Back to History Current Version

Raising Retail Productivity through Data Pooling: A Mobile Phone App Intervention

Last registered on August 08, 2025
View Trial History

Pre-Trial

Trial Information

General Information

Title
Raising Retail Productivity through Data Pooling: A Mobile Phone App Intervention
RCT ID
AEARCTR-0016469
Initial registration date
August 04, 2025

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
August 08, 2025, 6:55 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Primary Investigator

Name
Ajay Shenoy
Affiliation
U.C. Santa Cruz
Contact Primary Investigator

Other Primary Investigator(s)

PI Name
David Park
PI Affiliation
KDI School of Public Policy and Management
Contact Investigator
PI Name
Brenda Samaniego de la Parra
PI Affiliation
University of Notre Dame
Contact Investigator
PI Name
Emma Riley
PI Affiliation
University of Michigan
Contact Investigator

Additional Trial Information

Status
In development
Start date
2025-08-07
End date
2026-06-20
Keywords
Firms & Productivity
Additional Keywords
SMEs; retail productivity; data pooling; artificial intelligence; Zambia
JEL code(s)
O12; D22; L26; O33
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
This study examines whether limited access to high‑quality demand information constrains the growth of small retailers in developing countries and whether an AI‑driven mobile application that pools sales data across shops can overcome this barrier. One thousand Lusaka‑based retail shops will be randomly assigned to (i) an inventory‑management app only (control); (ii) the app plus product recommendations generated from data pooled across a relatively small number of similar shops (e.g., 10 shops); or (iii) the app plus recommendations generated from data on a larger number of similar shops (e.g., 100).. We will followup with the shops 4 times during and immediately after the intervention, and again 3 months later, in order to collect detailed data on adoption of the recommended products, weekly sales, and business income. Results will inform policies aimed at supporting SME productivity through digital data‑sharing platforms.
External Link(s)

Registration Citation

Citation
Park, David et al. 2025. "Raising Retail Productivity through Data Pooling: A Mobile Phone App Intervention." AEA RCT Registry. August 08. https://doi.org/10.1257/rct.16469-1.0
Sponsors & Partners
Experimental Details

Interventions

Intervention(s)
Intervention Components:
1. Inventory‑management mobile application (basic logging & analytics). Shops receive in‑person onboarding and three weeks of usage incentives to ensure high app engagement.
2. AI‑generated product recommendations based on pooled sales data from a small or large pool of similar shops.
3. Information about the size of the pool of shops used for generating the recommendations


Intervention Start Date
2025-09-15
Intervention End Date
2025-10-31

Primary Outcomes

Primary Outcomes (end points)
We will specify our primary and secondary outcomes in a pre-analysis plan, which will be lodged before followup data collection begins. We plan to capture data on the following:
Adoption of the recommended product
Sales & gross margin of recommended and existing products
Total sales revenue
Total business income / profit
Changes in suppliers or supplier search behaviour
Number of customers (footfall)
Number of distinct products stocked & inventory turnover
Introduction of other new products
Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
An individually randomised controlled trial with 1,000 shops divided equally across three arms.
Treatment Arms:
o T0 (Control) – App only
o T1 – App + recommendations from a smaller pool of comparable shops
o T2 – App + recommendations from a larger pool of comparable shops
Cross‑Randomisation: Within T1 and T2, shops are randomised as to whether the pool size (small vs large) is explicitly revealed.
Experimental Design Details
Not available
Randomization Method
Computer randomisation in Stata, using reproducible seed.
Randomization Unit
Individual firms
Was the treatment clustered?
No

Experiment Characteristics

Sample size: planned number of clusters
1000 firms
Sample size: planned number of observations
1000 firms
Sample size (or number of clusters) by treatment arms
T0 = 333 firms; T1 = 333 firms; T2 = 334 firms
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Supporting Documents and Materials
IRB

Institutional Review Boards (IRBs)

IRB Name
University of California Santa Cruz
IRB Approval Date
2025-06-25
IRB Approval Number
Protocol Number HS-FY2025-271
IRB Name
ERES Converge, Lusaka, Zambia
IRB Approval Date
2025-07-03
IRB Approval Number
Protocol Number 2025-Jun-038
IRB Name
University of Michigan
IRB Approval Date
2025-07-30
IRB Approval Number
HUM00276811
Analysis Plan