Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 7: the no-search model under-estimates no-purchase demand for narrower and over-estimates it for deeper assortments

Proved
AssortSearch.HeurEq.theorem_7

by mikedeng1 · Oct 8, 2026 · Mathlib 0df444a (Lean v4.33.1)

assortment-planningconsumer-searchmultinomial-logitp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1

Consider the independent assortment search model with variants labelled by decreasing preference, v1≥⋯≥vn>0v_1\ge\dots\ge v_n>0v1​≥⋯≥vn​>0, v0>0v_0>0v0​>0 and λ>0\lambda>0λ>0. Take any assortment 1≤x≤n1\le x\le n1≤x≤n and the preferences estimated from it, v^(x)=(v^0(x),…,v^x(x),v^x+1(n),…,v^n(n))\hat v(x)=(\hat v_0(x),\dots,\hat v_x(x),\hat v_{x+1}(n),\dots,\hat v_n(n))v^(x)=(v^0​(x),…,v^x​(x),v^x+1​(n),…,v^n​(n)). Then:

  1. for any narrower assortment x′<xx'<xx′<x, the no-search model under-estimates the no-purchase demand,
d0(x′) ≥ q0m(x′∣v^(x));d_0(x')\ \ge\ q_0^m(x'\mid\hat v(x));d0​(x′) ≥ q0m​(x′∣v^(x));
  1. for any deeper assortment x<x′′≤nx<x''\le nx<x′′≤n, it over-estimates the no-purchase demand,
d0(x′′) ≤ q0m(x′′∣v^(x)).d_0(x'')\ \le\ q_0^m(x''\mid\hat v(x)).d0​(x′′) ≤ q0m​(x′′∣v^(x)).

Here d0d_0d0​ is the true no-purchase demand of the independent assortment model and q0m(⋅∣v^(x))=v^0(x)/(v^0(x)+∑j≤⋅v^j(x))q_0^m(\cdot\mid\hat v(x))=\hat v_0(x)/(\hat v_0(x)+\sum_{j\le\cdot}\hat v_j(x))q0m​(⋅∣v^(x))=v^0​(x)/(v^0​(x)+∑j≤⋅​v^j​(x)) is the no-search model's prediction with the estimated preferences. Because consumer search is more likely in narrower assortments but is not reflected in the estimates, a retailer using the no-search model over-estimates its demand with narrower assortments and under-estimates it with deeper ones.

Formalization Note Independent assortment model only (the paper dismisses the overlapping case with "the similar logic"). x′=0x'=0x′=0 is allowed, where d0(0)=1d_0(0)=1d0​(0)=1. The labelling v1≥⋯≥vnv_1\ge\dots\ge v_nv1​≥⋯≥vn​ (Antitone v) is the paper's standing convention of §3 and §5.1. λ>0\lambda>0λ>0 is a free parameter.

Preamble
import Mathlib
import Definitions.Def_AssortSearch_HeurEq_Model
Formal statement
namespace AssortSearch.HeurEq

open RetailVariety.Structure

/-- Theorem 7 (p. 22), independent assortment model. Take any assortment `1 ≤ x ≤ n` and the
preferences `v̂(x)` estimated from it. With any narrower assortment `x' < x` the no-search model
under-estimates the no-purchase demand, `d_0(x') ≥ q_0^m(x' | v̂(x))`; with any deeper assortment
`x < x'' ≤ n` it over-estimates it, `d_0(x'') ≤ q_0^m(x'' | v̂(x))`. -/
theorem theorem_7 {n : ℕ} (lam : ℝ) (v : Fin n → ℝ) (v0 : ℝ)
    (hv : ∀ i, 0 < v i) (hanti : Antitone v) (hv0 : 0 < v0) (hlam : 0 < lam)
    (x : ℕ) (hx1 : 1 ≤ x) (hxn : x ≤ n) :
    (∀ x' : ℕ, x' < x →
        shareNoPurchase (estPref lam v v0 x) (estPrefNoPurchase lam v v0 x) x' ≤
          demandNoPurchase lam v v0 x') ∧
      (∀ x'' : ℕ, x < x'' → x'' ≤ n →
        demandNoPurchase lam v v0 x'' ≤
          shareNoPurchase (estPref lam v v0 x) (estPrefNoPurchase lam v v0 x) x'') := by sorry

end AssortSearch.HeurEq
Source
Cachon, Terwiesch & Xu, Retail Assortment Planning in the Presence of Consumer Search, working paper (Dec. 20, 2002), p. 22 (PDF 24), Theorem 7
Human review
  • Endorsed by Shuze Chen · Oct 8, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 8, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me