Government

Per-Mu Efficiency Evaluation Platform

Grades hundreds of industrial firms on output per unit of land: roster import, online filing, six-department review and automatic A–D grading, fully audited.

LiveGovernmentLarge projectPC · H5 · Dashboard

Client: Wenzhou Municipal Bureau of Economy and Information Technology
Industry: Government
Platforms: PC · H5 · Dashboard
Scale: Large project · 2018

01Background

Each year the Wenzhou Municipal Bureau of Economy and Information Technology runs a Per-Mu Efficiency Evaluation of several hundred industrial enterprises. (A mu is a Chinese land unit of about 667 m²; firms are graded on tax revenue, value added and similar output per unit of land.) The grades guide land allocation, policy support and risk monitoring. Company lists came from several agencies, firms reported offline, departments reviewed in separate spreadsheets and scoring was done by hand; the bureau needed one traceable online pipeline from list to final grade.

Client: Municipal government bureau (economy & IT)
Enterprises: Several hundred, above and below the statistical size threshold
Departments: Sub-districts plus tax, land & resources, planning, power, statistics, environment
Platforms: PC console, Zheliban H5, Zhezhengding, dashboard

02Challenges

1Scattered lists

Statistics, market regulation and tax each kept a list; sub-district surveys added duplicates and gaps.

2One reporting channel

Firms could only submit on PC or offline, so smaller firms rarely took part.

3Siloed reviews

Tax, land, power and environment data moved in separate spreadsheets with inconsistent definitions.

4Opaque grading

Weighted indicators plus bonuses, penalties, upgrades and vetoes were error-prone and hard to explain.

Scoring was never the hard part: sourcing the list, coordinating reviews and explaining results were.

03Solution

Shanku built an evaluation platform where the PC console, the Zheliban app for enterprises and the Zhezhengding app for officials share one backend. Weights, industry benchmarks and adjustment rules live in parameter tables, so annual changes rarely need code.

De-duplicated roster

Agency lists are merged and de-duplicated; sub-districts confirm size and sector, upload exemption proof and transfer firms between districts.

Online filing via Zheliban

Firms sign in with their unified social credit code; actual land use, plot ratio, power use and tax contribution are derived on save.

Six-department review

Tax, land, planning, power, statistics and environment each review their own data or bulk-import Excel; approval triggers a recalculation.

Automatic scoring

Indicators are scored against industry benchmarks and weights; separate A–D cut-offs for larger and smaller firms; batch runs and rollback.

Codified adjustments

Bonuses capped at 20 points, penalties at 40; upgrades, holds, downgrades and vetoes run in a fixed order with no grade skipping.

Yearly isolation & audit

Data is partitioned by evaluation year and read-only once archived; logs record operator, action, firm and channel.

04Core workflow

Evaluation pipeline

1
Import roster
Merge agency lists into one register
2
District survey
Confirm size, sector and eligibility
3
Firm filing
File on Zheliban; derived fields auto-filled
4
Joint review
Six departments review; approval recalculates
5
Score & grade
Weighted scoring, A–D cut-offs
6
Adjust
Bonuses, penalties, upgrades, vetoes in order
7
Report & archive
Word report export, year archived

A firm enters the scoring queue only after its sub-district and every department have approved; each score and adjustment carries a written rationale for firms and auditors.

05Results

One traceable pipeline

Every step from roster to final grade records status, operator and notes, replacing spreadsheet reconciliation.

Explainable results

Each firm’s indicator values, item scores, adjustments and grading reasons are written out and ready to share.

Configurable rules

Weights, benchmarks, grade ratios and adjustment rules are maintained as parameters, not code.

Mobile coverage

Firms file from their phones on Zheliban; officials work on Zhezhengding with the same permissions as on PC.

06Screens

Screens are prototypes or rebuilt with sample data. They contain no real client data, amounts or accounts.

07Tech stack

Backend: Spring BootMyBatis PlusScoring engine / scheduled jobs
Front end: Vue 3Element PlusVite
Mobile: H5 filing appZhelibanZhezhengding
Data & integration: MySQL (SSL)Excel bulk import / exportWord report templates
Deployment: Government intranetLinux + Tomcat