JapanRetail and e-commerce2023 – ongoingSyuppin
AIMD automatic purchase and sale pricing for used cameras
Syuppin revised camera prices about six times as often after AutoML models set prices without manual checks
Problem
Staff at Syuppin's Map Camera platform set purchase and sale prices by hand from market trends and past sales. Prices were revised too slowly. The company had worked with several vendors for about three years without reaching its goal.
Approach
Syuppin built AIMD with the consulting firm SIGMAXYZ, using Vertex AI AutoML. The models predict the number of sales and then adjust purchase and sale prices from it. Prices go to the websites and stores without a manual pre-check.
Results
Takeaway
The company dropped manual pre-checks to speed up revisions. It limits automatic pricing to items with enough trading data, and new products with little history remain a gap.
Categories
- Industry
- Retail and e-commerce
- Business function
- Marketing and sales
- Innovation type
- Process innovation
- AI technology
- Predictive analytics and ML
- Adoption
- Partner co-development
- Evidence
- Vendor case study
Sources
Source summaries
Google Cloud's customer story describes how Syuppin built the AIMD pricing system on Vertex AI AutoML for its used camera business. It reports about six times more price revisions, about 1,250 items under automatic revision, and wish-list emails rising from hundreds to about 40,000 a day. The page carries no date; the Internet Archive first saved it on January 29, 2023.
Full details are in the original sources.
