Why Enterprise Performance Management Needs Reform: A Data-Driven PDCA Architecture

Traditional enterprise performance management follows a hierarchical structure (supervisory, operational, management, execution layers) and a management process of strategy setting, goal definition, planning, tool selection, and method implementation, culminating in an operational cycle of task claiming, goal decomposition, responsibility clarification, process monitoring, and performance evaluation. However, this model faces four critical business pain points:

Inconsistent KPI definitions (口径杂) : Enterprise KPIs change frequently; business departments, assessment departments, and other units use different definitions. Each year indicators are redefined per strategy, often reused from prior years, with rules and algorithms distributed via Excel, leading to mixed definitions across the enterprise indicator system.

Slow information flow (信息慢) : Disconnection from operational data prevents timely KPI monitoring. Manual paper-based transmission introduces errors and delays; warning levels shift with management demands. Only calculated results are visible, not the process, so achievement risks cannot be exposed early.

Difficult improvement (改善难) : After risk identification, there is no grasp of corrective measure progress or risk closure. Abnormal indicators are distributed by KPI assessment departments to business units before monthly operational meetings, where responsible departments are named and reviewed, causing high latency and a fixed monthly cycle that precludes process monitoring and pre-emptive review.

Chaotic monitoring (监控乱) : Enterprise operational and KPI analysis lack system support; manual operations are slow and inefficient. Business leaders identify KPI risks and assign corrective actions but lack platforms to track measure progress, supervise execution, or feed back measure status and risk closure to leadership.

Existing solutions — manual performance management, commercial performance products, hybrid manual-product approaches, and self-developed systems — all fall short. Manual processes are cumbersome and slow; commercial products have misaligned system logic, complex management modes, and cannot integrate operational data, still requiring manual intervention; hybrid approaches suffer from weak institutional fit, messy manual data reconciliation, disconnection from operational meetings, and excessive subjective data; self-developed systems are difficult to build, have poor reusability, and involve massive effort for system initialization and indicator calculation reuse.

The proposed reform centers on real-time operational data-driven effective management , adopting a PDCA (Plan-Do-Check-Act) cycle to create a closed-loop, data-driven management platform. The architecture comprises three core centers:

1. KPI Maintenance Center (solves "Inconsistent Definitions")

Involves the assessment department and business departments. The assessment department maintains indicator basic information, target setting, data submission deadlines, version adjustments, and enables/disables assessment indicators. Business departments view claimed indicators and assessment logic, check submission deadlines, and inspect calculation logic to support underlying development. Key features include fuzzy query and multi-select filtering, filtering by enablement status and assessment method (default latest version year, historical versions queryable), batch export/import for modifications, pop-up indicator creation, per-indicator edit/view/delete, direct filtering by assessment department for submission, and direct selection/submission for deprecated indicator data.

Additional required functions: batch export/import, indicator editing, KPI viewing, KPI field option maintenance, and underlying key definitions.

2. Warning Item Management Center (solves "Slow Information")

Involves assessment department, business leaders, and business specialists. The assessment department defines indicator warning rules per annual strategy. Business leaders overview enabled indicators with warnings, edit/change warning content for upward reporting, and distribute warning indicators to specialists for content description or direct to-do creation. Business specialists edit/change warning content for reporting. The warning overview interface supports batch filtering of indicators by status, batch setting after selection, enabling/disabling warning function per indicator, and per-indicator warning rule editing.

Recommended additional functions: warning rule setting, warning to-do board, warning distribution, and warning handling.

3. To-Do Item Management Center (solves "Difficult Improvement")

Managed by business leaders and business specialists. Business leaders monitor key KPI achievement via the business operation center, initiate to-do items for unmet indicators, and track all departmental to-do items and their completion status. Business specialists execute and edit specific to-do items. The overview page provides multi-dimensional monitoring (all, incomplete, overdue incomplete, overdue completed, completed, closed), process tracking and result submission for each item, follow/unfollow actions displayed in personal portal, and unfollow for followed items.

To-do processing flows should be configured per enterprise requirements. This architecture aims to replace fragmented, manual, and delayed practices with a unified, real-time, data-driven platform that closes the loop from indicator definition through warning to corrective action tracking.