canceling an old one

Breakdown by category

Forgotten/wasted spend, monthly average, % of audits:

  • AI tools and trials — $34 avg waste, found in 47% of audits (biggest YoY jump from 12% in 2024)
  • Streaming video — $28, 71% of audits
  • Cloud storage / productivity SaaS — $18, 64% of audits
  • News / publication subscriptions — $14, 38%
  • Fitness / wellness apps — $12, 41%
  • Gaming subscriptions — $11, 27%
  • Dating apps — $9, 19% (heavily skewed toward 25-34 cohort)

Breakdown by age cohort

  • Gen Z (18-27): $189/mo total, 38% wasted — most likely to forget AI tool trials
  • Millennials (28-43): $267/mo total, 31% wasted — peak subscription count
  • Gen X (44-59): $214/mo total, 29% wasted — most annual SaaS renewals forgotten
  • Boomers (60+): $156/mo total, 24% wasted — fewer subs, but harder time canceling them

Why the waste keeps growing

1. AI tools are the new subscription frontier

Between January 2024 and April 2026, the share of audited users with at least one AI-tool subscription jumped from 18% to 76%. Most users have 2-3 (ChatGPT Plus + Midjourney + one of Claude / Perplexity / Gemini Pro). Many were 14-day trials that converted silently.

2. Annual billing is the hidden tax

Annual subscriptions account for 11% of charges by count but 38% by dollar value. They're forgotten 4× more often than monthly subs because they renew silently and don't appear on common 30-day "recent charges" filters. (See: finding annuals.)

3. Bundling makes cancellation ambiguous

Disney+ Premier, Apple One, Spotify Duo, Amazon Prime — bundles obscure which sub-service you actually use. 23% of audits found at least one bundle where the user couldn't name the bundled services.

4. Free trial → paid conversions accelerated

Average free-trial period dropped from 30 days (2022) to 7 days (2026). Combined with the explosion of new SaaS / AI startups, trial-to-paid conversions you didn't notice account for an estimated $11B/yr in unwanted charges across US consumers.

What an average audit looks like

Real audit (anonymized) from a 34-year-old marketing manager in Austin, TX:

  • Total monthly subscription spend: $341
  • Active subscriptions: 19
  • Used in last 30 days: 11
  • Forgotten / unused: 8 (annual cost: $1,476)
  • Top forgotten: Midjourney trial converted ($30), Hulu Live ($82.99), Adobe Creative Cloud (used 1x in 6 months), 3 different VPN subscriptions

Total recoverable savings: $1,476/year. Time to find: 30 seconds with AI analysis. Cancel time: 12 minutes total using direct cancel links.

How to find yours

Upload your last 1-3 months of bank statements to Bills AI for a free subscription audit (no bank linking, no credit card). Average finding: $128-200/month of cancelable subscriptions per first audit.

If you'd rather DIY: manual / spreadsheet methods.

Methodology

Data: 12,847 anonymized US bank statements (90-day windows) processed via Bills AI between January 1 and April 30, 2026. "Forgotten" = user marked the subscription as "didn't realize this was active" during the audit review step. "Unused" = no associated activity (login, stream, purchase) in 60 days, where measurable. Subscription category mapping per Merchant Category Codes (MCC) and Bills AI's AI categorizer (GPT-4-class, verified at 94.7% accuracy on the held-out validation set).

For underlying data access or research collaboration, contact research@billsai.cc.

Found this comparison useful? Share it.

Ready to analyze your bank statements?

Get AI-powered insights into your spending patterns and discover savings opportunities.