/ Library/ Not Enough Meaning/ Subadditivity Effect
Connect Bias № 141 · Last updated 6 June 2026

Subadditivity Effect.

"The bundle feels cheaper than the parts — until you price the parts separately wrong."

01Overview

Subadditivity effect describes underestimating combined quantities — probabilities, values, times — when they are grouped. Bundles feel like deals; joint probabilities feel smaller than parts; project timelines underestimate when split vs summed.

Designers bundle features for pricing psychology — subadditivity can help ethical clarity or obscure costs. Misapplied, it breaks user mental models: insurance add-ons, suite plans, and roadmap estimates that parts "should" add but psychologically do not.

02Detailed explanation

Subadditivity appears in pricing and planning:

  • Feature bundle priced below à la carte sum — feels like win; users skip items they would not buy alone.
  • Joint failure probability judged lower than sum of risks — planning fallacy cousin.
  • Time estimates for subtasks sum low; total sprint overflows.
  • Multi-peril insurance uptake patterns reflect subadditive probability judgment.

Conjunction fallacy misjudges joint likelihood of vivid stories; subadditivity is broader undervaluation of aggregates — related errors in probability and value.

03Why it exists

Heuristic averaging replaces addition under cognitive load.

Bundles simplify choice — brain rewards simplicity with undervaluation of omitted parts.

The short version

When users buy the bundle, do they want the whole — or are parts subadditive masking waste?

04Effects on users

Users buy bundles containing unwanted parts because subadditive pricing feels efficient — later feel cheated when unpacking value.

They underestimate compound risk when flows split permissions — each feels small.

05Effects on designers & teams

Teams use subadditivity in packaging and estimates:

  • Bundle anchoring. Decoy parts inflate deal feeling.
  • Roadmap summation error. Tasks underestimate in parallel.
  • Risk dashboards. Siloed low risks ignore aggregate.
  • Terms fragmentation. Small consents feel harmless together.

06Practical takeaways

  • Show bundle breakdown. Transparent part value builds trust.
  • Add probabilities correctly. Risk comms for compound events.
  • Planning: sum estimates explicitly. Fight subadditive scheduling.
  • Offer unbundle. Ethical choice when bundle is padded.
  • Test willingness to pay per part. Not only bundle.
  • Audit permission creep. Aggregate data ask scope.

07Design examples

Pricing

Suite illusion

Bundle lists five tools; users need one. Subadditive price feels cheaper than single SKU — later churn when unpacking reveals poor single-tool value.

Planning

Six one-week tasks

Each workstream estimates one week; parallel owners. Ship slips six weeks — subadditive time judgment at portfolio level.

Risk

Three low risks

Security reviews three 2% risks separately; aggregate outage underestimated. Incident post-mortem cites subadditivity in planning.

Consent

Three small asks

Users accept location, contacts, analytics in separate sheets. Combined surveillance surprise — each felt minor; aggregate did not.

08Ethical risks

Exploiting subadditivity to hide aggregate cost or data grab is bundling dark pattern.

Underestimating aggregate harm probability delays systemic fixes users need.

Self-test: What is the true sum of the parts users think they are getting — and does your bundle copy show it?

10Suggested reading